TASE - 1790

• (2025). A teleoperation control framework for a supernumerary robotic limb. TASE, 22, 23417--23429. (www)

  @article{ , abstract = {Supernumerary robotic limbs (SRLs) can significantly enhance human manipulation capability. However, it is difficult to achieve a safe and friendly interaction for collaboration tasks in complicated and dynamic external environments, and it easily leads to safety issues between a human user and the SRL robot. To address the aforementioned issues, this paper designs a new type of SRL and proposes a teleoperated control method for human-robot safe interaction. Specifically, inspired by the principle that human arms can autonomously adjust their stiffness to safely interact with the external environment, this paper proposes an SRL interaction control method based on the variable stiffness of the human upper limb. The proposed variable stiffness control parameters can be adaptively updated based on the characteristics of surface electromyography (sEMG) signals from the human upper limb. The effectiveness of the proposed framework is validated through experimental results. Note to Practitioners—This paper designs a novel SRL robot and develops a control framework for the SRL to improve safety and ease of interaction in human-robot collaboration tasks, especially in dynamic and complex environments. The proposed control framework has application potential in tasks requiring close human-robot interaction(HRI), such as collaborative assembly, object manipulation, and maintenance in hazardous or confined spaces. By using sEMG signals to adjust the robot’s stiffness in real-time based on human upper limb behavior, the method enables more intuitive and adaptive HRIs. This adaptive stiffness control can be particularly useful in applications like rehabilitation and assistive robotics, where safe and compliant interactions are essential.}, archive = {J_TASE}, author = {Jing Luo and Keao Wang and Tingyu Fei and Chao Zeng and Zhou Zhao and Jing Guo and Chenguang Yang}, doi = {10.1109/TASE.2025.3628247}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23417--23429}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A teleoperation control framework for a supernumerary robotic limb}, volume = {22}, year = {2025}, }
• (2025). Asynchronous dissipative control under adaptive event-triggered mechanism for nonhomogeneous markov jump systems with unknown probabilities. TASE, 22, 23406--23416. (www)

  @article{ , abstract = {This paper investigates the asynchronous dissipative control under adaptive event-triggered mechanism for nonhomogeneous Markov jump systems (MJSs) with unknown probabilities. The nonhomogeneous MJSs governed by time-varying transition probability matrix are modeled by polytope structure. A hidden Markov model (HMM) with unknown conditional probabilities (U-CPs) is established to characterize asynchronous relation between controller and system. To effectively save communication resources, an adaptive event-triggered mechanism (AETM) is utilized, which can adjust the triggering threshold in real time. Then, the criterion on ensuring both stochastic stability and strict $(Q,\;S,\;R)$ - $\alpha $ -dissipativity of the augmented closed-loop systems is established and mode-dependent asynchronous controller under AETM is designed. Finally, the effectiveness of the obtained results is verified by vertical take-off and landing (VTOL) helicopter model. Note to Practitioners—This paper deals with the asynchronous control problem for nonhomogeneous MJSs, which holds realistic significance in various control systems, such as helicopter systems, aircraft control systems and circuit systems. Considering the technical and resource constraints in practical engineering, we use a more general HMM to realize asynchronous control. In addition, the network environment has limited bandwidth, an AETM is designed to conserve communication resources and alleviate the network bandwidth burden. It is significant that the proposed asynchronous dissipative control method has promising applications in engineering, especially for systems with practical constraints and disturbances.}, archive = {J_TASE}, author = {Hongchao Li and Xin Liu and Yang Liu and Jiao Liu}, doi = {10.1109/TASE.2025.3628022}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23406--23416}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous dissipative control under adaptive event-triggered mechanism for nonhomogeneous markov jump systems with unknown probabilities}, volume = {22}, year = {2025}, }
• (2025). Hybrid memory-augmented neural control for real-time, model-free actuation of magnetic soft robots. TASE, 22, 23393--23405. (www)

  @article{ , abstract = {Magnetic soft robots have the potential to be used in biomedical applications, such as targeted drug delivery, minimally invasive surgery, and on-chip tissue manipulation, due to their untethered operation, rapid actuation, and physical adaptability. However, real-time control of these robots is challenging because of their inherent nonlinear dynamics, fabrication imperfections, and complex interactions with external magnetic fields. In this work, we present a model-free controller that uses Proximal Policy Optimization-collected experience and a hybrid architecture, EpisodicMemNet. A memory module returns stored actions for angle-matched states, while a multi-head network predicts actions when no match is found. In experimental validation on two magnetic soft robots, a four-legged dual-stem H-frame ( $6\times 10$ mm) and a three-legged asymmetric variant ( $8\times 11$ mm), an adaptive two-tier memory maintained a median lookup time of approximately 1.2 ms, and EpisodicMemNet achieved 87.4% balanced accuracy and 93.5% top-2 accuracy while sustaining 2 Hz real-time control. Furthermore, user-guided tasks such as target navigation, obstacle avoidance, and ramp climbing confirmed reliable performance and adaptability despite the system’s nonlinearities and data sparsity. The proposed method thus not only overcomes the limitations of traditional simulation-based and model-specific approaches but also paves the way for scalable, experience-driven control solutions in soft robotic applications. Note to Practitioners—Magnetic soft robots (MSRs) can be guided using external magnetic fields, removing the need for physical tethers and enabling tasks such as targeted drug delivery. Purely model-based control is brittle for MSRs in practice. Traditional reinforcement learning (RL) approaches for MSRs can require weeks of fine-tuning, extensive reward engineering, and domain expertise. This work explores the idea of using experience or example data to achieve real-time control. It combines an RL-based data collection process with a state-tolerance-based lookup and a neural network to approximate missing or uncovered experiences. The system reuses successful actions from memory and falls back to a classifier when no match exists. Practitioners can deploy this method using standard sensors and electromagnetic actuation hardware, as long as they collect sufficient training data under relevant conditions. In our tests, it met the above accuracy and latency targets without handcrafted policies. It has been validated on a symmetric dual-stem H-frame MSR and an asymmetric, ablated, three-legged MSR in various tasks such as target reaching, obstacle avoidance, and ramp climbing. Future work: richer locomotion primitives and additional actuation variables (e.g., frequency, waveform shape).}, archive = {J_TASE}, author = {Zakir Ullah and Dong Wang and Jinqiang Wang and Zixiao Zhu and Qing Xia and Peter B. Shull}, doi = {10.1109/TASE.2025.3628020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23393--23405}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hybrid memory-augmented neural control for real-time, model-free actuation of magnetic soft robots}, volume = {22}, year = {2025}, }
• (2025). Residual policy optimization with trust region constraints: A learning framework for stable and agile wheel-legged locomotion. TASE, 22, 23352--23365. (www)

  @article{ , abstract = {Wheel-legged robots integrate the adaptability of legged locomotion with the efficiency of wheeled movement, enabling agile traversal across diverse terrains. However, abrupt terrain transitions introduce substantial state variations, including velocity fluctuations, posture shifts, and slippage, which pose significant challenges to locomotion stability. To address these issues, we propose a state error compensation framework that integrates a residual network with a trust-region mechanism. The residual network implicitly captures nonlinear contact dynamics, enabling real-time correction of slippage-induced state deviations, while the trust-region mechanism regulates compensation amplitude to maintain stable locomotion. Furthermore, we introduce a dual-source contrastive learning strategy, which explicitly differentiates terrain-induced transitions from external perturbations, facilitating context-aware error recovery. The proposed framework is integrated into a model-free reinforcement learning pipeline, ensuring adaptability to previously unseen environments. To further enhance robustness, an uncertainty-aware calibration module is introduced. This module dynamically adjusts the trust region boundary in real time, leveraging sensory feedback to adaptively constrain residual corrections and prevent over-adjustment, thereby maintaining stability during diverse terrain transitions. Experimental results demonstrate that the proposed framework achieves a 96.7% terrain traversal success rate and 92% velocity tracking accuracy under dynamic disturbances. On unstructured and mixed terrains, it maintains a mean velocity tracking error of 0.15 m/s and stable posture, with pitch and roll angles constrained to ±0.04 rad and ±0.02 rad, respectively. Note to Practitioners—This work addresses the challenge of maintaining stable locomotion for wheel-legged robots when encountering abrupt terrain transitions. Conventional control methods often struggle with state variations such as slippage and posture shifts, leading to instability. To tackle this issue, we propose a state error compensation framework that integrates a residual network with a trust-region mechanism, ensuring real-time correction while maintaining control robustness. The framework is designed for seamless deployment in model-free reinforcement learning settings, requiring no prior terrain modeling. Our method is expected to facilitate real-world applications where reliable locomotion across diverse terrains is essential, such as autonomous exploration and disaster response.}, archive = {J_TASE}, author = {Naifeng He and Zhong Yang and Xiaoliang Fan and Wenqiang Que and Siyang Liu and Hongyu Xu and Chunguang Bu and Bi Zhang}, doi = {10.1109/TASE.2025.3620908}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23352--23365}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Residual policy optimization with trust region constraints: A learning framework for stable and agile wheel-legged locomotion}, volume = {22}, year = {2025}, }
• (2025). Adaptive video-conditioned imitation learning via bidirectional cross-domain skill transfer. TASE, 22, 23214--23227. (www)

  @article{ , abstract = {Imitation learning by watching humans offers a promising path to learning general-purpose robot skills with intuitive task specifications. While prior approaches for video-conditioned imitation learning directly extract skill embeddings from unstructured human videos and follow demonstrations step-by-step, such paradigm usually falls short in generalizing to unseen long-horizon tasks with a single human prompt video due to the significant embodiment and environment gap. To this end, our key insight is to infer local intentions from videos in order to retrieve robot skill memories from prior experience, and conversely select the feasible video clip to follow based on robot observations. Motivated by this, we introduce AdaMimic, a hierarchical imitation learning method that learns the bidirectional mapping of cross-domain sensorimotor skills and derives skill-based policy conditioned on adaptable latent plans. To enable generalization to unseen tasks given cross-domain human videos, AdaMimic leverages task-agnostic play data for interaction-aware skill embedding extraction and video-robot trajectory pairs for semantic and temporal human-to-robot skill alignment. In addition, our method exploits a skill adapter for robot-to-human alignment to adaptively align the robot with the skill intentions. We systematically evaluate AdaMimic on both simulated and real-world kitchen domains, demonstrating AdaMimic’s superiority over prior imitation learning methods in generalizing to novel long-horizon tasks with a single human prompt video. Note to Practitioners—In this work, we investigate the problem of imitating long-horizon manipulation tasks from raw human videos. Instead of collecting an unattainable amount of teleoperation data for learning unseen long-horizon tasks, we resort to easy-to-collect human videos in different environments to specify the intended robot goals, thus enabling new skill acquisition. Existing works on video-conditioned imitation learning are usually confined to relatively short-horizon tasks due to the high data supervision demands and similar layout constraints. We propose tackling these challenges via an adaptive skill-based imitation learning method AdaMimic, centered around taking human videos from a third-person perspective as input and producing robot actions conditioned on adaptable skill embeddings. To bridge cross-domain skills and capture the fine-grained knowledge encapsulated in human videos, we propose a bidirectional cross-domain skill transfer paradigm that maps video clips via transformer-based encoder to pre-trained robot skill embeddings for skill memory retrieval and inversely, aligns the robot with skill intentions in video clips based on current robot state for adaptively selecting most feasible video clips as the condition of low-level motion policies. Extensive experiments in both simulated and realistic kitchen environments show that our AdaMimic significantly outperforms other skill-based imitation learning baselines in generalizing one-shot to composite novel tasks and adapting to out-of-distribution tasks.}, archive = {J_TASE}, author = {Zhenyang Lin and Yurou Chen and Zhengwei Li and Xianxiang Zhang and Bin Liang and Zhiyong Liu}, doi = {10.1109/TASE.2025.3627627}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23214--23227}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive video-conditioned imitation learning via bidirectional cross-domain skill transfer}, volume = {22}, year = {2025}, }
• (2025). Design of a novel force-controlled end-effector with passive structural compliance and intrinsic contact sensing. TASE, 22, 23202--23213. (www)

  @article{ , abstract = {This paper presents a novel active force-controlled end-effector with passive structural compliance and intrinsic contact sensing capabilities. Slender elastic beams are introduced to provide the end-effector with structural compliance to accommodate inevitable fluctuations through deformation during operation. Meanwhile, flexible strain gauges are embedded into the slender elastic beams to endow the end-effector with intrinsic sensing capability. A kinetostatic model is established in closed-form to map the local deflections of strain gauges to the contact force with the environment. On this basis, model-based feedback control can be readily implemented to actively adjust the contact force between the end-effector and the workpiece. A prototype is developed, on which a variety of experiments are conducted to validate the effectiveness of the proposed end-effector. The results demonstrate that the end-effector can control the contact force accurately, with a maximum error of around $0.65N$ . Furthermore, a demonstration of robotic grinding using the developed prototype showcases its potential for practical applications. Note to Practitioners—This paper is motivated by the demand for intelligent force-controlled end-effectors in robotic manufacturing and interactive scenarios. In this work, a novel force-controlled end-effector for robotic grinding is presented. Slender elastic beams are utilized to output force and enable the end-effector with structural compliance. By integrating flexible strain gauges into the slender elastic beams, the end-effector achieves intrinsic contact sensing with the environment. Its structural compliance allows the end-effector to adapt to fluctuations in robot motion during the manufacturing process. The intrinsic force-sensing capabilities enable precise feedback control, critical for maintaining high manufacturing quality. The compact design of the end-effector improves its adaptability and facilitates easy installation in a variety of settings. Future research will focus on achieving more degrees of freedom in orientation/force hybrid control. Therefore, the presented end-effector has the potential for applications in robotic manufacturing, including tasks requiring compliant manipulation and precise force control, as well as human-robot collaboration environments.}, archive = {J_TASE}, author = {Tianyi Yan and Jianhuan Chen and Siyue Yao and Xuyang Duan and Yanjun Wang and Hao Wang and Genliang Chen}, doi = {10.1109/TASE.2025.3627234}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23202--23213}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design of a novel force-controlled end-effector with passive structural compliance and intrinsic contact sensing}, volume = {22}, year = {2025}, }
• (2025). Adaptive fault-tolerant control based on second-order sliding mode observer with nonlinear feedback for uncertain euler-lagrange systems. TASE, 22, 23190--23201. (www)

  @article{ , abstract = {This paper addresses the fault-tolerant trajectory tracking control problem of Euler-Lagrange (EL) systems subject to actuator faults and component faults (including system uncertainties). A novel finite-time Second-Order Sliding Mode Observer with Nonlinear feedback (SOSMON) incorporating adaptive fault estimators is proposed. Unlike linear feedback or complex sliding mode designs, the proposed SOSMON’s nonlinear feedback terms effectively handle the unknown high-order nonlinear terms in EL systems. To reconstruct the two faults, an adaptively weighted Radial Basis Function Neural Network (RBF NN) technique is introduced to estimate component faults without prior knowledge of fault characteristics, and an Adaptive actuator Fault Estimator (AFE) is designed. Then, accurately estimated faults allow for the simple design of a Sliding Mode Fault Tolerance Controller (SMFTC). Finite-time convergence of the observer and closed-loop stability are proved via Lyapunov theory. Finally, comparative simulations on an Autonomous Surface Vessel (ASV) demonstrate the superiority of the proposed method over conventional observers and the Back Propagation (BP) NN-based estimator regarding estimation accuracy and convergence speed. Note to Practitioners—Autonomous systems, such as marine vehicles and industrial manipulators, often suffer performance degradation or failures due to actuator faults (e.g., motor power loss), component damage (e.g., mechanical wear), and environmental disturbances. Traditional fault estimation solutions either rely on hard-to-obtain precise system models or require intensive computations, making real-time implementation challenging. This work presents a simplified control framework with low model dependency to rapidly detect and compensate for multiple simultaneous faults without prior knowledge of complex nonlinear dynamics. By deploying a sliding mode observer with two adaptive estimators that learn system anomalies directly from measurable position and velocity data, our approach estimates actuator failures and component faults within a guaranteed time while counteracting unknown disturbances. A straightforward control law automatically adjusts inputs to maintain trajectory tracking accuracy using the estimated faults. Simulation on autonomous marine vehicles demonstrates faster fault detection and higher tracking precision than conventional model-based observers. Future work will explore sensor fault tolerance and advanced observers, enhancing the reliability of automated systems.}, archive = {J_TASE}, author = {Zixuan Liang and Hao Wen and Baoheng Yao and Zhihua Mao and Lian Lian}, doi = {10.1109/TASE.2025.3626759}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23190--23201}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fault-tolerant control based on second-order sliding mode observer with nonlinear feedback for uncertain euler-lagrange systems}, volume = {22}, year = {2025}, }
• (2025). Predictive containment control for multi-agent systems subject to denial-of-service attacks. TASE, 22, 23177--23189. (www)

  @article{ , abstract = {This paper is concerned with the $H_{\infty } $ containment cooperative control problem for multi-agent systems (MASs) in a networked predictive control (NPC) framework, in which a recursive predictor is proposed to compensate for lost data caused by denial-of-service (DoS) attacks. Firstly, to facilitate system analysis, the DoS attacks are modeled by introducing a constraint on the attack duration. Secondly, to reproduce the lost states and fulfil containment task, we investigate a cooperative protocol based on prediction. This protocol uses states received from neighbor agents and the predictors. If the attacks are detected, the system control task is performed using predicted states; otherwise, true states are used. Thirdly, some sufficient conditions in terms of linear matrix inequalities are derived, which can guarantee that the error systems with proposed predicted-based protocol can achieve $H_{\infty } $ performance. Furthermore, the proposed predicted-based strategy is investigated with dynamic event-triggering mechanisms involving an auxiliary dynamic variable (ADV-DETMs) under switching topologies. A controller is designed to achieve the $H_{\infty } $ containment control for the MASs. Finally, we build an uncrewed surface vehicles (USVs) platform as an application of our methods numerically, which demonstrates the effectiveness of the proposed predict-based cooperative control strategy. Note to Practitioners—This paper is inspired from the problem of multiagent cooperation. Cooperative control strategies hold substantial practical value in a variety of engineering domains. For example, in a production line or collaboration task, multiple workstations are typically arranged in a series. These workstations exchange data with one another via an industrial BUS. If the data is corrupted during transmission, it will make an adverse effect on the control performance. In this paper, we consider cyber attacks as a factor affecting transmission. A prediction based approach is employed to mitigate the impact of the cyber attacks. With this approach, the predicted data is used to interact between workstations as the attacks occur and are detected. This paper presents the mathematical analysis about control systems with the proposed prediction mechanism. In particular, we demonstrate some criteria under which a multiagent system can be stabilized with certain performance constraint. Furthermore, some conclusions are made to design a satisfactory multiagent system. In future research, we aim to enhance our prediction mechanism to facilitate more practical problems and to achieve superior control performance.}, archive = {J_TASE}, author = {Chenxi Wang and Yanping Yang and Lei Zou and Bo Shen}, doi = {10.1109/TASE.2025.3626377}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23177--23189}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predictive containment control for multi-agent systems subject to denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Safe refined oil dispatching via constrained multiagent reinforcement learning with hierarchical action spaces. TASE, 22, 23164--23176. (www)

  @article{ , abstract = {The rise in urbanization and the demand for refined oil pose a significant challenge in improving transportation efficiency while ensuring the safety of gas station inventories. Traditional optimization methods, which rely on deterministic models or demand forecasting, enhance efficiency but struggle with uncertainties such as demand fluctuations. Multi-agent reinforcement learning (MARL) presents a promising approach for adaptive cooperative dispatching. This study models the refined oil dispatch problem as a partially observable constrained Markov game (CPOMG), in which agents make cooperative decisions under inventory constraints with limited observation, and proposes Hierarchical Action-Constrained MAPPO (HAC-MAPPO), a novel cooperation MARL algorithm to find the optimal cooperative dispatching policy of the game. Within the CPOMG framework, the HAC-MAPPO agents optimize order generation decisions based on the states of gas station inventories. The subsequent delivery routing for these orders is optimized using an integrated fuel routing tabu search algorithm, which is part of the environment’s state transition logic. HAC-MAPPO incorporates inventory safety thresholds to constrain order generation decisions across fuel types. It employs a dual-critic architecture and a shared hierarchical actor network for efficient decentralized policy learning. Crucially, the Lagrangian relaxation technique is applied to transform the constrained actor optimization objective (for order generation) into an unconstrained form. Evaluations carried out in real-world scenarios (based on actual geographic coordinates) and synthetic scenarios using multiple types of synthetic fuel consumption datasets demonstrate that the proposed model and algorithm significantly reduce inventory violations and improve dispatch efficiency and safety under varying demand patterns, outperforming baseline methods. Note to Practitioners—The efficient and stable dispatch of refined oil products is a critical challenge in the petroleum supply chain, directly impacting operational costs and service reliability. Traditional methods often rely on static demand forecasts, which struggle to adapt to dynamic market fluctuations, leading to inefficiencies such as delayed deliveries, stockouts, or overstocking. This paper presents a practical solution leveraging multi-agent reinforcement learning (MARL) to optimize dispatching decisions in real time, ensuring both efficiency and safety. Our approach, HAC-MAPPO, enables gas stations, depots, and transport vehicles to collaboratively adapt to changing demand patterns while adhering to inventory safety constraints. By integrating hierarchical action selection decision-making and adaptive routing algorithms, the system minimizes delivery delays and reduces inventory risks, such as shortages or excess stock. This method has been validated in both synthetic and real-world scenarios, demonstrating significant improvements in dispatch efficiency and operational stability. For industry practitioners, this work offers a scalable framework that can be integrated into existing logistics systems with minimal disruption. Potential applications include large-scale fuel distribution networks, where dynamic demand and complex routing are common challenges. Although the current implementation focuses on refined oil products, the methodology can be extended to other supply chain domains with similar requirements, such as chemical or liquid goods transportation. Future enhancements could involve incorporating additional real-world constraints, such as vehicle maintenance schedules or traffic conditions, to further improve practicality.}, archive = {J_TASE}, author = {Kun Tang and Chengwei Zhang and Wanting Liu and Xue Li and Qi Wang and Dou An and Furui Zhan}, doi = {10.1109/TASE.2025.3625392}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23164--23176}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe refined oil dispatching via constrained multiagent reinforcement learning with hierarchical action spaces}, volume = {22}, year = {2025}, }
• (2025). Multi-UAV cooperative decision-making under noisy and uncertain environments. TASE, 22, 23152--23163. (www)

  @article{ , abstract = {In the process of multi-UAV cooperative decision-making, the information evolution model plays a key role in the communication and fusion of information among multiple agents. However, traditional information fusion models, such as the DeGroot model and the Hegselmann-Krause bounded confidence (HK) model in opinion dynamics, have certain limitations in handling complex noisy environments, adjusting weights dynamically, and processing linguistic information. To address these challenges, this paper proposes an information fusion model that integrates the social network mechanism, the hyperbolic tangent function, and the Monte Carlo method. In each round of information iteration, the model uses the social network structure to simulate the communication relationships among agents. Specifically, the social network is represented as a binary adjacency matrix. Then, a distance measure between linguistic information is defined. Based on this distance, the hyperbolic tangent function is used to determine the weights of agents during the evolution process dynamically. To fully evaluate the model’s performance under noise, Monte Carlo simulations are used for statistical analysis. A case of multi-UAV cooperative reconnaissance is presented to validate the effectiveness of the model. In addition, detailed comparisons with several existing models are conducted. The experimental results show that the proposed model demonstrates clear advantages in terms of anti-interference capability, stability, and adaptability to various linguistic information environments. Note to Practitioners—This paper presents an information fusion model in multi-UAV cooperative decision-making. Traditional opinion dynamics models often struggle with noisy environments, fixed weighting, and complex linguistic information. To address these issues, the proposed model integrates social network structures, a hyperbolic tangent-based weighting mechanism, and Monte Carlo simulations for robustness evaluation. It supports dynamic information adjustment and is well-suited for uncertain scenarios. A case study on UAV cooperative reconnaissance demonstrates the model’s practical value in enhancing stability, adaptability, and resistance to interference in real-world multi-agent systems.}, archive = {J_TASE}, author = {Qianlei Jia and Francisco Javier Cabrerizo and Ignacio Javier Pérez and Enrique Herrera-Viedma}, doi = {10.1109/TASE.2025.3625085}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23152--23163}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-UAV cooperative decision-making under noisy and uncertain environments}, volume = {22}, year = {2025}, }
• (2025). Distributed data-driven control for adjustable current sharing and secure voltage restoration in DC microgrids. TASE, 22, 23141--23151. (www)

  @article{ , abstract = {For DC microgrids (MGs), real-time adjustment of current sharing ratios and secure voltage restoration are paramount for optimizing load allocation and enhancing dynamic performance. In this paper, a dual-objective distributed model-free adaptive control (MFAC) scheme is designed for the first time to guarantee voltage transient performance and adjustable current sharing. First, an output-constrained nonlinear MG model with ZIP (constant impedance, constant current and constant power) load is established, and subsequently it is converted into an equivalent unconstrained data model using system transformation and dynamic linearization techniques. Second, a new prescribed performance control algorithm with asymmetrical preset boundaries is proposed to restrict voltage transient responses. This algorithm is updated with real-time input and output data at discrete instants, making it independent of line resistance and ZIP load measurements. To enhance the robustness of the control method, an internal observer is designed to actively compensate for the unknown nonlinear dynamics generated by time-varying system parameters. The stability conditions of the transformed systems in the presence of ZIP loads and time-varying line resistance are derived, which can indirectly ensure the prescribed voltage performance of the original system. Finally, the effectiveness of the proposed control method is validated through some simulations and hardware experiments. Note to Practitioners—Current sharing and voltage restoration are two main control tasks in the DC microgrid. With the increasing penetration of renewable energy, real-time adjustment of the current sharing ratio according to the capacity of distributed generators (DGs) can prevent overload. When the system suffering large disturbances, perfect voltage transient response can protect sensitive loads and avoid malfunctions of protection equipment. However, the existing secondary control strategies cannot deal with these problems in a unified distributed MFAC framework. Therefore, this paper propose a new distributed MFAPPC method to achieve adjustable current sharing and secure voltage restoration for the DC MG. This data-driven approach can ensure the stability of the MG system in the presence of ZIP loads and time-varying line resistance, even if the model and load information are unavailable.}, archive = {J_TASE}, author = {Xiaojie Qiu and Bo Fan and Wenchao Meng and Yingchun Wang and Yan Xu and Zhaoyang Dong}, doi = {10.1109/TASE.2025.3624459}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23141--23151}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed data-driven control for adjustable current sharing and secure voltage restoration in DC microgrids}, volume = {22}, year = {2025}, }
• (2025). Real-time 3D motion prediction for human-robot collaboration via bayesian-optimized diffusion models. TASE, 22, 23128--23140. (www)

  @article{ , abstract = {Human motion prediction is a cornerstone of human-robot collaboration (HRC), as robots need to infer future movements of human workers based on observed motion cues to proactively plan their actions, ensuring safety in close collaboration scenarios. The diffusion model has demonstrated remarkable performance in predicting high-quality human motion with reasonable diversity, but suffers from a slow generative process that requires multiple times of model inference, hindering real-world applications. To enable real-time prediction, we propose training a one-step multi-layer perceptron-based (MLP-based) diffusion model using knowledge distillation and Bayesian optimization. Our method consists of two steps. First, we distill a pretrained diffusion-based motion predictor, TransFusion, directly into a one-step diffusion model with the same denoiser architecture. Then, to further reduce the inference time, we remove the computationally expensive components from the original denoiser and use knowledge distillation once again to distill the obtained one-step diffusion model into an even smaller model based solely on MLPs. Bayesian optimization is used to tune the hyperparameters for training the smaller diffusion model. To demonstrate the effectiveness of our model, we design a human-robot collaborative desktop disassembly task. The results showcase our model’s capability to forecast multiple realistic human motion in real time, addressing the uncertainty and multi-modality of human motion. Additional experimental studies are conducted on benchmark datasets to ensure fair comparisons with existing works, highlighting that our model significantly improves inference speed, achieving real-time prediction without noticeable degradation in performance. The project page is available at https://github.com/sibotian96/SwiftDiff Note to Practitioners—This paper was motivated by the problem of ensuring safety in human-robot collaborative disassembly, considering human motion uncertainty and multi-modality, but it also applies to other safety-critical scenarios involving humans, where human motion is naturally ambiguous and predicting multiple possible future motion instead of a single deterministic forecast is essential. State-of-the-art works build on diffusion models for better prediction quality, and utilize complex neural networks to generate diverse future motion, ignoring computational expense and inference speed, which are critical factors in determining whether the algorithms can be applied to real-world applications. For the first time, our method proposes a one-step diffusion model that uses a simple neural network as denoiser. We test and compare our method with state-of-the-art works on two benchmark datasets. The results show that our method speeds up the inference time by 98% compared to the fastest existing diffusion-based method, without compromising prediction quality. We also design a human-robot collaborative desktop disassembly experiment to emulate real-world e-waste recycling tasks. Preliminary results show that our model can generate multiple predictions with reasonable accuracy and diversity in just 0.01 seconds, making it suitable for real-time prediction in safety-critical tasks. In future research, we will design more complex and varied HRC scenarios and further test the proposed method.}, archive = {J_TASE}, author = {Sibo Tian and Minghui Zheng and Xiao Liang}, doi = {10.1109/TASE.2025.3624396}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23128--23140}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time 3D motion prediction for human-robot collaboration via bayesian-optimized diffusion models}, volume = {22}, year = {2025}, }
• (2025). Memory-driven dynamic output feedback control for networked active suspension system with time-varying input delay. TASE, 22, 23116--23127. (www)

  @article{ , abstract = {This study addresses the output feedback control synthesis problem for a quarter-vehicle active suspension system subject to input delay, using dynamic integral quadratic constraint. To account for the uncertain relationship between the real-time output of the controller and the delayed signal arriving at the actuator, a memory-driven control strategy is proposed. Furthermore, dynamic integral quadratic constraint multipliers are employed to reformulate the output feedback control synthesis problems into a set of linear matrix inequalities. A simplified memory-free dynamic output feedback controller is also introduced. Corresponding controller design algorithms are developed for both the memory-driven and memory-free schemes. Simulation results demonstrate the effectiveness of the proposed controllers in handling input delays, and comparative studies verify the superiority and feasibility of the proposed control approach. Note to Practitioners—Time-varying input delays in vehicle active suspension systems caused by communication delays or actuator limitations can degrade control performance and even destabilize the system. However, conventional controllers struggle to balance delay compensation with design complexity. To address this matter, this study proposes a memory-driven dynamic feedback control method that addresses delay-induced uncertainties between controller outputs and delayed actuator inputs. Specifically, the controller design is streamlined into a computationally efficient solving algorithm, achieving enhanced engineering practicality by avoiding intricate theoretical frameworks. Simulations demonstrate enhanced driving comfort and stability under severe delays, making it suitable for self-driving vehicles, industrial machinery, and other delay-sensitive applications. Ultimately, experimental results under severe delay conditions demonstrate marked improvements in ride comfort (reduced vibration transmission) and stability (enhanced road-holding capability), making it suitable for self-driving vehicles, industrial machinery, and other delay-sensitive applications.}, archive = {J_TASE}, author = {Hanzhe Shi and Guoliang Chen and Te Yang and Xingyue Liang and Jianwei Xia}, doi = {10.1109/TASE.2025.3624363}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23116--23127}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Memory-driven dynamic output feedback control for networked active suspension system with time-varying input delay}, volume = {22}, year = {2025}, }
• (2025). Spatial temporal attention-based target vehicle trajectory prediction for internet of vehicles. TASE, 22, 23103--23115. (www)

  @article{ , abstract = {Forecasting vehicle behavior within complex traffic environments is pivotal within Intelligent Transportation Systems (ITS). Though this technology plays a significant role in alleviating the prevalent operational difficulties in logistics and transportation systems, the precise prediction of vehicle trajectories still poses a substantial challenge. To address this, our study introduces the Spatio Temporal Attention-based methodology for Target Vehicle Trajectory Prediction (STATVTPred). This approach integrates Global Positioning System(GPS) localization technology to track target movement and dynamically predict the vehicle’s future path using comprehensive spatio-temporal trajectory data. We map the vehicle trajectory onto a directed graph, after which spatial attributes are extracted via a Graph Attention Networks (GATs). The Transformer technology is employed to yield temporal features from the sequence. These elements are then amalgamated with local road network structure maps to filter and deliver a smooth trajectory sequence, resulting in precise vehicle trajectory prediction. This study validates our proposed STATVTPred method on T-Drive and Chengdu taxi-trajectory datasets. STATVTPred achieves an AMR of 73.07% on the Beijing dataset, surpassing the Transformer by 6.38% and the LSTM Encoder-Decoder by 37.45%, while also reducing Distance Error (DE) by 26.93% and 20.95% in Beijing and Chengdu, respectively, also much lower than the baseline results. This is expected to establish STATVTPred as a new approach for handling trajectory prediction of targets in logistics and transportation scenarios, thereby enhancing prediction accuracy. Note to Practitioners—This article is motivated by the need for a high-precision trajectory prediction method for unmanned aerial vehicles (UAVs) in complex traffic scenarios. Especially in scenarios such as urban canyons or when tracking a target vehicle is lost, predicting the movement trajectory of the UAV with high precision and speed generates significant interest. In practice, as the trajectory of a target movement isn’t always known, and solely relying on satellite positioning is limited by the application scenario. It presents a challenge to determine the movement trajectory of the lost UAV in a short period of time. In response to these issues, we integrated GPS positioning technology, extracted spatial attributes through a graph attention network, and used Transformer technology to extract time features from a sequence. By combining these two kinds of attributes, we obtained comprehensive spatiotemporal trajectory data, thereby dynamically predicting the future path of the UAV. The target trajectory prediction technology developed in this paper can be expanded to multiple UAV trajectory prediction scenarios to achieve coordinated path optimization, thereby improving overall performance. This paper presents an innovative approach that would benefit practitioners in the field by providing a solution to the challenging problem of high precision trajectory prediction in complex traffic scenarios.}, archive = {J_TASE}, author = {Ouhan Huang and Huanle Rao and Xiaowen Cai and Tianyun Wang and Aolong Sun and Sizhe Xing and Yifan Sun and Gangyong Jia}, doi = {10.1109/TASE.2025.3624301}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23103--23115}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Spatial temporal attention-based target vehicle trajectory prediction for internet of vehicles}, volume = {22}, year = {2025}, }
• (2025). Stability analysis of networked stochastic coupled systems with impulsive control under round-robin protocol and deception attacks. TASE, 22, 23095--23102. (www)

  @article{ , abstract = {This paper addresses the stability problem of a class of networked stochastic coupled systems under impulsive control, Round-Robin protocol (RRP), and deception attacks. A novel system model is established that integrates stochastic dynamics, communication protocol, and deception attacks, while explicitly capturing the influence of the underlying network topology. Unlike existing studies on impulsive systems under RRP that predominantly consider impulses as undesirable disturbances, this work exploits impulsive control as an active stabilization tool, where the RRP are explicitly incorporated into the impulse timing design. By constructing topology-dependent Lyapunov functions and employing graph theory, new sufficient conditions are derived to ensure system stability. Finally, the theoretical framework is validated through case studies on a single-link robotic arm and a stochastic oscillator, demonstrating improved stability and performance under impulsive control and Round-Robin protocol. Note to Practitioners—Networked control systems (NCS) are widely used in industrial automation, robotics, and smart infrastructure due to their cost efficiency and flexibility. However, practical challenges such as limited communication bandwidth and cyber threats (e.g., deception attacks) can destabilize these systems. This paper proposes a novel impulsive control strategy combined with the Round-Robin protocol (RRP) to enhance the stability of stochastic coupled systems under adversarial conditions. Unlike traditional methods that treat impulses as disturbances, our approach actively leverages impulsive control to counteract attack-induced disruptions while efficiently managing network resources via RRP. The theoretical framework is validated through real-world applications, including robotic arms and oscillators. For engineers, this work offers actionable insights into designing resilient NCS under communication constraints and cyber threats, with potential applications in smart grid and multi-agent systems. Future directions include extending the method to large-scale networks and hardware-in-the-loop testing.}, archive = {J_TASE}, author = {Wenhua Wang and Qishen Chen and Wenxue Li}, doi = {10.1109/TASE.2025.3624252}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23095--23102}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stability analysis of networked stochastic coupled systems with impulsive control under round-robin protocol and deception attacks}, volume = {22}, year = {2025}, }
• (2025). Data-driven self-triggered MPC and stability analysis of linear time-invariant systems under noise disturbance. TASE, 22, 23083--23094. (www)

  @article{ , abstract = {This paper proposes a data-driven self-triggered model predictive control (MPC) scheme for constrained linear time-invariant (LTI) systems that are unknown and affected by bounded process and measurement noise. The proposed scheme relies solely on initial input-output data and accounts for noise in both offline and online measurements. A self-triggered control policy is developed based on optimal input-output trajectories obtained by solving the data-driven MPC optimization problem. Recursive feasibility of the control scheme is guaranteed and practical exponential stability of the closed-loop system is established under sufficiently small noise level and certain technical conditions. Two simulation examples are provided to validate the effectiveness of the theoretical results. Note to Practitioners—Model predictive control is a widely used technique in control engineering. However, its implementation typically relies on accurate system modeling and requires substantial computational resources. In real-world scenarios, the presence of process and measurement noise also makes precise modeling challenging. On the other hand, data-driven control has emerged as a promising alternative by designing controllers directly from measured input-output data, eliminating the need for explicit system identification. Therefore, this paper proposes a data-driven self-triggered MPC strategy that not only designs the controller directly from measured data but also determines the control update times based on system behavior rather than fixed sampling intervals. The proposed method explicitly accounts for both process and measurement noise and reduces the computational burden of the controller, while ensuring robustness and stability of the system. Finally, the effectiveness and practicality of the proposed method are validated through two simulation examples.}, archive = {J_TASE}, author = {Xinyun Yu and Feiqi Deng and Xueyan Zhao}, doi = {10.1109/TASE.2025.3623100}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23083--23094}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven self-triggered MPC and stability analysis of linear time-invariant systems under noise disturbance}, volume = {22}, year = {2025}, }
• (2025). Hierarchical control with steering mode switching for MDED-HDV via maneuver stability region analysis. TASE, 22, 23066--23082. (www)

  @article{ , abstract = {Modular distributed electric drive heavy-duty vehicles (MDED-HDV) integrate advanced technologies such as all-wheel steering (AWS) and distributed drive, achieving complete decoupling of the chassis’ motion degrees of freedom (DoFs). This architecture is considered a promising solution for enhancing the stability of heavy-duty vehicles (HDV). However, the impact mechanism of multi-axle steering configurations on stability remains inadequately understood, and the redundancy in control DoFs results in multiple feasible steering configurations. To address these challenges, this paper proposes a hierarchical control framework featuring steering mode switching based on stability region constraints. First, a dynamics model of MDED-HDV is established using rational polynomials. Subsequently, the sum-of-squares programming (SOSP) is employed to estimate the stability region, providing the first analysis of the effects of multi-axle steering on the stability region from the perspective of nonlinear system dynamics. Based on this analysis, a stability region-based steering mode switching strategy is developed. It incorporates vehicle states and road conditions to enable autonomous transitions among anti-phase, front-wheel, and in-phase steering modes. Finally, a hierarchical control framework is implemented. The upper layer selects the steering mode based on the estimated stability region. The lower layer executes a trajectory tracking controller with stability region constraints. The framework addresses the issue of multiple solutions caused by redundant DoFs. Experimental results demonstrate that the proposed steering mode switching strategy improves the tracking accuracy, while the stability region-based controller ensures maneuver stability. Note to Practitioners—MDED-HDVs achieve full decoupling of chassis motion DoFs through independently configured in-wheel and steering motors. This design fully unleashes the chassis’ motion potential and establishes a robust foundation for high mobility and stability control. However, the impact of multi-axle steering on the stability boundaries of HDVs has not been sufficiently explored from a theoretical perspective. Meanwhile, the multi-solution problem caused by actuator redundancy continues to lack effective control strategies. These challenges impede the intelligent control and practical application of MDED-HDVs. To address these issues, this study introduces a novel application of SOSP methods to MDED-HDVs. It provides the first systematic analysis of how AWS and multi-axle configurations affect the stability region. Based on this analysis, a stability region-constrained steering mode switching strategy is designed. This strategy enables intelligent and adaptive transitions among anti-phase, front-wheel, and in-phase steering modes. Furthermore, a hierarchical control architecture is constructed. The upper layer selects steering modes based on stability region estimation. The lower layer performs trajectory tracking under the corresponding stability constraints. This framework effectively resolves convergence issues caused by redundant control DoFs, ensuring both maneuverability and stability under complex operating conditions. This research offers a new theoretical foundation and engineering solution for the path tracking and stability control of MDED-HDVs. It fills a critical gap in the coordinated control of multi-axle heavy-duty vehicles with all-wheel steering. Moreover, it demonstrates promising application prospects in fields such as ultra-heavy logistics transportation, emergency rescue, and specialized engineering equipment.}, archive = {J_TASE}, author = {Ruiqi Fang and Tong Shen and Xin Bai and Jinhao Liang and Fanxun Wang and Weichao Zhuang and Guodong Yin}, doi = {10.1109/TASE.2025.3613526}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23066--23082}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical control with steering mode switching for MDED-HDV via maneuver stability region analysis}, volume = {22}, year = {2025}, }
• (2025). Multiobjective semi-supervised ensemble learning and its application in iron and steel industry. TASE, 22, 23051--23065. (www)

  @article{ , abstract = {High-quality predictions of key indicators are essential to maintain stable production in the iron and steel industry. However, most existing prediction methods rely on manually designed supervised learning models. These methods do not consider the effective utilization of unlabeled data, which is prevalent in practical production environments, to improve the model performance. As a result, the accuracy and generalization of these methods tend to perform poorly in real-world applications. To tackle this issue, this paper develops a multiobjective semi-supervised ensemble learning method with evolutionary neural architecture search (MOSSEL-ENAS), in which a one-dimensional convolutional neural network (OD-CNN) is employed as the base learner. MOSSEL-ENAS takes the input feature permutation and neural architecture of each OD-CNN as the decision variables, and an improved non-dominated sorting genetic algorithm is developed to simultaneously optimize two objectives, i.e., accuracy and complexity of each base learner. Moreover, an adaptive evaluation pattern adjustment strategy is proposed to adaptively adjust the learning paradigm (i.e., semi-supervised learning and supervised learning) of OD-CNNs by incorporating unlabeled data into the training process, further improving the learning performance of the algorithm. After evolution, a set of elite base learners with high-accuracy and low-complexity are selected and combined using a LightGBM-based stacking ensemble method to construct the final ensemble model. Experimental results on two key indicator prediction tasks in the iron and steel industry demonstrate that MOSSEL-ENAS achieves competitive or even better accuracy compared to other powerful learning methods, and also outperforms the existing task-specific learning methods. Furthermore, the competitiveness of a modified version of MOSSEL-ENAS on a classification task highlights the scalability of our proposal. Note to Practitioners— Many key indicators in the production process of the iron and steel industry cannot be tested online in a real-time manner, which poses challenges for monitoring, control, and operation optimization. The proposed multi-objective semi-supervised ensemble learning method with evolutionary network architecture search allows practitioners to leverage production process data to construct prediction models for key indicators in different production scenarios, thus achieving better process monitoring and quality optimization.}, archive = {J_TASE}, author = {Yao Wang and Xianpeng Wang and Yu Xue}, doi = {10.1109/TASE.2025.3625751}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23051--23065}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multiobjective semi-supervised ensemble learning and its application in iron and steel industry}, volume = {22}, year = {2025}, }
• (2025). Dexterity-guided dimensional synthesis and multi-task control for fingertip manipulation. TASE, 22, 23039--23050. (www)

  @article{ , abstract = {The geometric parameters severely impact the performance of dexterous manipulation, but manually adjusting them is time-consuming. In this paper, we employ the dexterity-guided dimensional synthesis to explore the geometric design space of a robotic hand, and propose a multi-objective optimization framework to improve its manipulation dexterity. Specifically, we first establish a screw-based mathematical model to describe the hand-object system’s kinetostatic properties. Three objectives are specified to achieve Pareto optimality: maximizing the system’s feasible position space, feasible orientation space, and manipulation stability. Furthermore, to validate the optimal design parameters in practice, we develop a multi-task object motion controller and apply it to a letter handwriting task. Finally, the optimized results are quantitatively analyzed by simulation, yielding a Pareto front and 112 optimal solutions. The applicability of the optimal parameters is assessed through a letter “O” handwriting experiment. Results show that the optimized hand can write the “O” with a maximum radius of 30.0 mm, which is 27.12% larger than that of the non-optimized hand. This controller is also used to schedule subtasks with varying priorities to avoid constrained regions. In contrast to the neural network-based PID controller, the designed controller prioritizes the task’s critical parts, ensuring the legibility of the letters. Note to Practitioners—The main motivation of this paper is to propose a multi-objective optimization framework for manipulation dexterity, aimed at enhancing the robotic hand’s ability to manipulate objects and validating its performance through simulation. Currently, most optimization methods primarily focus on a single objective or a single finger, often ignoring the coordinated movements between fingers. This paper begins by establishing a screw-based model to describe the hand-object system’s kinetostatic properties. It then proposes a multi-objective, multi-parameter optimization framework for manipulation dexterity, allowing for a quantitative assessment of how geometric parameters affect object manipulation. Results show that the optimized hand (hand-I) outperforms the non-optimized hand (hand-II) across all three objectives. Another motivation of this paper is to design a multi-task object motion controller to examine the physical applicability of the optimal design parameters through the letter handwriting tasks. Results show that hand-I outperforms hand-II in writing the letter “O”. This controller also addresses the problem of partially constrained object trajectories. The comparative writing experiment of the letter “S” highlights the necessity of this controller and its ability to coordinate multiple subtasks effectively. We hope that the optimization framework for manipulation dexterity, the hand-object system modeling method, and the multi-task object motion controller applied to letter handwriting can provide new insights for other practitioners.}, archive = {J_TASE}, author = {Congjia Su and Rui Wang and Shaowei Cui and Shuo Wang}, doi = {10.1109/TASE.2025.3621449}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23039--23050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dexterity-guided dimensional synthesis and multi-task control for fingertip manipulation}, volume = {22}, year = {2025}, }
• (2025). Distributed model predictive control for real-time automatic train regulation of metro networks with transfer connections. TASE, 22, 23023--23038. (www)

  @article{ , abstract = {To improve the operation and service efficiency of automatic train supervision systems in metro networks, this paper investigates the distributed train regulation control problem for metro networks in the presence of disturbances. Train-to-Train (T2T) communication technology, a key component of next-generation Communication-Based Train Control (CBTC) systems, has fundamentally shifted industry and academic focus in the field toward decentralized, adaptive, and infrastructure-light solutions. Leveraging T2T communication, we propose a distributed control system architecture to alleviate the computational burden of central traffic controllers. More precisely, each train is equipped with an onboard controller that makes decisions based on the state of the network, and a novel real-time train regulation strategy is developed, taking into account the metro’s operational and service constraints. The proposed approach aims to minimize traffic delays and optimally coordinate connections between trains of different lines. The goal is also to reduce the waiting time deviation for transferring passengers. Unlike existing studies that primarily focus on centralized approaches and single open lines, we model both one-direction loop lines and two-track open lines as loop lines and then propose a Distributed Model Predictive Control (DMPC) approach to address the regulation challenges of loop line configurations. Numerical simulations are performed on the Beijing metro network using real data, and the results are discussed to demonstrate the effectiveness and efficiency of the proposed automatic train regulating method. Note to Practitioners—The core concept of decentralized, train-to-train (or vehicle-centric) communication is actively being deployed under frameworks such as Communication-Based Train Control (CBTC) and Train Autonomous Control Systems (TACS). These systems rely on direct inter-train communication to optimize operations and enhance scalability. The automatic train supervision subsystem plays an important role in issuing automatic train operation commands and monitoring train operation in metro signaling systems. To ensure timely arrival and departure in small interval operations, the Automatic Train Regulation (ATR) technology is employed to achieve the precise control of trains. However, existing ATR technologies struggle to manage the growing complexity of metro networks that mix open and loop lines. Their limitations lead to inefficient line coordination and ultimately increase passengers’ waiting time at transfer stations. This paper addresses this gap by developing a practical ATR strategy for such complex networks. Specifically, we propose a novel Distributed Model Predictive Control method that formulates the dynamic interdependencies between different line types, ensures coordination and minimizes transfer waiting times, thereby improving both operational efficiency and service quality. Thanks to its computational efficiency, the proposed method is applicable to real-time regulating of large-scale metro networks. Our work aligns with the emerging needs of decentralized, adaptive, and infrastructure-light solutions for the CBTC and TACS systems, offering a framework that bridges current technological trends with future operational demands.}, archive = {J_TASE}, author = {Yin Tong and Graziana Cavone and Jiate Luo and Carla Seatzu and Mariagrazia Dotoli}, doi = {10.1109/TASE.2025.3616600}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23023--23038}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed model predictive control for real-time automatic train regulation of metro networks with transfer connections}, volume = {22}, year = {2025}, }
• (2025). SFD-net: Enhancing small fruit detection before thinning period using query vectors. TASE, 22, 23007--23022. (www)

  @article{ , abstract = {Fruit detection in an unstructured orchard is a foundational visual task employed in plant phenotyping. This is a challenging task due to the minimal contrast between the foreground target and the background color. This situation seriously affects the accuracy of the target fruit detection model, especially for small pear fruits which are more difficult to recognize. Addressing this challenge, an end-to-end small object fruit detection model based on query vectors, SFD-Net, is proposed to achieve more accurate detection of young pear fruits and thus provide guidance for fruit thinning operations in pear orchards for farmers. Specifically, SFD-Net employs the lightweight MobileNetv2 network as a feature extraction framework to obtain the distinctive features of green fruits, then these features are subsequently fed directly into the decoder, where they are integrated with embedded query vectors for fruit identification and localization prediction. This approach efficiently detects small-scale fruits, benefiting from the global sensing capability of query vectors. Importantly, the model avoids the inclusion of redundant multi-scale feature pyramid networks and feature encoders, achieving an AP value of 51.9% and an AP ${}_{\mathbf {S}}$ value of 25.1% on the pear dataset, which escalates to an AP ${}_{\mathbf {S}}$ value of 27.9% after undergoing additional training iterations. Experimental results unequivocally demonstrate that SFD-Net outperforms other state-of-the-art methods, offering a significant advantage, particularly when considering its suitability for deployment on agricultural smart devices. Note to Practitioners—This paper introduces SFD-Net, a novel approach for detecting small fruits in complex, unstructured orchard environments, focusing particularly on green pears. The model leverages lightweight MobileNetv2 for feature extraction and utilizes query vectors to improve the detection of small, hard-to-detect fruits. The system is designed to overcome common challenges in agricultural environments, such as poor contrast between fruits and foliage, occlusions, and varying lighting conditions. For practitioners in smart agriculture, SFD-Net offers a promising solution that balances accuracy with computational efficiency, making it suitable for deployment on mobile and embedded devices. This could be particularly useful in scenarios where real-time performance is crucial, such as precision fruit thinning, automated harvesting, and yield estimation.}, archive = {J_TASE}, author = {Weikuan Jia and Ruina Zhao and Ru Jiang and Guoliang Liu}, doi = {10.1109/TASE.2025.3624608}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {23007--23022}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SFD-net: Enhancing small fruit detection before thinning period using query vectors}, volume = {22}, year = {2025}, }
• (2025). Single-frame point-pixel registration via supervised cross-modal feature matching. TASE, 22, 22995--23006. (www)

  @article{ , abstract = {Point-pixel registration between LiDAR point clouds and camera images is a fundamental yet challenging task in autonomous driving and robotic perception. A key difficulty lies in the modality gap between unstructured point clouds and structured images, especially under sparse single-frame LiDAR settings. Existing methods typically extract features separately from point clouds and images, then rely on hand-crafted or learned matching strategies. This separate encoding fails to bridge the modality gap effectively, and more critically, these methods struggle with the sparsity and noise of single-frame LiDAR, often requiring point cloud accumulation or additional priors to improve reliability. Inspired by recent progress in detector-free matching paradigms, we revisit the projection-based approach and introduce the detector-free framework for direct point-pixel matching between LiDAR and camera views. To further enhance matching reliability, we introduce a repeatability scoring mechanism that acts as a soft visibility prior. This guides the network to suppress unreliable matches in regions with low intensity variation, improving robustness under sparse input. Extensive experiments on KITTI, nuScenes, and MIAS-LCEC-TF70 benchmarks demonstrate that our method achieves state-of-the-art performance, outperforming prior approaches on nuScenes (even those relying on accumulated point clouds), despite using only single-frame LiDAR. Note to Practitioners—In many practical robotic and autonomous driving systems, only sparse LiDAR sensors (e.g., 16- or 32-line) are available due to cost and deployment constraints. This sparsity makes online LiDAR-camera calibration particularly challenging, since traditional methods often rely on dense multi-frame point clouds or additional calibration targets. Our proposed detector-free framework directly establishes point-pixel correspondences from a single sparse LiDAR scan, ensuring robust calibration even under noisy and low-density conditions. This enables practical online re-calibration after sensor maintenance or accidental impacts without requiring specialized setups. Furthermore, the approach allows camera localization against previously recorded sparse point cloud maps, making it valuable for infrastructure perception, historical-map-based navigation, and deployment on low-cost robotic platforms.}, archive = {J_TASE}, author = {Yu Han and Zhiwei Huang and Yanting Zhang and Fangjun Ding and Shen Cai and Xiaoyu Tang and Yanchao Dong and Rui Fan}, doi = {10.1109/TASE.2025.3621651}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22995--23006}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Single-frame point-pixel registration via supervised cross-modal feature matching}, volume = {22}, year = {2025}, }
• (2025). Safe edge intelligent control with cross-loop constraints and stability guarantees. TASE, 22, 22979--22994. (www)

  @article{ , abstract = {As modern control systems, which integrate cutting-edge technologies, such as edge computing and data analytics, are transforming industries, industrial control systems have evolved into multitier cyber-physical systems. Most current work targets single control loop or multiple independent control loops. However, in practice, multiple control loops may have close relations and dependencies among each other. It is nontrivial to optimize control performances of multiple control loops with interleaving constraints under tight timing restrictions of real-time control, while guaranteeing stability. In this work, we propose a cloud-edge cooperation framework tailored for intelligent multi-loop edge control. We design an intelligent multi-loop edge controller (IMEC) to efficiently generate the control policy for multi-loop control systems, leveraging both analytical and data-driven approaches. And we apply the transfer learning (TL) method to improve the robustness and efficiency under varying control scenarios. Furthermore, we present an edge-end cooperative Simplex architecture (EESA) to guarantee the stability of the system. Extensive evaluation over two case studies, linear and nonlinear multi-loop control systems, on semi-physical simulation platforms illustrate that IMEC significantly reduces the computation cost and improves control performance compared with analytical optimal control. Transfer learning effectively enhances the robustness under different control scenarios. And EESA can guarantee the safety during the running time. Note to Practitioners—Modern control systems are embracing multi-tier frameworks integrating cloud platforms, edge servers, and local devices Efficiently managing multi-loop control systems with interdependent constraints and varying parameters is a persistent challenge in industrial automation, especially in scenarios such as multi-agent coordination or tightly integrated manufacturing processes. Traditional analytical control algorithms often struggle to achieve the balance between computational efficiency and real-time control performance. This work proposes a cloud-edge-end framework to address these challenges. The intelligent multi-loop edge controller (IMEC) integrates analytical and data-driven approaches to reduce computational overhead while maintaining robust control performance. A transfer learning mechanism further enhances adaptability across varying control scenarios. To ensure both stability and constraint satisfaction, even in the presence of network imperfections, we introduce the edge-end cooperative Simplex architecture (EESA). Semi-physical simulations conducted on both linear and nonlinear systems demonstrate that IMEC surpasses traditional approaches in terms of computation and control efficiency, while EESA ensures runtime stability. The results indicate the applicability potential of this framework in automation systems requiring both high control performance and safety guarantees, such as multi-robot coordination in assembly lines, multi-agent formation control, collaborative robotic arms in precision manufacturing, and real-time coordination in automated warehouses.}, archive = {J_TASE}, author = {Chenhui Xue and Yehan Ma}, doi = {10.1109/TASE.2025.3620556}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22979--22994}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe edge intelligent control with cross-loop constraints and stability guarantees}, volume = {22}, year = {2025}, }
• (2025). Joint trajectory and data acquisition optimization in multi-UAV assisted IoT. TASE, 22, 22962--22978. (www)

  @article{ , abstract = {To tackle the problems of large data volume, multiple devices, and difficult data acquisition in large-scale sensor networks, we propose a multi-UAV (Uncrewed Aerial Vehicle) cooperative trajectory planning and data acquisition method based on the fusion of multi-traveling salesman problem (mTSP) and multi-agent reinforcement learning. Considering the factors such as the wide and uneven distribution of sensing nodes, limited transmission range, and limited energy of UAVs, firstly, large-scale sensing nodes perform intra-cluster data fusion processing by cluster heads, and multi-UAV cooperative trajectory planning acquires cluster head node data to balance the working time of each UAV, thereby improving energy efficiency. Then, the multi-UAV data acquisition problem is modeled as a mTSP, and an improved genetic algorithm with 2-opt optimization operator to optimize the sequence of data acquisition for multi-UAV, so that each UAV acquires the data of the cluster head nodes in its own area in the shortest time. To optimize the UAV’s trajectory based on the data acquisition sequence, we model it as a partially observable Markov decision process, train the model with the counterfactual multi - agent policy gradient method, and solve the agents’ belief allocation problem using the advantage function based on difference reward. Simulation results show that the proposed algorithm has good performance in terms of flight time, data acquisition and energy consumption. Note to Practitioners—In large-scale Internet of Things applications such as smart agriculture, intelligent logistics, and industrial monitoring, collecting data from numerous distributed sensors is a practical challenge, especially when ground access is limited or inefficient. This work addresses the difficulty of timely and energy-efficient data collection using multiple UAVs (Uncrewed Aerial Vehicles). Practitioners often face trade-offs between completing tasks quickly and conserving UAV energy, particularly in complex environments. Our proposed method combines a clustering strategy with an improved genetic algorithm and multi-agent reinforcement learning to guide UAVs to visit data-rich locations efficiently ad avoid collisions. This joint approach allows UAVs to collect more data in less time while maintaining safe and energy-conscious operations. Although reinforcement learning methods can be computationally intensive to train, they offer the advantage of adaptability and robustness once deployed. The techniques developed in this work are especially useful for scenarios where sensor coverage is uneven, and real-time coordination among UAVs is critical. Future extensions could focus on reducing training time and enhancing robustness under dynamic environmental conditions.}, archive = {J_TASE}, author = {Lingyu Zhao and Shiyang Li and Xiaorong Zhu}, doi = {10.1109/TASE.2025.3619537}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22962--22978}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Joint trajectory and data acquisition optimization in multi-UAV assisted IoT}, volume = {22}, year = {2025}, }
• (2025). Fixed-time cooperative model-free sliding mode control for fractional-order multi-motor systems. TASE, 22, 22950--22961. (www)

  @article{ , abstract = {Speed inconsistency in multi-motor systems may lead to mechanical structure damage, low cooperative precision, and even cause equipment breakdown. Traditional control methods may induce pronounced consensus error fluctuations and prolonged convergence time under complex conditions such as system parameter mismatch, sudden load changes, and inaccurate mechanism models. To address these challenges, we propose a fixed-time cooperative model-free sliding mode control (FCM-SMC) method based on a fractional-order ultra-local model (FOULM). Specifically, the proposed approach integrates fractional-order dynamics into the system model to better capture the non-integer order characteristics of multi-motor systems. Then, to estimate the unknown terms of FOULM, a novel fixed-time disturbance observer (FTDO) is designed. Additionally, grounded in the framework of FOULM, the FCM-SMC protocol is designed to ensure that the consensus error of the multi-motor systems converges within a fixed time. Taking motor #1 as an example, compared with existing methods, the error metrics (ME, MAE, and RMSE) of the proposed method are reduced by up to 64.8%, 70.2%, and 69.7%, respectively. Note to Practitioners—This paper was motivated by the problem of speed consensus for multi-motors. This problem presents certain challenges in relevant industrial applications, such as urban rail trains and electric vehicles. In this research, the dynamics of the multi-motor system are formulated in accordance with the fractional-order theory and model-free theory. Subsequently, we demonstrate how to design the FTDO and the FCM-SMC protocol according to the state equations, so that the speed consensus error of the multi-motor system can converge within a fixed-time upper bound. However, it is unable to precisely provide the specific convergence moment. Ultimately, the experimental results validate the practical effectiveness of the proposed approach. The proposed FCM-SMC protocol exhibits the potential for broader applicability and can be effectively extrapolated to diverse categories of multi-motor systems. In the subsequent research, we will further explore the application of the predefined-time algorithm in multi-motor systems under the consideration of time-varying topologies on this basis.}, archive = {J_TASE}, author = {Guanyang Hu and Dezhi Xu and Bin Jiang and Tinglong Pan and Wei Hua}, doi = {10.1109/TASE.2025.3618245}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22950--22961}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time cooperative model-free sliding mode control for fractional-order multi-motor systems}, volume = {22}, year = {2025}, }
• (2025). Autonomous local navigation of spring-based continuum choledochoscopes for cholecystolithotomy. TASE, 22, 22938--22949. (www)

  @article{ , abstract = {Choledochoscopy, a natural orifice transluminal endoscopic surgery, represents an alternative to cholecystolithotomy in comparison to cholecystectomy and open gallbladder-preserving surgery. To reduce surgeon fatigue and increase surgical autonomy, this paper presents an autonomous local navigation (ALN) system for cholecystolithotomy, comprising a novel spring-based choledochoscope (CDS) with a monocular camera, a dataset-independent dynamic adaptive threshold segmentation (DATS) algorithm, and an ALN strategy. During navigation, the DATS detects bile duct bifurcations from the CDS view, selecting the target bifurcation based on bile duct anatomy with a wall-following mechanism. The CDS control method including bending and rotation control is then proposed based on the relationship between the position and direction of the bile duct center and the CDS characteristics. To ensure precise navigation, a closed-loop ALN strategy utilizing three interchangeable proportional controllers is developed, with task-specific proportionality factors. The efficacy of the DATS and CDS is validated through series experiments. Finally, both rigid and soft phantom experiments demonstrate the potential of the ALN system for cholecystolithotomy. Note to Practitioners—This work addresses the growing need for surgical autonomy, particularly in reducing surgeon fatigue during high-volume procedures like cholecystolithotomy. We developed an ALN system featuring a novel spring-based CDS and an ALN strategy to enable efficient and reliable navigation within bile ducts. The system’s key features include a DATS algorithm for bile duct bifurcation detection and a control method for CDS maneuvering, improving the feasibility of autonomous navigation in complex lumen environments. While our approach enhances procedural reliability, limitations remain. The system only relies on monocular camera data, and future improvements could incorporate additional sensing technologies such as contact sensing to further improve navigation safety. Although validated in laboratory experiments, further animal experiments is required before clinical application. Beyond cholecystolithotomy, the ALN system could also benefit other procedures involving tubular anatomy, such as bronchoscopy and ureteroscopy, expanding its application potential.}, archive = {J_TASE}, author = {Jie Wang and Hee Guan Khor and Yingni Wang and Xueling Wei and Yihua Sun and Yiguang Yang and Xu Lu and Longfei Ma and Hongen Liao}, doi = {10.1109/TASE.2025.3615597}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22938--22949}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous local navigation of spring-based continuum choledochoscopes for cholecystolithotomy}, volume = {22}, year = {2025}, }
• (2025). Observer-based dynamic event-triggered finite-time switching control for networked discrete implicit DC circuit systems under aperiodic DoS attacks. TASE, 22, 22925--22937. (www)

  @article{ , abstract = {This article investigates the finite-time switching control issue based on switched observer for networked discrete implicit systems under dynamic event-triggered protocol (DETP) and denial-of-service (DoS) attacks. First, regarding the intermittent characteristic of DoS attacks, a switched observer adapted to such attacks is devised to tackle the unavailability of state information. In addition, to enhance the efficiency of network resource utilization, an improved DETP is introduced, which not only relies on the latest transmission data and the current sampling data, but also combines the changes in their average values to dynamically adjust the triggering threshold. Next, by adopting the singular value decomposition approach, the iterative technique and by constructing the Lyapunov-Krasovskii (L-K) functional correlated with the DoS attack signals, the criteria of singular finite-time boundedness for the networked controlled implicit systems are acquired. The proposed cooperative scheme for DETP and observer-based switching controller promotes both communication efficiency and the resilience of the implicit systems to DoS attacks. Finally, the validity of the mentioned collaborative design switching control method is confirmed by taking direct current motor model and tunnel diode circuit as simulation examples. Note to Practitioners—In real engineering, numerous dynamic systems inevitably need to be modeled as implicit systems, for instance, circuit systems, power systems, etc. It is noteworthy that the core characteristic of implicit systems lies in the inherent algebraic constraints, which necessitates considering not only stability but also regularity and causality in system analysis, thereby making the study of implicit systems more complicated. Furthermore, it is difficult to obtain directly the system state information, and the openness of network environment combined with limited bandwidth resources makes the information transmission channels vulnerable to cyber-attacks, which may degrade the performance of the networked controlled implicit systems or even cause significant damage. To cope with the above challenges, this paper proposes the observer-based attack mode-dependent switching controller design strategy for implicit systems subject to aperiodic DoS attacks, achieving finite-time bounded control of the controlled systems.}, archive = {J_TASE}, author = {Mengjuan Hao and Guangming Zhuang and Guozeng Cui and Jianwei Xia}, doi = {10.1109/TASE.2025.3624969}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22925--22937}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based dynamic event-triggered finite-time switching control for networked discrete implicit DC circuit systems under aperiodic DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Asymptotic tracking control for n-link flexible-joint robots under state constraints and disturbances: Design and experiment. TASE, 22, 22912--22924. (www)

  @article{ , abstract = {This paper presents an asymptotic tracking control strategy for an $n$ -link flexible-joint (FJ) robotic system subject to full-state constraints and external disturbances. First, the original system is transformed into a chained system structure, then a state-dependent function (SDF) is applied to convert the constrained system into an unconstrained one. Subsequently, a novel disturbance observer (DO) is constructed to asymptotically estimate and compensate for unknown disturbances. Concurrently, a neural adaptive tracking controller incorporating a first-order filter is designed, with an adaptive law utilized to approximate the maximum value between the norm of weight vector and the upper bound of approximation error. Compared with the existing DOs and backstepping-based methods to realize the tracking control, the developed method guarantees the tracking error asymptotically converges to zero without violating the predefined constraints. The efficacy of the proposed control algorithm is validated through simulation and experimental studies on a two-link manipulator system. Note to Practitioners—The motivation of this work is to achieve the asymptotic tracking control problem of FJ robots under full-state constraints and external disturbances. In practical engineering applications, FJ manipulators need to move within a certain range to avoid collision, so state constraint is a significant issue. In addition, in the actual industrial control process, the performance and stability of the control system are often affected by unknown external disturbances, which can result in reduced tracking accuracy, longer response time, and even system instability. Therefore, how to deal with the disturbances is also necessary. However, existing literature on FJ robots rarely addresses both state constraints and disturbances simultaneously, thereby limiting their practical applicability. In this paper, a new DO is proposed, so that the asymptotic estimation and compensation for disturbances are realized, even in the presence of mismatched disturbances in the FJ robot system. The proposed approach ensures that the two-link FJ robot system achieves asymptotic tracking of the desired trajectory without violating the predefined state constraints.}, archive = {J_TASE}, author = {Nan Jiang and Zhongcai Zhang and Yang Gao and Weiwei Sun and Yuqiang Wu}, doi = {10.1109/TASE.2025.3624751}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22912--22924}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asymptotic tracking control for n-link flexible-joint robots under state constraints and disturbances: Design and experiment}, volume = {22}, year = {2025}, }
• (2025). Distributed secondary control of DC microgrids under unreliable communication networks. TASE, 22, 22900--22911. (www)

  @article{ , abstract = {This paper investigates the voltage restoration and optimal load sharing problem of direct-current (DC) microgrids with considering random switching topologies and communication delays. The topological random switch is driven by a continuous-time Markov process with its state corresponding to all communication topologies among distributed energy resources (DERs). A distributed communication-based secondary control protocol is proposed to achieve voltage restoration across multiple DERs while optimizing load distribution for cost-effective operation in the mean square sense. Such control protocol is implemented in a fully distributed manner, with each controller only using the local information and the information from its neighbors, thereby enhancing its robustness, flexibility, and scalability. A sufficient condition is given to ensure the mean square exponential stability of the stochastic time-varying delayed systems by constructing a novel topology-dependent degenerate Lyapunov functional. Finally, the effectiveness and robustness of the proposed control protocol are verified by comparing with the random switching topologies and communication delays based on a set of simulation examples. Note to Practitioners—Real-world DC microgrids are vulnerable to unpredictable communication disruptions, such as random denial-of-service (DoS) attacks, random link failures, or packet dropouts, which destabilize coordination among DERs. Traditional secondary control strategies often fail to address the inherent randomness of these disruptions, limiting their applicability in dynamic or adversarial environments. This work introduces a Markovian modeling framework to capture the stochastic nature of communication topology changes, explicitly linking topology randomness to real-world phenomena like cyber-attacks or intermittent connectivity. By treating topology switches as probabilistic events governed by a continuous-time Markov process, the proposed approach enables robust voltage restoration and optimal load sharing, even under communication delays or sudden network failures. The proposed controller ensures mean-square exponential stability using only local and neighbor information. Its key practical advantage is that it remains stable and automatically recovers even when links are temporarily lost, as long as the union graph is connected. This property allows the use of inexpensive, low-bandwidth communication infrastructure and reduces the need for continuous network supervision or manual reconfiguration. The fully distributed design also facilitates scalable deployment as DERs are added or removed.}, archive = {J_TASE}, author = {Haihua Guo and Xiaoran Dai and Siqi Bu and Zijun Zhang}, doi = {10.1109/TASE.2025.3624598}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22900--22911}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed secondary control of DC microgrids under unreliable communication networks}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time active disturbance rejection control for nonlinear systems with mismatched uncertainties: A non-separation principle approach. TASE, 22, 22886--22899. (www)

  @article{ , abstract = {This paper addresses the problem of prescribed-time output feedback stabilization for a class of nonlinear uncertain systems with both external disturbances and mismatched uncertainties. By utilizing a non-separation principle design approach, a novel active disturbance rejection control (ADRC) scheme is proposed. Firstly, a prescribed-time extended state observer (PTESO) is designed to simultaneously estimate unmeasured system states and external disturbances. Secondly, a feedforward-feedback composite controller is developed by integrating the PTESO with a constructive backstepping procedure. Furthermore, the prescribed-time stability of the entire closed-loop system is rigorously analyzed by using a Lyapunov function method. Finally, numerical simulations validate the effectiveness of the proposed control method. Note to Practitioners—In time-critical applications such as missile guidance, spacecraft docking, and target interception, achieving high-precision control within a prescribed timeframe is essential. Prescribed-time control has gained significant attention in recent years, as it allows designers to directly specify a convergence time that meets practical requirements. Moreover, real-world systems often exhibit nonlinear behavior and are subject to various uncertainties—both matched and mismatched—as well as external disturbances. To improve convergence speed, disturbance rejection, and robustness, this paper employs an ADRC framework to solve the prescribed-time output feedback control problem for a class of nonlinear systems with external disturbances and mismatched uncertainties. Compared to existing prescribed-time ADRC methods, the proposed approach relaxes constraints on nonlinear dynamics and accommodates non-differentiable nonlinear terms. This flexibility broadens the algorithm’s applicability, particularly in systems with strong nonlinearities and non-smooth dynamics.}, archive = {J_TASE}, author = {Xixi Shen and Yanzhi Wu and Jiangping Hu and Qingrui Zhou}, doi = {10.1109/TASE.2025.3621472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22886--22899}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time active disturbance rejection control for nonlinear systems with mismatched uncertainties: A non-separation principle approach}, volume = {22}, year = {2025}, }
• (2025). Remote state estimation against essential-information-aware attack tactic in cyber-physical systems. TASE, 22, 22875--22885. (www)

  @article{ , abstract = {From the perspective of attackers, this article proposes a creative essential-information-aware (EIA) attack policy targeting remote state estimation in cyber-physical systems (CPSs), in which adversaries aim at maximizing the trace of estimation error covariance by selectively leveraging crucial information during data transmission. Through manipulating the transmitted signals, the attacker can recognize critical packets exchanged among sensing nodes and allocate higher attack energy level to them, thereby enlarging the attack success probability (ASP) and amplifying the disruptive effect. Then, a scientific power allocation scenario has been designed with the aid of signal-to-interference-plus-noise ratio (SINR), while explicitly accounting for the confinement imposed by limited channel transmission capacity. Furthermore, the relationships among the attack coefficient, ASP, attack cost, and the channel capacity are rigorously analyzed utilizing stochastic methods. The upper bounds for attack coefficient are derived to balance attack impact with resource limitations. Finally, two examples are provided to verify the effectiveness and practicality of the attack strategy. Note to Practitioners—This work has discussed the security problem of remote state estimation in cyber-physical systems (CPSs) from an adversarial stand, with the aim of offering practical insights into system protection. The proposed essential-information-aware (EIA) attack strategy provides a theoretical yet applicable framework for selectively targeting essential information during transmission. By exploiting the real-time importance of transmitted signals, the attacker can intelligently allocate energy resources to maximize estimation degradation. The inclusion of signal-to-interference-plus-noise ratio (SINR) considerations and the quantitative relationship between attack expense and channel capacity further closes to the real-word scenario. The theoretical results and two studies can guide engineers in designing more robust sensor scheduling and transmission encryption mechanisms to counteract such attacks.}, archive = {J_TASE}, author = {Haotong Lv and Chao Cheng and Yiyang Chen and Engang Tian}, doi = {10.1109/TASE.2025.3621293}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22875--22885}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Remote state estimation against essential-information-aware attack tactic in cyber-physical systems}, volume = {22}, year = {2025}, }
• (2025). GateSleepNet: A dual-level spatiotemporal graph-transformer architecture for automatic sleep staging. TASE, 22, 22860--22874. (www)

  @article{ , abstract = {Automatic sleep staging is critical for understanding sleep patterns and diagnosing sleep-related disorders, yet traditional manual scoring methods remain time-consuming, labor-intensive, and subjective. Existing research often fails to fully exploit both cross-spatial and local-global temporal information in polysomnography (PSG) data. To address these challenges, we propose GateSleepNet, a novel dual-level spatiotemporal framework designed to handle the unique properties of PSG data, which comprises multi-channel physiological signals such as EEG, EOG, and EMG that capture temporal and cross-spatial interactions. GateSleepNet combines a Global Spatial Encoder and a Temporal Vision Transformer (ViT) Encoder to effectively capture both local and global temporal features. To enhance cross-spatial understanding, the framework incorporates a Global Spatial Encoder that models inter-channel cross-spatial relationships using sparsely connected graphs. A key innovation of GateSleepNet is the dual-level surrogate loss, which combines global epoch-level accuracy with local temporal consistency, ensuring alignment between predictions and actual sleep patterns. Experimental results on PSG datasets demonstrate the effectiveness of GateSleepNet, achieving high performance in classifying sleep stages and outperforming existing methods. The proposed framework provides a powerful solution for clinical and research applications in sleep medicine, with potential for broader adoption in automated health diagnostics. Note to Practitioners—To capture the unique spatiotemporal characteristics of PSG data, we propose GateSleepNet, a dual-level spatiotemporal framework for automated sleep staging. Implementing this approach requires attention to three key factors. First, multi-channel data must be accurately collected in accordance with the American Academy of Sleep Medicine (AASM) guidelines to ensure reliable input signals. Second, segmenting the signals into multiple patches allows for capturing both local and global temporal features. The number and length of these patches can be tuned to meet specific scenario requirements. Finally, when constructing the graph structure for multi-channel data, adjusting the degree of the regular graph or the number of virtual nodes can modify graph sparsity. These modifications yield computational savings while maintaining the framework’s effectiveness in sleep staging.}, archive = {J_TASE}, author = {Yikun Feng and Yu An and Hongqiang Sun and Zhihong Li and Xi Zhang}, doi = {10.1109/TASE.2025.3621217}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22860--22874}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GateSleepNet: A dual-level spatiotemporal graph-transformer architecture for automatic sleep staging}, volume = {22}, year = {2025}, }
• (2025). Event-triggered asynchronous quasi-zero-sum games for heterogeneous multiagent systems under nonlinear disturbances. TASE, 22, 22845--22859. (www)

  @article{ , abstract = {This paper investigates the cooperative control of heterogeneous networked multiagent systems (MASs) in the presence of unknown nonlinear disturbances through one novel event-triggered asynchronous quasizero-sum game (ET-AQZSG) scheme. The individuals in the MASs are divided into principals and agents on the basis of their respective state values. A sufficient condition is established to prevent Zeno behavior in the ET-AQZSG scheme, ensuring its practical applicability. Additionally, a separable nonlinear functions (SNLFs) is proposed for assessing global stability. By leveraging norm inequalities and ordinary differential equations, a set of slackness conditions is formulated to ensure both feasibility and solution accuracy and promote convergence to the optimal solution. The proposed approach enhances communication flexibility without compromising computational efficiency. Finally, two numerical examples based on the topology of an electromagnetic actuator system and the topology of IEEE 39-Bus system are presented to demonstrate the efficacy of the ET-AQZSG scheme. Note to Practitioners—This work presents an ET-AQZSG scheme for heterogeneous networked MASs under unknown nonlinear disturbances, offering practical solutions to enhance cooperation and reduce computational burdens. In real-world applications, such as UAVs, smart grids and intelligent transportations, MASs often face challenges like limited communication flexibility, high computational burdens, and instability due to nonlinearities. The proposed ET-AQZSG scheme addresses these issues by categorizing agents into principals and subordinates based on state values, enabling asynchronous event-triggered communication. This approach avoids Zeno behavior, ensures global stability via a novel separable nonlinear functions (SNLFs), and simplifies computations by leveraging Euclidean norm analysis instead of $\infty $ -norm, eliminating the need to identify saddle points or extrema. Key advantages, including higher communication efficiency, lower computational overhead, and robust performances. For industry, it can implement this framework in scenarios requiring distributed control, such as cooperative robotics or power grid management, where heterogeneous agents operate under dynamic conditions. By adopting the ET-AQZSG scheme, practitioners can achieve more resilient and scalable multiagent systems with lower resource demands.}, archive = {J_TASE}, author = {Jiansheng Wang and Mingwei Wei and Daqing Wang and Man Jiang and Lifu Gao}, doi = {10.1109/TASE.2025.3621139}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22845--22859}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered asynchronous quasi-zero-sum games for heterogeneous multiagent systems under nonlinear disturbances}, volume = {22}, year = {2025}, }
• (2025). Optimizing multitask industrial processes with predictive action guidance. TASE, 22, 22833--22844. (www)

  @article{ , abstract = {Monitoring complex assembly processes is critical for maintaining productivity and ensuring compliance with assembly standards. However, variability in human actions and subjective task preferences complicate accurate task anticipation and guidance. To address these challenges, we introduce the Multi-Modal Transformer Fusion and Recurrent Units (MMTF-RU) Network for egocentric activity anticipation, utilizing multi-modal fusion to improve prediction accuracy. Integrated with the Operator Action Monitoring Unit (OAMU), the system provides proactive operator guidance, preventing deviations in the assembly process. OAMU employs two strategies: 1) Top-5 MMTF-RU predictions, combined with a reference graph and an action dictionary, for next-step recommendations and 2) Top-1 MMTF-RU predictions, integrated with a reference graph, for detecting sequence deviations and predicting anomaly scores via an entropy-informed confidence mechanism. We also introduce Time-Weighted Sequence Accuracy (TWSA) to evaluate operator efficiency and ensure timely task completion. Our approach is validated on the industrial Meccano dataset and the large-scale EPIC-Kitchens-55 dataset, demonstrating its effectiveness in dynamic environments. Note to Practitioners—Manual assembly remains crucial in industries that require human dexterity and adaptability for tasks that are too complex, customized, or low-volume for full automation. We propose a system to enhance the accuracy of these processes by predicting operator actions and providing real-time guidance through multi-modal data. Its primary function is to ensure that tasks are completed in the correct sequence, thereby reducing human error and boosting productivity. The system is particularly suited for industries like automotive and electronics manufacturing, where precision and flexibility are critical to maintaining high product quality. In addition to providing real-time feedback, the system continuously monitors the sequence and timing of operator actions using multi-modal data. It detects anomalies such as delays, repeated attempts, or missed steps, flags them as bottlenecks, and provides immediate corrective guidance. Such functionality helps operators stay aligned with the expected task flow, minimizes inefficiencies, and reduces cycle time, thereby reducing the need for corrective measures and post-process inspections while delivering immediate operational benefits. This solution offers significant potential for transforming manual assembly workflows, improving accuracy, efficiency, and consistency, while proactively addressing inefficiencies. It represents a key step toward creating smarter, human-centric manufacturing environments that align with the goals of Industry 5.0.}, archive = {J_TASE}, author = {Naval Kishore Mehta and Arvind and Shyam Sunder Prasad and Sumeet Saurav and Sanjay Singh}, doi = {10.1109/TASE.2025.3621005}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22833--22844}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimizing multitask industrial processes with predictive action guidance}, volume = {22}, year = {2025}, }
• (2025). Resilient control of trajectory tracking for cloud-based intelligent connected vehicle under DoS attacks. TASE, 22, 22817--22832. (www)

  @article{ , abstract = {Cloud-based intelligent connected vehicles (CICVs) are among the essential future applications of autonomous driving. Communication between the vehicle and the cloud is carried out through a wireless network, which is susceptible to cyber attacks due to its openness. In order to solve the problem of denial-of-service (DoS) attacks that result in large trajectory tracking errors for CICV, a cross-layer collaborative resilient control strategy is proposed, which integrates offensive-defensive games at the network layer with $H\infty $ control at the physical layer. Firstly, based on the vehicle dynamics model and the DoS attack model, the closed-loop vehicle trajectory tracking control model under DoS attack in sensor-controller (S-C) and controller-actuator (C-A) communication networks is established. Secondly, an offensive-defensive gaming framework in network layer is established, and a cost function considering vehicle control error, attack frequency and attack duration of DoS is designed to optimise the hybrid gaming strategy, thereby constraining the packet loss rate (PLR) caused by DoS attacks. Then, a controller design framework integrating $H\infty $ control and multi-objective co-optimisation is proposed to maintain the comprehensive control performance of vehicle system under DoS attack. Finally, the results of simulation and cloud controller-in-the-loop experiment under typical conditions show that the proposed offensive-defensive gaming strategy in network layer can effectively suppress DoS attacks, and the designed controller improves trajectory tracking accuracy by 47.5% and 147.6% compared to the conventional model predictive controller (MPC) and linear quadratic regulator (LQR), respectively, which significantly enhances the cyber-physical security of the CICV. Note to Practitioners—The motivation for this work stems from the need to develop a secure control strategy for CICVs. Although advanced control systems offer significant advantages for CICVs, developing effective tracking control schemes under cyber attacks remains a challenging issue. Due to packet loss in the uplink and downlink of vehicle-cloud communication caused by DoS attacks, it is unrealistic to receive state and control variables in real-time during tracking control. This packet loss in the communication link can significantly degrade the vehicle’s control performance, such as position and heading errors. To provide a secure operation strategy for CICVs, we consider the actions of a defense system in practical vehicle-cloud communication networks and propose a robust control strategy based on an offensive-defensive game at the network layer. This method can be conveniently applied to the safety control of various networked vehicles and can effectively address the control design problem through convex optimization algorithms.}, archive = {J_TASE}, author = {Chuanlin He and Xing Xu and Haobin Jiang and Jiachen Jiang and Te Chen and Yifeng Long}, doi = {10.1109/TASE.2025.3620997}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22817--22832}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient control of trajectory tracking for cloud-based intelligent connected vehicle under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Distributed robust adaptive consensus control for uncertain nonlinear multi-agent systems with output constraints. TASE, 22, 22807--22816. (www)

  @article{ , abstract = {This paper tackles the consensus tracking control challenge for a category of uncertain nonlinear multi-agent systems subject to unknown time-varying disturbances and output constraints. Even if only a portion of followers can directly receive the leader’s information, by developing a new barrier Lyapunov function and integrating it with the backstepping technique, a distributed robust adaptive state feedback controller is developed, which not only guarantees that the system outputs remain within prescribed constraints but also ensures that the output error between any two followers and the tracking error converge to the origin asymptotically. Neither control design nor theoretical analysis relies on eigenvalue information of Laplacian, thus eliminating the demand for global information and enhancing the utilization efficiency of local communication resources. The effectiveness of the control strategy is ultimately demonstrated through a practical example of single-link manipulators. Note to Practitioners—Consensus tracking has always been a focal issue in multi-agent systems. Practical systems are subject to external disturbances on account of various factors. Moreover, resulting from safety requirements and environmental limitations, the system outputs are constrained. In this paper, a distributed robust adaptive consensus tracking control strategy for output-constrained multi-agent systems is proposed, applying the robust control method, adaptive techniques, and the barrier Lyapunov function to multi-agent systems. The theoretical analysis manifests that even when the system is affected by external disturbances, the proposed strategy can achieve asymptotic consensus tracking of all followers to the leader while ensuring that the output constraints are always satisfied. The effectiveness and practical application value of the control strategy are verified by a simulation of multiple single-link manipulators.}, archive = {J_TASE}, author = {Keli Liu and Zong-Yao Sun and Jiao-Jiao Li and Chih-Chiang Chen}, doi = {10.1109/TASE.2025.3620837}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22807--22816}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed robust adaptive consensus control for uncertain nonlinear multi-agent systems with output constraints}, volume = {22}, year = {2025}, }
• (2025). Intelligent critic design with policy transfer for wastewater treatment processes. TASE, 22, 22796--22806. (www)

  @article{ , abstract = {Wastewater treatment plays a meaningful role in environmental protection, water recycling and public health, and its optimal control offers significant economic and social value. However, the wastewater treatment plant is a large-scale nonlinear system, and its own operation is affected by a series of uncontrollable factors such as the flow rate and composition of the incoming wastewater. This makes it extremely challenging for traditional control methods to realize precise control. To address these issues, this paper proposes a novel knowledge-guided policy optimization control method by integrating policy transfer and adaptive dynamic programming. First, an adaptive critic policy transfer framework based on the source domain knowledge selection method is introduced to construct knowledge extraction and data mining structures to enhance the learning efficiency of the target domain agent. Second, a source domain knowledge selection module is proposed to adaptively adjust the knowledge extracted by the target domain agent. Third, an adaptive termination module is designed to determine when the immature policy in the source domain should terminate during the target domain learning process. Finally, the theoretical stability of the method in this paper is demonstrated by designing a reasonable Lyapunov function. The system performance of the wastewater treatment plant under dry, rainy, and stormy weather conditions is evaluated, ultimately demonstrating the superior performance and adaptability of the proposed method. Note to Practitioners—The motivation of this work is to mitigate the performance degradation of the agent caused by environmental changes in the wastewater treatment process through the use of a knowledge transfer approach. Traditional control methods, which often rely on single-environment learning and prioritize current environment objectives, struggle to generalize across varying conditions (e.g., weather changes and diverse control goals). To bridge this gap, we propose a novel knowledge-transfer-based learning framework. It enhances agent adaptability and learning efficiency through policy transfer. To address source-target domain knowledge mismatch, a source domain knowledge selection module is introduced, enabling adaptive selection and adjustment of relevant source domain knowledge during training for improved target task performance. An adaptive termination module is designed to prevent negative transfer by determining the optimal time to discard immature source domain policies during target domain learning. Finally, to address the lack of rigorous theoretical analysis in the optimal control of wastewater treatment based on policy transfer, this paper proposes a reasonable Lyapunov function to ensure the theoretical stability of this method.}, archive = {J_TASE}, author = {Ding Wang and Ning Gao and Xin Li and Kai Ye and Junfei Qiao}, doi = {10.1109/TASE.2025.3620541}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22796--22806}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intelligent critic design with policy transfer for wastewater treatment processes}, volume = {22}, year = {2025}, }
• (2025). Quantified event-triggered control for finite-time projective synchronization of fractional fuzzy inertial delayed neural networks with uncertainties. TASE, 22, 22785--22795. (www)

  @article{ , abstract = {This paper focuses on the finite-time projective synchronization (FTPS) of fractional fuzzy inertial delayed neural networks (FFIDNNs) with parameters perturbation via several quantified event-triggered mechanisms (QETMs). First, applying Laplace transform, a linear fractional finite-time inequality is established for the infinite time interval rather than a finite time interval. Next, three kinds of QETMs are designed, which can respond flexibly to practical challenges. By the established inequalities and the designed controller, some novel criteria are derived to guarantee FTPS for the considered system. Moreover, the settling time is estimated and the event interval has a positive lower bound to avoid the Zeno behavior. Finally, the accuracy of theoretical findings is confirmed via numerical examples. Note to Practitioners—Numerous engineering systems, such as autonomous vehicles, UAV formation control, and smart grids, demand finite-time control to enhance tracking accuracy. Thus, this paper investigates FTPS of FFIDNNs via QETMs. The proposed QETMs can offer a communication-efficient solution that markedly reduces communication and computational loads. The parameters in these QETMs are flexibly adjusted according to engineering scenarios. For resource-constrained systems such as smart grids, increasing the threshold parameter $h_{2}$ reduces communication loads. For high-precision scenarios such as autonomous vehicles, decreasing $h_{2}$ improves tracking accuracy. For complex UAV formation control, decreasing $h_{2}$ ensures high accuracy during take-off or tight formation, while increasing $h_{2}$ cuts communication loads during cruise.}, archive = {J_TASE}, author = {Weiwei Zhang and Zhenjie Wang and Hai Zhang and Dingyuan Chen and Jinde Cao and Mahmoud Abdel-Aty and Zhao-Dong Xu}, doi = {10.1109/TASE.2025.3617500}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22785--22795}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quantified event-triggered control for finite-time projective synchronization of fractional fuzzy inertial delayed neural networks with uncertainties}, volume = {22}, year = {2025}, }
• (2025). Mathematical evolution of origami structures and their applications in soft robotics. TASE, 22, 22766--22784. (www)

  @article{ , abstract = {Origami, as an ancient art form, has become a critical design paradigm in science and engineering. While numerous studies have explored various origami structures and materials, they often overlook the fundamental principles that unite them. This review seeks to bridge that gap by examining origami engineering through the perspective of mathematical evolution. We outline a continuous developmental process, from kinematics-driven planar patterns to innovative non-rigid, energy-responsive systems and mechanics of active materials. A thorough comparison of key origami families reveals the essential balance between geometric simplicity and mechanical complexity. Furthermore, we introduce advanced modeling and design techniques, including numerical simulation, inverse design, and active material mechanics, particularly in the application of soft robotics. This study establishes a foundational framework for future research and inspires the design of sophisticated, functional systems. Note to Practitioners—Origami structures, particularly those exhibiting bistable properties, exhibit potential in diverse fields such as soft robotics, deployable systems, and flexible electronics. However, the design and optimization of these structures for real-world applications are still challenging. This paper transcends the categorization of origami patterns by uncovering the evolutionary connections among various origami families, providing a more in-depth theoretical foundation for selecting appropriate structures. Another focus of this work is a critical analysis of engineering trade-offs, particularly between deployability and stiffness, which are essential for optimizing applications in dynamic environments like soft robotics. We also provide a comprehensive overview of state-of-the-art modeling tools and introduce innovative reverse design strategies that streamline the transition from performance specifications to conceptual prototypes. For practitioners in the field of robotics, this review investigates the integration of stimulus-responsive materials and the associated modeling challenges, aiming to facilitate a deeper incorporation of origami technology into engineering practice.}, archive = {J_TASE}, author = {Tao Ren and Yifan Chen and Yujia Li and Yang Yang and Yonghua Chen and Simon X. Yang and Yingtian Li}, doi = {10.1109/TASE.2025.3616407}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22766--22784}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mathematical evolution of origami structures and their applications in soft robotics}, volume = {22}, year = {2025}, }
• (2025). Adaptive sub-nanometer control of a piezoelectric positioning platform. TASE, 22, 22755--22765. (www)

  @article{ , abstract = {This paper reports an asymmetric Bouc-Wen (ABW) hysteresis model and a hybrid control algorithm based on multi-modal Bayesian gradient optimization (MBGO) for trajectory tracking in the micro-positioning phase of a piezoelectric positioning platform. First, a system-level dynamic model capable of expressing hysteresis nonlinearity is established based on the asymmetric Bouc-Wen model. Second, an MBGO parameter identification algorithm based on Particle Swarm Optimization (PSO) is proposed to improve the characterization capability of the hysteresis model. Subsequently, a feedforward adaptive fuzzy PID (FF-AFPID) composite controller is designed by compensating the hysteresis nonlinearity through the ABW inverse model while dynamically adjusting PID parameters with adaptive fuzzy rules. Through triangular and sinusoidal trajectory tracking experiments, the root mean square errors were reduced to 0.112 nm and 0.103 nm by the FF-AFPID, with an improvement of 74.944%, 57.088% (triangular) and 77.511%, 57.083% (sinusoidal) over the FF-PID and FF-FPID algorithms, respectively. The results demonstrate that the trajectory tracking of performance the positioning platform in micro-positioning phase was significantly enhanced by FF-AFPID, with the maximum error being suppressed to sub-nanometer levels. Note to Practitioners—The piezoelectric positioning platform, characterized by millimeter-scale travel range and nanometric positioning accuracy, is recognized as a critical component in atomic-scale manufacturing systems. However, the trajectory tracking accuracy in micro-positioning phase is fundamentally limited by inherent piezoelectric hysteresis nonlinearities, strong parameter coupling, and dynamic disturbances, making sub-nanometer accuracy particularly challenging to achieve. To address these limitations, a novel control methodology is developed in this paper to enhance trajectory tracking accuracy and improving disturbance rejection capabilities during micro-positioning phase.}, archive = {J_TASE}, author = {Siyuan Meng and Jiankang Jiang and Qian Ju and Yong Wang and Dongmei Wu and Wei Dong and Ming Pang and Junhui Zhu and Changhai Ru}, doi = {10.1109/TASE.2025.3615572}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22755--22765}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive sub-nanometer control of a piezoelectric positioning platform}, volume = {22}, year = {2025}, }
• (2025). MMFSeg: Multi-structure multi-feature fusion for segmentation of road potholes. TASE, 22, 22742--22754. (www)

  @article{ , abstract = {Road pothole segmentation is important for the driving safety of autonomous vehicles, especially in unstructured or rural environments. Recently, many effective multi-modal fusion networks have been proposed for road pothole segmentation. However, most of them adopt two encoders with the same type of structure, such as only convolutional neural network (CNN) or only Transformer, to extract features from different modalities. This overlooks the fact that the information richness of features extracted from different modalities varies across the types of features, such as CNN features or self-attention features. To provide a solution to this issue, we design a novel RGB-Disparity segmentation network, named MMFSeg, by adopting the two types of structures as encoders. Specifically, we adopt CNN and Transformer as encoders to extract two types of features from each modality, and also propose a late-fusion multi-feature alignment fusion module to fuse the two types of features with different numbers of channels. Experimental results demonstrate that our network outperforms well-known networks, and can trade-off between accuracy and efficiency. Note to Practitioners—This study is driven by the challenge of segmenting road potholes to ensure the safety of autonomous driving. Our proposed network can utilize different structures to extract diverse feature information from each modality of data, thereby achieving more accurate road pothole segmentation results. Our proposed fusion module effectively fuses diverse features from each modality through a late fusion strategy. Our method validates effectiveness on multiple datasets. This work contributes to the safety of autonomous vehicles by enhancing their driving performance under poor road conditions. Our network achieves superior performance compared with existing networks, thereby enabling more effective deployment of downstream tasks, such as path planning and navigation, in autonomous driving systems.}, archive = {J_TASE}, author = {Zhen Feng and Yanning Guo and Rui Fan and Yuxiang Sun}, doi = {10.1109/TASE.2025.3613629}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22742--22754}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MMFSeg: Multi-structure multi-feature fusion for segmentation of road potholes}, volume = {22}, year = {2025}, }
• (2025). Learning-aided evolutionary algorithm for solving energy-minimized deadline-constrained task scheduling problem in human-cyber-physical systems. TASE, 22, 22729--22741. (www)

  @article{ , abstract = {This work addresses an energy-minimized deadline-constrained task scheduling problem in human-cyber-physical systems. It consists of three subproblems: processor allocation, task sequencing, and processor frequency scaling. A Learning-aided Evolutionary Algorithm (LEA) is proposed to efficiently find its reliable and high-quality solutions. It incorporates a bidirectional long short-term memory network-embedded autoencoder trained via end-to-end self-supervised learning. The model extracts the interconnections among the three strongly-coupled subproblems, enabling effective global search in a low-dimensional feature space. A parallel framework with two co-evolved subpopulations, one using the autoencoder and another undergoing regular evaluation in the original search space, is constructed. To balance LEA’s exploration and exploitation, a deep reinforcement learning-based search operator selection scheme is introduced, using a novel feedback-based reward function to guide operator selection for each subpopulation. Numerical experiments demonstrate that LEA surpasses several recently developed methods in finding high-quality schedules in a reasonable time. Note to Practitioners—In a human-cyber-physical system, heuristics are commonly used to solve task scheduling problems. However, fast dispatching rules tend to perform poorly. Evolutionary algorithms can identify relatively high-quality schedules but are highly time-consuming, especially for population-based methods that iteratively evaluate fitness functions. To balance computational burden and solution quality, our idea is to combine two machine learning methods with evolutionary algorithms. First, a self-supervised autoencoder enhances global search capability by reducing the complexity of the search space. Second, a deep reinforcement learning-based operator selection scheme balances exploration and exploitation. This hybrid approach enables engineers to find high-quality schedules for the considered problems in a short time. Theoretical analysis and experimental results demonstrate that our proposed method outperforms its competitive peers.}, archive = {J_TASE}, author = {ZhengCai Cao and HaoRan Zhou and ChengRan Lin}, doi = {10.1109/TASE.2025.3611625}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22729--22741}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-aided evolutionary algorithm for solving energy-minimized deadline-constrained task scheduling problem in human-cyber-physical systems}, volume = {22}, year = {2025}, }
• (2025). Edge cascading failure model for command and control networks: Integrating edge importance and edge hierarchy. TASE, 22, 22715--22728. (www)

  @article{ , abstract = {Most existing cascading failure models for command and control networks (C2Ns) primarily focus on node-level failures while overlooking edge-level failures. To address this issue, this paper proposes a novel edge-based cascading failure model that integrates edge importance and hierarchical structure. First, both local and global edge importance are quantified through edge degree and ego network edge betweenness, and an improved definition of edge importance is introduced by incorporating an enhanced bridging coefficient. Second, based on the hierarchical structure of network nodes, edge hierarchy is defined, and a new initial edge load formulation is proposed by combining edge hierarchy and importance. Third, to reflect load imbalance in realistic networks, a nonlinear edge capacity model is constructed, and a non-uniform, adjustable load redistribution strategy based on the residual capacity of neighboring edges is developed to accommodate network hierarchy. Simulation experiments are conducted to validate the proposed model. The results show that, in two representative C2N networks, the degradation rates of average efficiency and connectivity coefficient are 22.42% and 73.70% for Network 1, and 13.84% and 14.52% for Network 2, respectively. These results demonstrate that the proposed model significantly improves the robustness and resilience of C2Ns under cascading failure scenarios. Note to Practitioners—This study addresses a critical issue in C2Ns: cascading failures, with a particular focus on edge-level failures, which are often overlooked compared to traditional node-based approaches. The motivation stems from the increasing demand for robust and reliable C2Ns in military operations, where communication link disruptions can lead to severe operational consequences. The proposed model integrates edge importance and hierarchical structure to offer a more comprehensive approach to understanding and mitigating cascading failures. Edge importance is defined through a combination of edge degree, ego network edge betweenness, and an improved bridging coefficient, capturing both local and global network features, as well as structural and functional characteristics. This enables more accurate identification of critical edges that are vital for maintaining network stability. A nonlinear edge capacity model and a non-uniform, adjustable load redistribution strategy based on residual neighboring capacity are introduced to enhance the model’s ability to manage load imbalances under adversarial conditions. Practitioners can apply the proposed model—developed from an abstraction of real-world combat systems and incorporating edge importance and hierarchical structures—to enhance the resilience and stability of C2Ns. By identifying critical edges and deploying effective load redistribution strategies, the model addresses vulnerabilities in large-scale network infrastructures. The current model has been validated under simulation conditions; future work will incorporate hardware-in-the-loop (HIL) simulation to evaluate its performance and practicality in more realistic environments. Ongoing research will focus on key challenges such as dynamic network topologies and fluctuating traffic loads, while progressively adapting the model to meet the demands of dynamic C2N environments and cross-domain collaborative defense, thereby further enhancing overall network resilience.}, archive = {J_TASE}, author = {Bo Chen and Lingdong Sun and Yufeng Chen and Xiue Gao and Zhengtao Xiang and Yulin Zhang and Shifeng Chen}, doi = {10.1109/TASE.2025.3621461}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22715--22728}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Edge cascading failure model for command and control networks: Integrating edge importance and edge hierarchy}, volume = {22}, year = {2025}, }
• (2025). A robotic system for medical hoses manipulation and quality check. TASE, 22, 22699--22714. (www)

  @article{ , abstract = {This paper proposes a robotic based system, equipped with suitably developed mechatronic tools, able to fine manipulate medical hoses directly from the production line and during the execution of quality check tasks. The system combines suitably developed technologies (i.e., tactile sensors and cutting machine) with methodological approaches (i.e., tactile indicator, blind search methods, image processing) in order to automatize the manipulation of very thin and soft linear deformable objects, where the main challenge is the very small forces involved during manipulation. The work presents the system testing during the execution of two real quality check scenarios: verification of the hose internal diameter and checking of the hose section shape and dimensions. For the internal diameter check, a suitably defined contact indicator based on tactile signals is proposed, and it is combined with a blind search method, based on the use of a spiral path, in order to generalize the task execution, by overcoming limitations due to the possible misalignment among grasped area and hose end. For the inspection of the hose section, a specific cutting tool has been developed and image processing methods have been implemented, by means of microscope feedback, for the centering and focus of the hose sample to be analyzed. Note to Practitioners—This paper is motivated by the problem of robotic manipulation for quality check of medical hoses, which are deformable and very thin (typical diameters are equal to $0.5-2{\,}$ mm). On the basis of customer application, the producer is requested to check up to 100% of the lot. Currently, quality control operations are carried out manually by human operators, executing very alienating tasks, which lead to an increasing percentage of hoses check errors over long working shifts. In this paper, both hardware and software details needed to reproduce the system are reported: tactile sensors compatible with commercial parallel gripper, used to manipulate hoses during the interaction with the other mechanical tools; a cutting machine with a microscope used for the inspection of the hose section; an approach for the execution of standard Go/No-Go tests, exploited for hose internal diameter evaluation; an approach for the centering and focus of microscope images, implemented for the inspection of the hose section. The proposed approaches are also presented in algorithmic fashion in order to simplify the implementation for interested reader. The system has been demonstrated in a real production scenario. Further refinements are planned in order to speed up the system, to manage the failures and to extend the proposed solution to longer hoses and hoses with additional type of sections.}, archive = {J_TASE}, author = {Andrea Govoni and Gianluca Laudante and Michele Mirto and Olga Pennacchio and Gianluca Palli and Salvatore Pirozzi and Nicola Spatarella}, doi = {10.1109/TASE.2025.3620597}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22699--22714}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A robotic system for medical hoses manipulation and quality check}, volume = {22}, year = {2025}, }
• (2025). Online opacity verification of networked discrete event systems modeled with labeled petri nets. TASE, 22, 22686--22698. (www)

  @article{ , abstract = {This paper addresses the online verification of current-state and $k$ -step opacity in networked discrete event systems modeled with labeled Petri nets. In networked transmissions, there are two main communication disruptions, namely, delays and losses, which may result in incomplete observation sequences, thus affecting the accuracy of traditional online opacity verification methods. To solve such an issue, the notions of current-state and $k$ -step opacity of an observation are generalized by incorporating the effects of the two disruptions. As shown in the literature, a secret is described as a subset of markings for the generality of the approach. Under the assumptions that the unobservable subnet is acyclic and the system is bounded, integer linear programming problems (ILPPs) are formulated. Then, online algorithms are designed by solving ILPPs to respectively verify the current-state and $k$ -step opacity of an observation that may be incomplete due to communication delays and losses. Furthermore, in the absence of communication disruption, fewer constraints and variables are considered in the proposed ILPPs compared with the existing approaches, which contributes to a lower computational overhead. Finally, the efficiency and efficacy of the proposed methods are illustrated through an example and a case study. Note to Practitioners—Opacity is a critical property in highly automated systems such as manufacturing systems, power plants, and smart grids, ensuring that certain system behaviors remain opaque to external entities. This research presents a novel approach for online opacity verification in discrete event systems modeled by labeled Petri nets, explicitly accounting for network-induced delays and losses. By formulating the verification process as an integer linear programming problem, this method provides a computationally efficient and practical solution that can be applied to real-world networked systems, helping engineers ensure information security and privacy in dynamic environments.}, archive = {J_TASE}, author = {Tengbo Li and Huorong Ren and Ruotian Liu and Maria Pia Fanti and Zhiwu Li}, doi = {10.1109/TASE.2025.3614662}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22686--22698}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online opacity verification of networked discrete event systems modeled with labeled petri nets}, volume = {22}, year = {2025}, }
• (2025). Observer-based dual-memory controller for lurie-type CPSs subject to synergy delays and DoS attacks: Application to DC motor system. TASE, 22, 22674--22685. (www)

  @article{ , abstract = {This paper concerns the observer-based dual-memory control for Lurie-type cyber-physical systems (CPSs) subject to denial of service (DoS) attacks. First, an augmented switching Lurie-type CPSs model is established to analyze the impact of DoS attacks and synergy time delays on controller performance. Second, the memory-dynamic event-triggered scheme (MDETS) efficiently utilizes bandwidth resources by adapting the number of historical data packets transmitted based on system state fluctuations, meanwhile, the controller design incorporates constant time-delay compensation to improve system stability. Third, the exponential stability conditions for the closed-loop Lurie-type CPSs are derived via Lyapunov theory. Additionally, the co-design of the observer gain, controller gain, and event-triggering matrices is formulated using linear matrix inequalities (LMIs). Finally, the effectiveness of the proposed control scheme is validated through application to a DC motor system. Note to Practitioners—This study addresses the stability challenges faced by industrial automation systems—particularly DC motor control—under the influence of network delays and denial-of-service (DoS) attacks. Traditional sampling control methods often suffer from redundant communication or delayed responses during network disruptions. Our proposed dual-memory controller integrates a memory-dynamic event-triggered mechanism (MDETS) and delays compensation method, adaptively adjusting historical data transmission to reduce communication load (triggering rate decreased from 23.31% to 16.04%) while suppressing signal fluctuations. Co-designed using Lyapunov theory and linear matrix inequalities (LMIs), the framework enhances resilience against DoS attacks and ensures robust real-time control. Current limitations include idealized assumptions of periodic attacks and simplified network conditions. Future work will focus on adaptive triggering mechanisms, hardware-in-the-loop validation, and extensions to multi-agent CPSs (e.g., smart grids, and autonomous manufacturing). This research contributes to the development of efficient and secure control strategies for industrial automation systems, offering practical solutions for real-world applications.}, archive = {J_TASE}, author = {Qing Zhu and Kaibo Shi and Bin Guo and Da Chen and Jinglei Tan and Xiao Cai}, doi = {10.1109/TASE.2025.3614384}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22674--22685}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based dual-memory controller for lurie-type CPSs subject to synergy delays and DoS attacks: Application to DC motor system}, volume = {22}, year = {2025}, }
• (2025). Source seeking for quad-rotor aircraft with actuator dynamic uncertainties via small-gain approach. TASE, 22, 22663--22673. (www)

  @article{ , abstract = {This paper investigates the source seeking problem for quad-rotor aircraft system in the presence of actuator dynamic uncertainties and presents the semi-global asymptotic practical stability for the closed-loop extremum-seeking system. A new hierarchical two-layer extremum-seeking control framework is presented which consists of a slow quasi-steady-state extremum-seeking algorithm and a fast transient-state dynamics reference-tracking system, the notion of input-to-output practical stability is adopted to describe the performance of different subsystems. Compared to the existing results, this paper eliminates the need for the assumption that the controlled system must exhibit exponential stability at a specific equilibrium point or manifold. By leveraging the nonlinear small-gain theorem and incorporating a robust parameter design approach under small-gain condition, the semi-global practical stability of the overall system is ensured, even in the presence of uncertainties in the actuator dynamics. Finally, the effectiveness of the proposed method is verified through physical experiment and numerical simulation. Note to Practitioners—The rapid advancement of control and optimization algorithms for autonomous systems has significantly broadened the deployment of drones and autonomous ground vehicles across a wide range of industrial and societal domains. These technologies offer promising solutions to complex control problems in applications such as precision agriculture, disaster response, and industrial automation. However, real-world physical systems are often subject to dynamic uncertainties, unmodeled dynamics, and external disturbances, which can result in conservative control strategies and degraded system performance. Therefore, the development of robust and adaptive control methods that can accommodate diverse environmental conditions and inherent system constraints remains a fundamental challenge in control engineering. This study addresses the source-seeking problem for drones operating in spatially distributed signal fields. A novel theoretical framework is proposed to mitigate the restrictive assumptions commonly imposed by conventional extremum-seeking algorithms on the tracking performance of controlled systems. By leveraging this relaxed formulation, the scalability and applicability of perturbation-based extremum-seeking control are significantly enhanced for drones with heterogeneous performance characteristics. To validate the proposed method, physical experiments are conducted in outdoor environments. The results demonstrate the ability of drones to autonomously navigate toward and localize signal sources in the presence of modeling uncertainties and external disturbances. These findings highlight the practical relevance of the proposed approach to real-world source-seeking tasks, such as fire source localization, hazardous gas leak detection, and related applications. Future research will explore the extension of this framework to cooperative multi-agent source-seeking tasks to further improve the efficiency, scalability, and robustness of autonomous localization systems.}, archive = {J_TASE}, author = {Jingwei Xu and Zhanxiu Wang and Zhenghong Jin}, doi = {10.1109/TASE.2025.3623448}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22663--22673}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Source seeking for quad-rotor aircraft with actuator dynamic uncertainties via small-gain approach}, volume = {22}, year = {2025}, }
• (2025). Model-and-data-driven adaptive frequency control for microgrid systems. TASE, 22, 22655--22662. (www)

  @article{ , abstract = {This paper proposes a novel model-and-data-driven adaptive frequency control (MDAFC) for microgrid (MG) systems. The proposed MDAFC includes two control loops, i.e., an adaptive model predictive control (AMPC) loop and a model free adaptive control (MFAC) loop. The AMPC loop not only utilizes the exact model information to improve the control performance but also employs an unscented Kalman filter (UKF) to estimate the unknown parameters of the internal prediction model to improve the robustness to a certain degree. The MFAC loop is designed to address the unmodeled dynamics, nonlinear uncertainty and disturbance of the MG system by virtue of its adaptation mechanism and the data-driven characteristics. Therefore, the MFAC loop can compensate the poor impact of the inaccuracy model information on the AMPC method. To validate the effectiveness of the proposed method, a low inertia MG system containing renewable energy sources (RESs) is considered in this research. Simulation results show that the proposed method can achieve a high control performance. It can effectively cope with the frequency fluctuation caused by RESs, large load consumption, and other unknown uncertain factors. Compared with the existing single AMPC loop and the single proportional integral control loop, the proposed dual-loop-based MDAFC performs better since the two control loops cooperate with each other to leverage their advantages and compensate for their shortcomings. Note to Practitioners—Frequency is a critical variable in ensuring stability and power quality of the MG systems. The frequency dynamics of the MG can be represented using a well-established mathematical model. However, the increasing integration of RESs, the multi-energy coupling between regional grids and equipment, and the diverse electric loads have significantly increased the uncertainty of the model and the nonlinearity of the MG systems, thus posing many difficulties in frequency control. Therefore, this study proposes an MDAFC for the MG systems by combining an AMPC loop with an MFAC loop. The AMPC loop utilizes mechanism model of MG to achieve a good control performance with the help of the predictive information. The MFAC loop is used to address the nonlinear unmodeled dynamics, the system uncertainties and the exogenous disturbances by virtue of the input and output data. The complementary interaction between the two control loops can improve frequency control performance. Furthermore, the proposed MDAFC method adapts an iterative computational structure, which can offer straightforward implementation for the nonlinear and uncertain MG systems with light computation burden, and the proposed MDAFC is easy to implement in practical applications.}, archive = {J_TASE}, author = {Weichao Wang and Ronghu Chi and Yang Liu and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3622101}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22655--22662}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model-and-data-driven adaptive frequency control for microgrid systems}, volume = {22}, year = {2025}, }
• (2025). Enhanced disturbance attenuation for PMSM speed control based on advanced fast reaching and iterative composite compensation. TASE, 22, 22638--22654. (www)

  @article{ , abstract = {Torque ripples and disturbances pose significant obstacles to achieving superior speed regulation performance of the permanent magnet synchronous motor (PMSM). This study proposes an enhanced sliding mode control (SMC) approach to further elevate the dynamic responsiveness and antidisturbance ability of the PMSM. First, an advanced fast reaching law (AFRL), which introduces system errors into the power and exponential terms, is proposed to simultaneously reduce the reaching time and sliding mode chattering. A modified sliding mode observer (MSMO) is then constructed to assess the variation of the load torques, and an iterative learning law is designed to learn the periodic disturbances. The integration of the MSMO and iterative learning law forms an iterative composite compensation strategy, which effectively elevates the observation accuracy of system disturbances and strengthen the system’s robustness. The enhanced sliding controller is consequently developed according to the AFRL and the iterative composite compensation. The stability of the AFRL and the closed-loop PMSM system, as well as the convergence of tracking errors, are rigorously substantiated using the Lyapunov theory. Experimental results reveal that the proposed controller exhibits smaller torque ripples, faster convergence speed, reduced chattering, and considerable antidisturbance capability. Note to Practitioners—The inspiration for this article arises from the difficulties encountered in attempting to apply a PMSM to potential high-end industrial applications, such as robotics, numerical machining, and precise instruments. These scenarios demand remarkable dynamic responsiveness and antidisturbance ability, requiring a sophisticated control system. Most of the existing control methods exhibit contradictions in dynamic performance, including overshoot, convergence speed, and disturbance rejection capability. Therefore, this article proposes an enhanced disturbance attenuation control strategy for high-performance control of a PMSM system. This approach can be applied to a servo system equipped with PMSMs to enhance the dynamic performance and increase the robustness of the system.}, archive = {J_TASE}, author = {Yong Yang and Gaofeng Yu and Xia Liu and Deqing Huang and Yanan Li}, doi = {10.1109/TASE.2025.3621903}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22638--22654}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhanced disturbance attenuation for PMSM speed control based on advanced fast reaching and iterative composite compensation}, volume = {22}, year = {2025}, }
• (2025). Reactive self-collision avoidance for dual-arm robots using a temporal feature modeling and fusion network. TASE, 22, 22625--22637. (www)

  @article{ , abstract = {Dual-arm robots (e.g., humanoid robots) possess substantial potential for executing collaborative tasks in universal scenarios. However, the workspaces of the individual arms frequently overlap, rendering self-collision avoidance crucial for maintaining safe robotic operations. Existing motion planning-based methods demonstrate inadequate real-time efficacy, and learning-based distance proxy methods are subject to elevated false positive rates. To address these challenges, we present a novel minimum distance prediction neural network for reactive collision avoidance of dual-arm robots, which considers the continuous motion of the robotic arms and the interrelation of joint configurations. The temporal joint configurations are encoded and divided into historical and current features. A state-space model is utilized to capture the temporal dependency of historical features. A self-attention mechanism is employed to model the hidden relationships among current features. The integration of historical and current features via a cross-attention mechanism followed by a gated fusion module allows for precise prediction of the minimum distance between the dual arms. Simulations and real-world experiments, including human-robot interaction and autonomous grasping tasks, were conducted using two redundant robotic arms. The proposed method achieves an average error of 1.804 cm in minimum distance prediction. In dual-arm autonomous grasping experiments, an average error of 1.471 cm is attained. Our approach has improved accuracy by 38.08% over the state-of-the-art methods. No collisions occurred throughout all real-world experiments. This method holds promise for extensive applications of dual-arm robots. The code is accessible at https://github.com/XuejinLuo/SelfCollision Note to Practitioners—Dual-arm robots have shown great potentials in collaborative tasks. The collision between the two redundant robotic arms has to be addressed during the motion. A straightforward solution is predicting the minimum distance between the two arms in real time and incorporating the distance information into the framework of optimal control for self-collision avoidance. Existing state-of-the-art methods have accuracy issues in the distance prediction. We propose a novel neural network for real-time minimum distance prediction for reactive collision avoidance of dual-arm robots. we introduce a temporal model and self attention mechanisms into the network, which is supposed that the temporal correlations between the joint configurations of adjacent frames can be effectively modeled and fused, leading to more accurate and smoother prediction of the minimum distance. Our proposed network outperforms the state-of-the-art methods, achieving higher accuracy in minimum distance prediction with a reduced false positive rate. The proposed network is further incorporated into a quadratic optimization framework for realizing the self-collision avoidance during task execution. Comprehensive real-world experiments have demonstrated the effectiveness of our method.}, archive = {J_TASE}, author = {Xuejin Luo and Runshi Zhang and Siqin Yang and Zhen Sun and Ruizhi Zhang and Junchen Wang}, doi = {10.1109/TASE.2025.3620883}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22625--22637}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reactive self-collision avoidance for dual-arm robots using a temporal feature modeling and fusion network}, volume = {22}, year = {2025}, }
• (2025). A Q-learning-based particle swarm optimization for aircraft routing and scheduling in airport terminal area. TASE, 22, 22609--22624. (www)

  @article{ , abstract = {As the airport terminal area becomes progressively crowded, costly delays and adverse environmental impact due to excessive fuel burn require effective aircraft routing and scheduling in the airport terminal area. Aircraft routing and scheduling in the airport terminal area refers to the safe and efficient movement of aircraft among airport facilities such as runways, aircraft stands and taxiways. It is a hybrid optimization problem that involves both the airport ground movement problem and the aircraft sequencing problem. The paper investigates the hybrid problem with constraints such as the safe distance between aircraft, queuing limit at the runways, and release time difference for sake of safety. The objective is to minimize the average taxiing time of the aircraft. A particle swarm optimization and Q-learning based aircraft routing and scheduling algorithm (PSO-QL-ARS) is proposed for the problem considered. The proposal adopts the main framework of PSO. It consists of three major components: the ideal shortest path algorithm, the semi-no-wait schedule generation method and the Q-learning based particle evolving method. The shortest path algorithm takes into consideration turning time according to the turning angles on the path and different taxiing speeds depending on the taxiway types. The semi-no-wait schedule generation method is presented to compute the feasible trajectory plan for an aircraft, including the waiting time at each vertex and the taxiing time on each taxiway. It attempts to place the waiting time of an aircraft on the aircraft stand. The Q-learning-based particle evolving method employs a Q-table to select the finest evolving action which is used to update the position of the given particle. The proposed algorithm is compared with four baseline algorithms. The experimental results show that the proposal outperforms the compared baseline algorithms in effectiveness and robustness. Note to Practitioners— To improve utilization of airport facilities (e.g., aircraft stands, taxiways and runways) and reduce aircraft delays, we propose a particle swarm optimization and Q-learning based aircraft routing and scheduling algorithm for a hybrid problem involving the airport ground movement problem and the aircraft sequencing problem. The outer layer of PSO-QL-ARS inherits the main framework of the PSO to search for the best aircraft routing and scheduling solutions; meanwhile a Q-table is designed for evolving particles. For each particle, the inner layer of PSO-QL-ARS employs semi-no-wait timetabling methods to generate the feasible trajectory plans satisfying distance constraint, queuing constraint, and release time difference constraint for all aircraft. Numerical experiments in both real and simulated instances demonstrate the effectiveness of the proposal in the application of Air Traffic Flow Management (ATFM). PSO-QL-ARS can achieve an improvement of at most 9.01% in reducing the taxiing time of aircraft. The framework of the proposal can be further extended to problems involving both the path planning and task scheduling, e.g., order dispatch in logistics.}, archive = {J_TASE}, author = {Jie Zhu and Guangke Han and Peishan Shang and Zheng Zhao and Haiping Huang and Fu Xiao}, doi = {10.1109/TASE.2025.3620472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22609--22624}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A Q-learning-based particle swarm optimization for aircraft routing and scheduling in airport terminal area}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant H-infinity stabilization for networked cascade control systems with novel adaptive event-triggered mechanism. TASE, 22, 22597--22608. (www)

  @article{ , abstract = {Networked cascade control systems (NCCS) are susceptible to performance degradation from network-induced delays, communication constraints, and component failures. To address these interconnected challenges, this paper develops an integrated fault-tolerant H-infinity control framework for discrete-time nonlinear NCCS. A key contribution is a novel, disturbance-aware adaptive event-triggered mechanism (AETM) that uniquely incorporates measurable external disturbances into its triggering logic. This design dynamically reduces communication frequency while preserving performance. Furthermore, a unified co-design of the controller parameters and the AETM’s weighting matrix is established, while stability is rigorously substantiated through theoretical analysis. Simulation results for a boiler control system validate the proposed method’s effectiveness, demonstrating significant improvements in both system robustness and communication efficiency. Note to Practitioners—This study proposes an adaptive event-triggered mechanism for networked cascade control systems, which holds broad applicability in industrial processes such as boiler power generation and chemical processing. Unlike conventional static approaches, the proposed AETM dynamically adjusts communication frequency in response to measurable external disturbances, thereby optimizing resource utilization while ensuring system stability under network-induced delays, bandwidth constraints, and equipment failures. For industrial applications, the AETM significantly reduces communication overhead—our simulation on a boiler system demonstrated a 49.02% reduction in bandwidth usage compared to traditional event-triggered mechanisms, representing a 28.77% improvement over existing adaptive methods. When integrated with fault-tolerant H-infinity control, the framework maintains reliable control performance during actuator failures or disturbance spikes, enhancing operational efficiency in resource-constrained industrial environments.}, archive = {J_TASE}, author = {Zhaoping Du and Chen Chen and Changjiang Li and Xiaofei Yang and Jianzhen Li}, doi = {10.1109/TASE.2025.3619932}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22597--22608}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant H-infinity stabilization for networked cascade control systems with novel adaptive event-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Intelligent optimization setting of froth velocity by integrating multi-source flotation operation information. TASE, 22, 22587--22596. (www)

  @article{ , abstract = {Froth velocity is one of the key control indices affecting the performance of froth flotation. In the flotation production process, the froth velocity is usually set based on manual experience, which is subjective and arbitrary, which is highly subjective. Moreover, the relationship between the flotation process characteristics and froth velocity is complex, making it difficult to optimize the froth velocity setting. Therefore, an intelligent optimization setting method for froth velocity that integrates multi-source flotation operation information is proposed. In this study, we first analyze the relationship between different operational conditions and the flotation speed, establishing a model for the flotation speed. Subsequently, we employ case-based reasoning (CBR), combined with a flotation speed evaluation and learning model, to obtain the set values for the flotation speed. Finally, an intelligent compensation prediction model based on extreme trees regression (ETR) is utilized to derive the final predicted flotation speed values. Throughout these steps, a flotation production index prediction model is applied to forecast various flotation production indices under the predicted flotation speed. Experiments utilizing historical process data from a flotation production experimental platform demonstrated that the froth velocity settings derived from this method resulted in reduced fluctuation in production indicators and a 12% increase in the comprehensive qualification rate of flotation process indicators. Additionally, the concentrate output rose by 13.9%. This approach significantly enhances flotation production conditions and key performance indicators, confirming its practical application value. Note to Practitioners—This paper primarily addresses the issue of subjectivity and uncertainty in manually setting the flotation velocity in froth flotation due to the lack of objective criteria. To tackle this, an intelligent optimization setting method for froth velocity is proposed. The aim is to leverage various operational data to intelligently predict and set the flotation velocity, thereby enhancing the efficiency of the flotation process. In this study, we propose an intelligent optimization setting method for froth velocity based on the analysis of the working principle of the flotation cell and the relationship between flotation velocity and key flotation indicators. We integrate the flotation production indicator prediction model, the evaluation and learning model that is based on the discrete particle swarm algorithm, and the intelligent compensation prediction model that is based on extreme trees regression. Preliminary experiments indicate that this method is theoretically feasible, but it has not yet been tested in the flotation production process. Future research will consider constraints of on-site conditions, such as the delay in setting the flotation velocity, for further prediction optimization.}, archive = {J_TASE}, author = {Ming Lu and Xianke He and Yingcong Li and Lei He and Ying Zou and Zunhui Yi and Pei Li}, doi = {10.1109/TASE.2025.3619416}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22587--22596}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intelligent optimization setting of froth velocity by integrating multi-source flotation operation information}, volume = {22}, year = {2025}, }
• (2025). Multi-view stereo with geometric encoding for large-scale dense scene reconstruction. TASE, 22, 22568--22586. (www)

  @article{ , abstract = {Multi-view stereo (MVS) implicitly encodes photometric and geometric cues into the cost volume for multi-view correspondence matching, transferring insufficient geometric cues essential to depth estimation and reconstruction. This paper proposes GE-MVS, a novel multi-view stereo network with geometric encoding for more accurate and complete depth estimation and point cloud reconstruction. First, the cross-view adaptive cost volume aggregation module is proposed to strengthen the encoding of multi-view geometric cues during cost volume construction. Then, the depth consistency optimization is performed in 3D point space during learning by invoking ground-truth depth cues from adjacent views. Finally, the surface normal geometries are explicitly encoded to refine the sampled depth hypotheses to be consistent in the local neighbor regions. Extensive experiments on the standard MVS benchmarks including DTU, Tanks and Temples, and BlendedMVS demonstrate the state-of-the-art depth estimation and point cloud reconstruction performance of GE-MVS. The GE-MVS is further deployed in real-world experiments for UAV-based large-scale reconstruction, where our method outperforms the prevalent industrial reconstruction solutions in terms of reconstruction efficiency and effectiveness. Supplementary video can be found at https://youtu.be/Z4tGROatVjU Note to Practitioners—Multi-view stereo (MVS) enables dense point cloud reconstruction of target scenes from calibrated multi-view images and has been widely adopted in robotic navigation, exploration, and manipulation. Recently, learning-based MVS methods have significantly improved reconstruction accuracy and completeness compared to traditional approaches. This work aims to enhance geometric modeling during network learning by utilizing ground-truth depth cues from adjacent views and encoding surface normal geometries. Extensive experiments conducted on both datasets and real-world scenarios validate the effectiveness, scalability, and efficiency of the proposed method. The proposed method can provide accurate depth and dense point cloud representations for applications such as aerial path planning, robotic manipulation, autonomous driving, and virtual and augmented reality. Future work will focus on adapting the proposed method from terrestrial to underwater domains for real-world underwater dense scene reconstruction.}, archive = {J_TASE}, author = {Guidong Yang and Rui Cao and Junjie Wen and Benyun Zhao and Qingxiang Li and Xi Chen and Yun-Hui Liu and Ben M. Chen}, doi = {10.1109/TASE.2025.3619093}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22568--22586}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-view stereo with geometric encoding for large-scale dense scene reconstruction}, volume = {22}, year = {2025}, }
• (2025). Causal forest-guided partitioning for robust operation optimization: A case study on draw ratios allocation. TASE, 22, 22553--22567. (www)

  @article{ , abstract = {The draw ratios allocation critically determines the final performance of carbon fiber. However, the stochastic nature in component fluxes and concentrations variations introduces uncertainty into draw ratios, causing frequent parameter fluctuations. Therefore, we propose a robust operation optimization framework that formulates a biobjective model to minimize linear density and maximize strength. Within this framework, a causal forest-guided partitioning for robust operation optimization algorithm is developed to solve the model. Specifically, the algorithm uses perturbations on individual decision variables to form treatment groups from an initial control population, enabling causal forests to infer the overall causal effects of each decision variable on the objectives. Next, the computed mean conditional average treatment effects are clustered via K-means to partition the variables by their robustness relevance. Then, a tailored optimization strategy, augmented by an external archiving mechanism, is performed on each group to efficiently search for the robust Pareto optimal set. Finally, adaptive utopian point-based decision making is used to determine the optimal setpoint. The proposed framework is validated on benchmark problems and in simulation study, achieves a 29.93% reduction in linear density and 99.41% increase in strength, thereby confirming its practical applicability. Note to Practitioners—This study aims to develop an optimal draw ratios allocation framework in a dry-jet wet spinning for polyacrylonitrile-based carbon fiber under uncertainty. To this end, we propose a robust operation optimization framework comprising two modules: operation optimization modeling for draw ratios allocation and causal forest-guided partitioning for robust operation optimization. In the modeling phase, practitioners can construct relevant models through data-driven or mechanistic approaches. Then, the causal forest-guided partitioning of decision variables based on causal forests infers the overall causal effects of each variable on the objectives under perturbations, and the K-means clustering is subsequently used to achieve adaptive variable partitioning. Furthermore, practitioners must consider both data requirements and computational resources. Because robust operation optimization depends on accurate models, practitioners at small- to medium-scale facilities should compile a dataset of at least several thousand validated production records. Optimization problems involving dozens of decision variables can typically be handled on a standard industrial computer. The proposed framework has been validated on simulation data for draw ratios allocation. Moreover, it is directly applicable to multistage drawing setups and can be readily extended to wet spinning or other mainstream carbon fiber manufacturing.}, archive = {J_TASE}, author = {Quan Liu and Kunlun Li and Yilin Fang and Liang Jin and Zude Zhou}, doi = {10.1109/TASE.2025.3618923}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22553--22567}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Causal forest-guided partitioning for robust operation optimization: A case study on draw ratios allocation}, volume = {22}, year = {2025}, }
• (2025). Incremental joint learning of depth, pose, and implicit scene representation on monocular camera in large-scale scenes. TASE, 22, 22541--22552. (www)

  @article{ , abstract = {Dense scene reconstruction for photo-realistic view synthesis has various applications, such as VR/AR, and robotics navigation. Existing dense reconstruction methods are primarily designed for small room scenarios, but in practice, the scenes encountered by robots are typically large-scale environments. Most existing methods have difficulties in large-scale scenes due to three core challenges: (a) inaccurate depth input. Depth information is crucial for both scene geometry reconstruction and pose estimation. Accurate depth input is impossible to get in real-world large-scale scenes. (b) inaccurate pose estimation. Existing methods are not robust enough with the growth of cumulative errors in large scenes and long sequences. (c) insufficient scene representation capability. A single global radiance field lacks the capacity to scale effectively to large-scale scenes. To this end, we propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale dense scene reconstruction. For depth estimation, a vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. For pose estimation, a feature-metric bundle adjustment (FBA) method is designed for accurate and robust camera tracking in large-scale scenes and eliminates pose drift. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields to enhance the scalability of 3D scene representation. In local radiance fields, we propose a tri-plane based scene representation method to further improve the accuracy and efficiency of scene reconstruction. We conduct extensive experiments on various datasets, including our own collected data, to demonstrate the effectiveness and accuracy of our method in depth estimation, pose estimation, and large-scale scene reconstruction. The code has been open-sourced on https://github.com/dtc111111/incre-dpsr Note to Practitioners—In practical robot deployment scenarios, dense scene reconstruction can play a crucial role, such as in precise localization and obstacle avoidance. However, existing scene reconstruction methods are mainly suitable for small-room environments and struggle with larger-scale settings like long corridors or extended sequences. Our method introduces an incremental joint learning framework that simultaneously performs depth estimation, pose estimation, and scene reconstruction. Experimental results demonstrate the accuracy and effectiveness of this approach across different scene types.}, archive = {J_TASE}, author = {Tianchen Deng and Nailin Wang and Chongdi Wang and Shenghai Yuan and Jingchuan Wang and Hesheng Wang and Danwei Wang and Weidong Chen}, doi = {10.1109/TASE.2025.3617654}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22541--22552}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Incremental joint learning of depth, pose, and implicit scene representation on monocular camera in large-scale scenes}, volume = {22}, year = {2025}, }
• (2025). Multi-modal LLMs in agriculture: A comprehensive review. TASE, 22, 22510--22540. (www)

  @article{ , abstract = {Given the rapid emergence and applications of Multi-Modal Large Language Models (MM-LLMs) across various scientific fields, insights regarding their applicability in agriculture are still only partially explored. This paper conducts an in-depth review of MM-LLMs in agriculture, focusing on understanding how MM-LLMs can be developed and implemented to optimize agricultural processes, increase efficiency, and reduce costs. Recent studies have explored the capabilities of MM-LLMs in agricultural information processing and decision-making. Despite these advancements, significant gaps persist, particularly in addressing domain-specific challenges such as variable data quality and availability, integration with existing agricultural systems, and the creation of robust training datasets that accurately represent complex agricultural environments. Moreover, a comprehensive understanding of the capabilities, challenges, and limitations of MM-LLMs in agricultural information processing and application is still missing. Exploring these areas is crucial to providing the community with a broader perspective and a clearer understanding of MM-LLMs’ applications, establishing a benchmark for the current state and emerging trends in this field. To bridge this gap, this survey reviews the progress of MM-LLMs and their utilization in agriculture, with an additional focus on 11 key research questions (RQs), where 4 RQs are general and 7 RQs are agriculture focused. By addressing these RQs, this review outlines the current opportunities and challenges, limitations, and future roadmap for MM-LLMs in agriculture. The findings indicate that multi-modal MM-LLMs not only simplify complex agricultural challenges but also significantly enhance decision-making and improve the efficiency of agricultural image processing. These advancements position MM-LLMs as an essential tool for the future of farming. For continued research and understanding, an organized and regularly updated list of papers on MM-LLMs is available at https://github.com/JiajiaLi04/Multi-Modal-LLMs-in-Agriculture Note to Practitioners—Motivated by the need to optimize agricultural practices, this paper investigates the use of Large Language Models (MM-LLMs) to improve efficiency and decision-making in agriculture. We delve into critical RQs to reveal the capabilities and challenges of MM-LLMs, and their potential applications in the agricultural sector. Looking ahead, our findings suggest a promising future for the integration of MM-LLMs in agriculture, potentially revolutionizing how we manage and operate farms.}, archive = {J_TASE}, author = {Ranjan Sapkota and Rizwan Qureshi and Muhammad Usman Hadi and Syed Zohaib Hassan and Ferhat Sadak and Maged Shoman and Muhammad Sajjad and Fayaz Ali Dharejo and Achyut Paudel and Jiajia Li and Zhichao Meng and John Shutske and Manoj Karkee}, doi = {10.1109/TASE.2025.3612154}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22510--22540}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-modal LLMs in agriculture: A comprehensive review}, volume = {22}, year = {2025}, }
• (2025). Tracking control of nonlinear switched multi-agent systems via model-dependent edge-node-combined dynamic event-triggered strategies. TASE, 22, 22497--22509. (www)

  @article{ , abstract = {This research investigates the problem of exponential tracking consensus for nonlinear switched multi-agent systems (NSMASs) under limited communication resources. To alleviate the communication load, edge-event-triggered strategies are applied to switched multi-agent systems for the first time. Due to the nonlinear switched dynamics in NSMASs, addressing the edge-event-triggered tracking consensus problem is more challenging compared to systems with linear or nonlinear fixed dynamics explored in previous studies. To tackle this challenge, the novel dynamic event-triggered rule for the leader and novel dynamic edge-event-triggered rules for edges are developed, both relying on model-dependent designs. Different triggering conditions are established for diverse system models to achieve enhanced triggering performance. Furthermore, the introduction of model-dependent dynamic auxiliary and edge-auxiliary variables contributes to a significant reduction in signal transmission frequency. Then, novel model-dependent edge-node-combined distributed dynamic event-triggered controllers are formed to ensure the global exponential consensus of NSMASs, with Zeno phenomenon explicitly excluded. Finally, the effectiveness, practicability and superiority of proposed methods are validated through two numerical examples and a practical physical simulation. Note to Practitioners—In practical applications, numerous engineering problems can be abstracted as NSMASs, such as the switching of loading robots between different operating conditions or amphibious robots transitioning between water and land modes. Under constrained communication resources, continuous information exchange among agents would result in significant resource wastage. To minimize energy consumption, improve system efficiency, and extend operational time under limited energy resources, this study, for the first time, introduces edge-based event-triggered strategies into switched multi-agent systems. The proposed novel model-dependent edge-node-combined triggering control approach not only ensures tracking consensus but also significantly reduces triggering frequency and enhances resource utilization, demonstrating substantial potential for practical applications.}, archive = {J_TASE}, author = {Jinhai Liu and Shuo Zhang and Huaguang Zhang and Wei Wang and Zhigang Zhang}, doi = {10.1109/TASE.2025.3621540}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22497--22509}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tracking control of nonlinear switched multi-agent systems via model-dependent edge-node-combined dynamic event-triggered strategies}, volume = {22}, year = {2025}, }
• (2025). Finite time prescribed performance adaptive consensus control for nonlinear multi-agent systems and its application. TASE, 22, 22484--22496. (www)

  @article{ , abstract = {This study designs an adaptive consensus output feedback quantized controller for high-order nonlinear multi-agent systems with full-state and tracking error constraints. The key features of this control strategy include: 1) A finite-time prescribed performance barrier function with online parameter adjustment, combined with dynamic surface technology, which simplifies the controller design and ensures all agent states remain within predefined limits; 2) An error threshold function is used to meet tracking error constraints, improving response speed and maintaining constraint compliance; 3) The use of quantizers in the system reduces communication load during signal transmission. Stability analysis shows that all signals in the closed-loop system remain consistently and semi-globally uniformly bounded. Additionally, the effectiveness of the control strategy is validated through experiments on a multi-agent cooperative control platform. Note to Practitioners—Modern engineering systems typically exhibit complex and ever-changing characteristics, particularly in high-order nonlinear multi-agent systems, whose control involves not only the entire system state but also strict adherence to tracking error constraints. In practical applications, to ensure that the state of each agent consistently remains within predefined safety and performance bounds, it is essential to consider dynamic performance requirements and error threshold limitations within a finite time. Moreover, the adoption of quantization techniques effectively reduces the communication burden during signal transmission, thereby enhancing the system’s real-time performance and reliability. The adaptive consensus output feedback quantized controller proposed in this paper, which employs a finite-time prescribed performance barrier function and an online parameter adjustment mechanism in conjunction with dynamic surface control, achieves effective regulation of both the full state and tracking error constraints. Experimental results further demonstrate the practicality and reliability of the control strategy on a multi-agent collaborative control platform, thereby offering novel insights and a robust theoretical foundation for engineering practice.}, archive = {J_TASE}, author = {Guoqiang Zhu and Xuecheng Zhang and Haoqi Li and Xiuyu Zhang and Chun-Yi Su}, doi = {10.1109/TASE.2025.3621082}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22484--22496}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite time prescribed performance adaptive consensus control for nonlinear multi-agent systems and its application}, volume = {22}, year = {2025}, }
• (2025). Modeling and trajectory tracking control of continuum robot with magnetic spacer disks and soft drives. TASE, 22, 22468--22483. (www)

  @article{ , abstract = {This paper proposes efficient modeling and control methods for a continuum robot (CR) composed of a three-segment robot body (two proximal segments and a distal segment with an endpoint) and soft drives. Each segment incorporates multiple mutually exclusive magnetic spacer disks for a flexible structure, and the soft drives are designed based on pneumatic soft actuators (PSAs) to ensure operational safety. For both computational efficiency and tracking performance, an open-loop control method based on the kinematic model of the robot body is adopted to operate the two proximal segments, and a feedforward-feedback control method based on the dynamic model of the distal segment derived from the Euler-Lagrange method is proposed to control the endpoint trajectory. During the dynamic modeling process of the distal segment, the magnetic potential energy is considered additionally by simplifying the magnetic spacer disks to multiple magnetic dipoles. Moreover, the relationship between the input pressures of the soft drives and the generalized force of the dynamic model is derived from the three-element model and the principle of virtual power. Experiments demonstrate the effectiveness of the proposed modeling and control method. Note to Practitioners—Continuum robots (CRs) are highly promising for use in complex environments due to their excellent deformability. For practical applications of CRs, the modeling and endpoint trajectory tracking control considering efficiency and tracking performance are urgent problems. This paper focuses on a CR including a three-segment robot body and soft drives, and proposes an efficient control method that combines kinematic and dynamic models of the investigated CR. Experiments confirm the effectiveness of the method, demonstrating the potential for practical applications in areas such as detection and medical treatments.}, archive = {J_TASE}, author = {Shiying Zhao and Qingxin Meng and Xuzhi Lai and Jinseok Woo and Jinhua She and Edwardo F. Fukushima and Min Wu}, doi = {10.1109/TASE.2025.3620829}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22468--22483}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling and trajectory tracking control of continuum robot with magnetic spacer disks and soft drives}, volume = {22}, year = {2025}, }
• (2025). Online reduced-order data-enabled predictive control. TASE, 22, 22455--22467. (www)

  @article{ , abstract = {Data-enabled predictive control (DeePC) has garnered significant attention for its ability to achieve safe, data-driven optimal control without relying on explicit parametric models. Traditional DeePC methods use pre-collected input/output (I/O) data to construct a Hankel matrix offline and then formulate a predictive control framework online for linear, weakly nonlinear, and weakly stochastic systems. However, in systems with evolving dynamics, incorporating real-time data into the DeePC framework becomes crucial to enhance control performance. This paper proposes an online DeePC framework designed for strongly nonlinear and/or time-varying systems (i.e., systems with evolving dynamics), enabling the algorithm to update the Hankel matrix online by adding real-time informative signals. By exploiting the minimum non-zero singular value of the Hankel matrix, the developed online DeePC selectively integrates informative data and effectively captures evolving system dynamics. Additionally, a numerical singular value decomposition technique is introduced to reduce the computational complexity for updating a reduced-order Hankel matrix. Simulation results on three cases, linear time-varying system, vehicle anti-rollover control, and Li-ion battery fast charging, demonstrate the effectiveness of the proposed online reduced-order DeePC framework. Note to Practitioners—With the increasing complexity of modern control systems and the expanding availability of data, there is a growing preference for data-driven control schemes. Different from model-based control schemes that require precise system modeling, data-driven controllers generate control policies by using collected input/output (I/O) data. Data-driven control can be categorized into two paradigms: indirect data-driven control, which includes system identification and model-based control design, and direct data-driven control, which bypasses the system identification process and directly designs control strategies based on I/O data. The direct data-driven control offers greater flexibility and potential by avoiding a specific parametric model in control design. However, for time-varying systems, offline collected I/O data is not reliable, and incorporating real-time data into the control framework becomes crucial.}, archive = {J_TASE}, author = {Amin Vahidi-Moghaddam and Kaixiang Zhang and Xunyuan Yin and Vaibhav Srivastava and Zhaojian Li}, doi = {10.1109/TASE.2025.3619434}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22455--22467}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online reduced-order data-enabled predictive control}, volume = {22}, year = {2025}, }
• (2025). Appointed-time prescribed performance control for shipborne SideArm with DRL-based kinematic compensation. TASE, 22, 22438--22454. (www)

  @article{ , abstract = {During the recovery of fixed-wing autonomous aerial vehicles (AAVs) by a SideArm, the motion of the desired docking point (DDP) on the arresting cable located at the SideArm’s tip is heavily influenced by the ship’s movements. The movements increase the complexity and risk of the docking procedure. As a means of stabilizing the spatial position of the DDP, this paper presents a novel control framework that integrates kinematic solutions and dynamic control strategies. The proposed framework first compensates for the ship’s six-degrees-of-freedom motion, then generates the basic joint signals of the SideArm using a geometry-based inverse kinematic (IK) method. The twin delayed deep deterministic policy gradient (TD3) algorithm is applied to compensate for these signals, forming the IK-TD3 approach that enhances the accuracy of the DDP’s spatial position. An appointed-time prescribed performance control method is designed to ensure that the system’s performance remains within predefined constraints and converges within a preassigned time. A fixed-time disturbance observer is introduced to estimate the lumped disturbances and uncertainty dynamics of the SideArm. A comprehensive set of simulations and experiment considering different sea states demonstrates the reliability of the proposed approach. Note to Practitioners—This paper is motivated by the problem of recovering AAVs in unpredictable sea conditions using a SideArm system located on a ship’s deck. From a deployment perspective, a fixed-size SideArm mitigates the need to resize equipment based on wave height, but simultaneously increases the difficulty of finding accurate and reliable IK solutions, especially in the case of large ${}^{\mathrm { }}$ amplitude deck motion. Traditional geometric-based IK methods are computationally efficient, but struggle to maintain sufficient accuracy under these conditions. To overcome this limitation, we propose an IK framework that integrates the TD3 algorithm, thus enhancing the SideArm’s ability to compensate for varying ship motions. We improve the controller design using prescribed performance control, thereby ensuring both transient and steady-state performance without requiring an explicit residual set calculation. Numerical simulations and experiments under various sea states demonstrate that the proposed approach achieves robust, high-accuracy joint angle commands and improved system stability, making it a practical and reliable solution for AAV recovery operations in maritime environments.}, archive = {J_TASE}, author = {Zikang Su and Jialiang Fan and Xingling Shao and Chuntao Li and Xinru Wang}, doi = {10.1109/TASE.2025.3618920}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22438--22454}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Appointed-time prescribed performance control for shipborne SideArm with DRL-based kinematic compensation}, volume = {22}, year = {2025}, }
• (2025). FeRF-BEVIO: A feature radiance field-based bidirectionally enhanced visual-inertial odometry. TASE, 22, 22423--22437. (www)

  @article{ , abstract = {To address the issues of inaccurate feature association and insufficient utilization of the complementarity between the visual and inertial information in visual-inertial navigation systems, this paper proposes a feature radiance field (FeRF)-based bidirectionally enhanced visual-inertial odometry (FeRF-BEVIO). FeRF is a novel method for describing environment features, incorporating both the grayscale and position of point and line features. It provides a unified framework for representing and storing features, regardless of whether their depth has converged. Then, a bidirectional visual-inertial enhancement method is designed based on FeRFs. Specifically, on one hand, the information of the inertial measurement unit (IMU) is utilized to aid visual feature association within FeRFs. On the other hand, the visual information is employed to refine the IMU parameter estimation. This bidirectional enhancement process is iterated to improve the integration of the visual and inertial data. At last, the system jointly optimizes the robot poses and point-line feature parameters within a sliding window and updates FeRFs accordingly. Comparative experiments on public datasets and in the real-world environments demonstrate that FeRF-BEVIO outperforms state-of-the-art visual-inertial odometry (VIO) systems in both accuracy and robustness. Therefore, FeRF-BEVIO is highly suitable for navigation and simultaneous localization and mapping (SLAM) of micro aerial vehicles (MAVs). Note to Practitioners—This paper is motivated by the robust localization and mapping for MAVs. We propose a novel visual-inertial odometry system FeRF-BEVIO. FeRF is a novel environmental representation that encodes both point and line features in a compact and unified framework. In addition, we design a bidirectional enhancement mechanism, where visual features help optimize inertial parameters, and the resultant inertial information in turn improves visual feature association. FeRF-BEVIO maintains high accuracy under fast motion and low-texture environments. The experiments are carried out on MAVs in various indoor and outdoor scenes, and the results demonstrate the robustness and effectiveness of the proposed framework. FeRF-BEVIO shows great potential for MAV autonomous navigation and AR/VR applications.}, archive = {J_TASE}, author = {Yuanxi Gao and Jing Yuan and Shizhuo Yu and Xuebo Zhang}, doi = {10.1109/TASE.2025.3618843}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22423--22437}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FeRF-BEVIO: A feature radiance field-based bidirectionally enhanced visual-inertial odometry}, volume = {22}, year = {2025}, }
• (2025). Closed-loop parameter optimization for robotic machining using physics-informed machine learning and multiobjective optimization. TASE, 22, 22410--22422. (www)

  @article{ , abstract = {In practical applications, the simultaneous optimization of numerous design parameters in time-consuming multi-objective optimization experiments is recognized as a significant bottleneck across various scientific and engineering disciplines. A prominent example is the optimization of machining parameters for achieving efficient and precise robotic belt grinding (RBG). This paper presents a closed-loop machining parameter optimization approach, which comprises two key stages: forward multi-task prediction and backward multi-objective parameter optimization. In the first stage, a physics-informed neural network (PINN) method is introduced, which integrates the multi-gate mixture-of-experts multi-task learning method with an RBG mechanism model to simultaneously predict material removal depth and averaged surface roughness. In the second stage, a powered multi-objective particle swarm optimization (MOPSO) method is developed, which combines a standard MOPSO method with a non-linear Powerball technique, to efficiently optimize the RBG machining parameters with a limited number of training iterations based on the learned PINN model. Two optimal machining parameter solutions are generated and recommended for the RBG machining process. The effectiveness and superiority of the proposed closed-loop parameter optimization method are validated through comparative experiments, which demonstrate its advantages in both coprediction accuracy and optimization efficiency. Note to Practitioners—This paper addresses the challenge of identifying robotic machining parameters that effectively balance machining efficiency and surface quality. Traditional trial-and-error methods for adjusting these parameters are both time-intensive and costly, given the vast number of possible combinations. To overcome these limitations, this paper proposes an intelligent optimization approach that leverages historical machining data to automatically determine optimal machining parameters. Our approach integrates artificial intelligence techniques with robotic machining mechanisms, ultimately recommending two sets of parameters for robotic machining. Preliminary experiments demonstrate the feasibility of this approach, but it has not yet been incorporated into a robotic machining system or tested in production. Future research will focus on dynamically optimizing robotic machining parameters by combining dynamic time-series signals with static machining parameters.}, archive = {J_TASE}, author = {Guijun Ma and Zidong Wang and Weibo Liu and Zeyuan Yang and Desheng Huang and Han Ding}, doi = {10.1109/TASE.2025.3618242}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22410--22422}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Closed-loop parameter optimization for robotic machining using physics-informed machine learning and multiobjective optimization}, volume = {22}, year = {2025}, }
• (2025). Path coordination for robust, fast, and scalable multi-agent path finding under unforeseen delays. TASE, 22, 22396--22409. (www)

  @article{ , abstract = {Multi-agent path finding (MAPF), i.e., finding collision-free paths for multiple agents on graphs, is widely applied to domains such as warehouse automation, robotic manufacturing, and autonomous vehicle coordination. One major challenge that hinders MAPF algorithm deployments is that, agents often suffer from unforeseen delays without prior knowledge during path execution because of exogenous disturbances (e.g., blocked by humans, uncontrollable or third-party vehicles in the workspace). Existing approaches can hardly achieve robustness, computational efficiency and scalability simultaneously. To address this issue, we propose a novel approach based on the classical decoupled path coordination, which firstly plans paths for each agent in space domain, and then schedules paths in time-space domain to avoid collisions or deadlocks. To guarantee liveness and improve computational efficiency, it incorporates deadlock alleviation, prevention, and repair mechanisms utilizing several graph structures. Theoretically, necessary conditions are provided and formally proved for the safety and liveness in the presence of unforeseen delays. Empirically, the proposed approach is shown to be robust, fast and scalable. It also significantly outperforms state-of-the-art MAPF robust execution approaches on success rate, runtime, and solution quality. Note to Practitioners—Robust and real-time multi-agent path finding for large-scale multi-robot systems in logistics, order fulfillment systems, and manufacturing remains a major challenge. Since its inception, path coordination has proven to be a time-tested approach in real plants for decades. On one side, it is decoupled and easy to be implemented; on the other side, it is incomplete and suffers from a major drawback: deadlocks are unavoidable and should be tackled during coordination. Its drawback is even worsened when path coordination is deployed in large-scale scenarios with hundreds of robots. Experienced with years’ cooperation with one of the largest 3C manufacturing companies in China, we propose a novel path coordination approach that can achieve robustness, computational efficiency and scalability simultaneously. It alleviates, prevent or repair deadlocks using several existing or newly introduced graph structures. To deploy the proposed method in real plants, practitioners should: 1) adapt the proposed approach to roadmap layouts; 2) consider concrete robot/vehicle/map constraints and costs; 3) add geometric collision detection module; and 4) use multi-thread computation to further improve the runtime performance.}, archive = {J_TASE}, author = {Weibin Ye and Hanfu Wang and Weidong Chen}, doi = {10.1109/TASE.2025.3615989}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22396--22409}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Path coordination for robust, fast, and scalable multi-agent path finding under unforeseen delays}, volume = {22}, year = {2025}, }
• (2025). A disturbance rejection scheme for ASR heading control based on an improved extended state observer and experiment research. TASE, 22, 22384--22395. (www)

  @article{ , abstract = {This paper addresses the problem of stable heading control for Autonomous Surface Robots (ASRs) under unknown external time-varying disturbances. An Improved Extended State Observer (IESO) is proposed, along with an adaptive law for adjusting the disturbance observation frequency based on the observation error. In this way, unknown time-varying disturbances can be estimated, and robust disturbance rejection for heading control is ensured. Meanwhile, inspired by the concept of Model-Free Adaptive Control, an Improved Non-Linear Second-order Error Feedback heading controller (INLSEF) is designed. This approach reduces the quantity of control parameters when maintaining excellent control performance. The stability of the control system is analyzed using singular perturbation theory. Numerical simulation and field experiments are conducted using a small ASR “Yu Lang III” as the study object. The results indicate that, once control stability is achieved, the proposed method leads to a significant reduction in heading error, highlighting its effectiveness and practical value. Note to Practitioners—In the field of industrial automation, maintaining the stability of equipment operations is of utmost importance. This paper proposes a novel anti-disturbance control scheme for the attitude control of Autonomous Surface Robots (ASRs), which effectively observes and compensates for unknown disturbances from both internal and external environments without relying on the robot’s detailed control model. By accurately observing disturbances and compensating for them, this scheme reduces oscillations in the actuators, significantly improving the robot’s operational efficiency and stability. Extensive simulations and field experiments were conducted to validate this method, and the results demonstrate that the scheme substantially enhances the robot’s performance stability in complex environments. For industrial practitioners, it is highly recommended to integrate this control strategy during the initial design phase to optimize the robot’s overall disturbance resistance and operational precision.}, archive = {J_TASE}, author = {Tianyu Wei and Yulei Liao and Lei Wan and Chenguang Yang and Tuosheng Zhang and Ming Zhang}, doi = {10.1109/TASE.2025.3614016}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22384--22395}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A disturbance rejection scheme for ASR heading control based on an improved extended state observer and experiment research}, volume = {22}, year = {2025}, }
• (2025). Interpretable multi-agent reinforcement learning for traffic signal control: Influence mechanism and piecewise linear approximation. TASE, 22, 22371--22383. (www)

  @article{ , abstract = {Traffic signal control plays a crucial role in intelligent transportation systems, with cooperative control being challenging to implement but essential for its effectiveness. Many methods model multi-intersection traffic networks as grids and address the problem using multi-agent reinforcement learning (RL). Despite these existing studies, there is an opportunity to further enhance our understanding of the connectivity and globality of the traffic networks by capturing the spatiotemporal traffic information with efficient neural networks in deep RL. In this paper, we propose a novel multi-agent actor-critic framework based on an interpretable influence mechanism with a centralized learning and decentralized execution method. Specifically, we first construct an actor-critic framework, for which the piecewise linear neural network (PWLNN), named biased ReLU (BReLU), is used as the function approximator to obtain a more accurate and theoretically grounded approximation, and exhibits interpretability. Then, to model the relationships among agents in multi-intersection scenarios, we introduce an interpretable influence mechanism based on efficient hinging hyperplanes neural network (EHHNN), which derives weights by analysis of variance (ANOVA) decomposition among agents and extracts spatiotemporal dependencies of the traffic features. Finally, our proposed framework is validated on two synthetic traffic networks and a real road network to coordinate signal control between intersections, achieving lower traffic delays across the entire traffic network compared with benchmark performance. Note to Practitioners—The motivation of this paper is to develop an interpretable multi-intersection control framework that coordinates traffic signals across urban networks to mitigate congestion. When congestion occurs on the global traffic network, the strategy can collaborate with signal controllers at different intersections to minimize the vehicle waiting time on the road network and improve traffic flow. The strategy introduces an interpretable influence mechanism and an efficient training paradigm to better understand how decisions at one intersection affect traffic flow at upstream and downstream intersections, and thereby improve the effectiveness of the control outcomes. The developed strategy has been validated through co-simulation using Python and the Simulation of Urban Mobility (SUMO) traffic simulator in two synthetic traffic networks and a real road network to show its effectiveness.}, archive = {J_TASE}, author = {Zhiyue Luo and Hongyi Li and Jun Xu and Fanglin Chen}, doi = {10.1109/TASE.2025.3618243}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22371--22383}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Interpretable multi-agent reinforcement learning for traffic signal control: Influence mechanism and piecewise linear approximation}, volume = {22}, year = {2025}, }
• (2025). Exoskeleton-assisted stance and kneeling balance and work task evaluation in construction. TASE, 22, 22356--22370. (www)

  @article{ , abstract = {Construction workers experience serious safety and health risks in hazardous working environments. Quiet stance and kneeling are among the most common postures performed by construction workers during their daily work. This paper first analyzes lower-limb joint influence on neural balance control strategies using the frequency behavior of the intersection point of ground reaction forces. To evaluate the impact of elevation and wearable knee exoskeletons on postural balance and welding task performance, we design virtual- and mixed-reality (VR/MR) to simulate elevated environments and welding tasks. A linear quadratic regulator-controlled triple- and double-link inverted pendulum model is used for balance strategy quantification in stance and kneeling, respectively. Extensive multi-subject experiments are conducted to evaluate the usability of wearable exoskeletons in destabilizing construction environments. The quantified balance strategies capture the significance of the knee joint during balance control of quiet stance and kneeling gaits. Results show that center of pressure sway area is reduced up to 62% in quiet stance and 39% in kneeling gait for subjects tested in high-elevation worksites with knee exoskeleton assistance. The balance and multitask evaluation confirm and provide guidance on exoskeleton design to mitigate the fall risk in construction. Note to Practitioners—Construction workers commonly perform tasks that require prolonged quiet stance or kneeling gaits on high elevations. Worker balance can be undermined by chronic knee injuries, musculoskeletal disorders, and destabilizing visual perturbations caused by occupational activities. Wearable knee exoskeletons have evolved as promising interventions to reduce knee joint stress across a variety of work gaits in construction. Emerging technologies such as virtual- and mixed-reality (VR/MR) provide an enabling tool to study underlying balance strategies to complete tasks in dynamic environments. The VR/MR- generated immersive elevated welding environment are leveraged to examine the effects of threatening visual stimuli, wearable exoskeletons, and construction tasks on worker balance and skill performance. Intersection point height frequency analysis is used to quantify the neural balance strategy during various testing scenarios. We particularly explore the often-neglected role of the knee joint to facilitate research on knee exoskeleton assisted balance in stance and kneeling gaits. The experimental results provide insight into the efficacy of knee exoskeletons in improving worker’s stability and task performance in construction. The results highlight the need for a holistic approach to exoskeleton design to ensure that developed solutions can be safely and successfully integrated into the workplace environment.}, archive = {J_TASE}, author = {Gayatri Sreenivasan and Chunchu Zhu and Jingang Yi}, doi = {10.1109/TASE.2025.3538532}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22356--22370}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exoskeleton-assisted stance and kneeling balance and work task evaluation in construction}, volume = {22}, year = {2025}, }
• (2025). Whole-body stabilization of a cable-suspended multirotor platform carrying a slung load. TASE, 22, 22337--22355. (www)

  @article{ , abstract = {Suspended multirotor platforms are fascinating systems that can be employed in construction applications to provide safe transportation of heavy loads. Such a system comprising a cable-suspended platform with attached load features seven degrees of freedom (DoF) motion for the whole system. In this paper, we propose a composite whole-body control framework for the stabilization of the suspended multirotor platform system, leveraging singular perturbation theory to exploit the inherent three time-scale dynamics of the system. The control strategy computes the underactuated 3-DoF wrench space generated by the platform’s actuation units for the stabilization of the complete system. Building upon this, we develop a superposition-based shared control approach and then compare the two controllers. Moreover, to address specific cases where the time-scale separation between two dynamics of the triple-spherical pendulum becomes negligible, we design an operational space controller. The control approaches are validated using both extensive numerical simulations and experiments in different scenarios. We also carried out numerical robustness and stability analysis of the whole system. Note that our system relies on only onboard sensors for state estimation, which makes it effective for real-life outdoor applications. Note to Practitioners—This paper is motivated by the transportation of heavy loads using overhead cranes in construction sites. Due to the safety criticality of such applications, it has to be ensured that no oscillations of the attached load arise during the whole process. These oscillations can result from wind disturbances or due to the motion of the crane. In this work, we ensure that the proposed control approaches stabilize the attached load by utilizing the actuation power of an aerial platform suspended from the crane, and thus ensure safety in the involved construction sites. The practical implications of this work are validated using extensive simulation studies and experimental tests in two different lab settings, while demonstrating the numerical stability and robustness of the proposed framework to tackle the different uncertainties and external disturbances that can arise in construction sites.}, archive = {J_TASE}, author = {Hemjyoti Das and Grazia Zambella and Christian Ott}, doi = {10.1109/TASE.2025.3588767}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22337--22355}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Whole-body stabilization of a cable-suspended multirotor platform carrying a slung load}, volume = {22}, year = {2025}, }
• (2025). Robotic inspection and data analytics to localize and visualize the structural defects of concrete infrastructure. TASE, 22, 22324--22336. (www)

  @article{ , abstract = {This paper presents an innovative robotic inspection system designed to enhance the detection and analysis of structural defects in concrete infrastructure. The proposed inspection system is comprised of three modules: a robotic data collection module, a visual inspection module, and a subsurface mapping module. The robotic data collection module features an omnidirectional robotic platform, designed to move sideways without spinning. It is equipped with Ground Penetrating Radar (GPR) and RGB-D cameras, facilitating systematic data collection across construction sites. The visual inspection module employs a learning-based method, InspectionNet++, to analyze the frames for surface defects such as cracks, spalls, and stains, providing high accuracy and metric measurements of the defects. The subsurface mapping module processes the GPR data to detect and visualize hidden defects, creating a comprehensive map that correlates these with visible surface anomalies. Field tests demonstrate the system’s ability to automate construction structural inspection with improved efficiency and precision. Additionally, the customized visualization software is introduced to enable intuitive and interactive exploration of the detected defects within a unified interface. By automating data collection and enhancing defect detection through learning algorithms, the system not only speeds up the inspection process but also increases the reliability of infrastructure evaluations, supporting more informed maintenance decisions. Note to Practitioners—This paper introduces a robotic solution for inspection and condition assessment of concrete infrastructure. The system uses an omnidirectional robot equipped with GPR and RGB-D cameras to automatically collect data across construction sites. By harnessing vision-based positioning technology, our system empowers the robot to scan the ground surface in free motion pattern. This eliminates the need for time-consuming grid line setup traditionally required for manual GPR data collection. Our approach combines multi-sensor data analytics with advanced software, which enables the detection and visualization of both surface defects (cracks, spalls, stains) and subsurface anomalies. For practitioners, this automated approach offers several key benefits over manual inspections: increased inspection speed and coverage, rapid data collection enabled by robotic free motion, higher detection accuracy, quantitative defect measurements, and unified visualization correlating surface/subsurface conditions. This system has the potential to revolutionize infrastructure assessment practices. By facilitating more frequent and reliable inspections with minimal human intervention, it paves the way for proactive maintenance and ensures the sustainability of critical infrastructure. A current limitation is the relatively small training dataset for visual inspection, which may affect generalizability across diverse concrete structures. Future work aims to expand the dataset, improve irrelevant feature filtering, and utilize other NDE sensors (e.g., impact echo) for infrastructure inspection.}, archive = {J_TASE}, author = {Jinglun Feng and Bo Shang and Ejup Hoxha and César Hernández and Yang He and Weihan Wang and Jizhong Xiao}, doi = {10.1109/TASE.2025.3535227}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22324--22336}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic inspection and data analytics to localize and visualize the structural defects of concrete infrastructure}, volume = {22}, year = {2025}, }
• (2025). AI enabled automatic mobile robot intelligent navigation in construction with obstacle awareness. TASE, 22, 22308--22323. (www)

  @article{ , abstract = {In the construction industry, the integration of artificial intelligence (AI) and robotics has led to significant advancements in automating various tasks. One critical aspect is the intelligent navigation of automatic mobile robot (AMR) within construction sites, where dynamic environments pose challenges such as inaccurate robot state estimation, irregular and textureless obstacle detection. To solve these issues, this paper designs an AI enabled obstacle-aware AMR intelligent navigation approach while ensuring robot safety and efficiency. Specifically, the robot is equipped with the complementary RGB camera, inertial measurement unit (IMU) and wheel encoder to estimate the states (i.e., positions, orientations and velocities), which have been tightly fused in the optimization framework. Furthermore, the RGB-Depth images are jointly fed into the AI model. It combines of Mask-RCNN and multi-scale attention network, to detect and segment the obstacles, and estimate the corresponding relative depth. It should be noted that the system incorporates obstacle awareness mechanisms to dynamically switch the robot’s velocities in response to obstacles in the environment, ensuring smooth and collision-free movement. Experimental results demonstrate the effectiveness and robustness of the proposed AI-enabled navigation system in real-world construction scenarios, showcasing its potential to enhance productivity and safety in construction automation.Note to Practitioners —Motivated by the booming of AI and robotics, this paper aims to design an approach for AMR intelligent navigation with obstacle awareness in the scenarios of construction sites and buildings using the multimodal sensing units. The monocular camera, IMU and wheel encoder are jointly used to estimate the robot states in a tightly-coupled manner. Apart from that, we combine the advanced AI models, i.e., Mask-RCNN and multi-scale attention network, to discover the forward obstacles and calculate the corresponding depths. Eventually, the smooth control operation on vehicle velocity will be implemented while considering the robot running states and the obstacle relative depths, to ensure the robot safety on one hand and improve the productivity on the other hand, which is helpful in robot-assisted construction automation.}, archive = {J_TASE}, author = {Yinlong Zhang and Yuanhao Liu and Yunge Cui and Ziming Zeng and Wei Liang}, doi = {10.1109/TASE.2025.3558498}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22308--22323}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AI enabled automatic mobile robot intelligent navigation in construction with obstacle awareness}, volume = {22}, year = {2025}, }
• (2025). Attention-based pipe recognition network with a hybrid dataset. TASE, 22, 22296--22307. (www)

  @article{ , abstract = {A building information model for existing pipes is essential for maintenance tasks such as repairs, reconstruction, and modernization. However, current automatic recognition methods are not well-suited for handling complex piping systems on mobile devices. To address this, an attention mechanism-based recognition network is proposed. First, a decoder network is designed to minimize information loss during feature extraction. Its ability to integrate with various encoder networks allows for flexible network design tailored to different applications. Second, an extended convolution pyramid, tailored to the geometric characteristics of pipes, is introduced to enhance computational efficiency. The proposed network’s performance was evaluated through ablation studies and comparison experiments with state-of-the-art methods. To address the challenge of data scarcity, a hybrid dataset was developed using neural style transfer techniques, and an evaluation experiment demonstrated its effectiveness. Note to Practitioners—This paper presents an advanced recognition network aimed at improving the maintenance of complex piping systems in buildings, which often require frequent repairs, reconstruction, and modernization. Leveraging an attention mechanism, the network enables accurate identification of pipe components from images on mobile devices—a task that traditional methods struggle with due to system complexity. Key innovations include a decoder network that minimizes information loss during feature extraction and an extended convolution pyramid optimized for pipe shapes, enhancing computational efficiency. These advancements streamline maintenance tasks and provide a foundation for extending similar technologies to other complex industrial systems, offering substantial efficiency improvements across various sectors.}, archive = {J_TASE}, author = {Yang Tian and Yueh Feng Lin and Shugen Ma}, doi = {10.1109/TASE.2025.3567080}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22296--22307}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attention-based pipe recognition network with a hybrid dataset}, volume = {22}, year = {2025}, }
• (2025). Speak the same language: Global LiDAR registration on BIM using pose hough transform. TASE, 22, 22282--22295. (www)

  @article{ , abstract = {Light detection and ranging (LiDAR) point clouds and building information modeling (BIM) represent two distinct data modalities in the fields of robot perception and construction. These modalities originate from different sources and are associated with unique reference frames. The primary goal of this study is to align these modalities within a shared reference frame using a global registration approach, effectively enabling them to “speak the same language”. To achieve this, we propose a cross-modality registration method, spanning from the front end to the back end. At the front end, we extract triangle descriptors by identifying walls and intersected corners, enabling the matching of corner triplets with a complexity independent of the BIM’s size. For the back-end transformation estimation, we utilize the Hough transform to map the matched triplets to the transformation space and introduce a hierarchical voting mechanism to hypothesize multiple pose candidates. The final transformation is then verified using our designed occupancy-aware scoring method. To assess the effectiveness of our approach, we conducted real-world multi-session experiments in a large-scale university building, employing two different types of LiDAR sensors. We make the collected datasets and codes publicly available to benefit the community. Note to Practitioners—Our proposed registration method leverages walls and corners as shared features between LiDAR and BIM data, making it particularly well-suited for scenarios with well-defined structural layouts. Accumulating a larger LiDAR submap provides richer structural information, which further aids in achieving accurate alignment. To optimize computational efficiency, we recommend constructing the descriptor database offline and loading it during runtime, enabling a theoretical retrieval complexity of $O(1)$ . Despite its advantages, our approach has certain limitations. First, it primarily focuses on planar structures, which limits its effectiveness in utilizing non-planar features. Second, the method may underperform in cases where significant deviations exist between the as-designed BIM and as-is LiDAR data. Lastly, in ambiguous scenarios, such as long corridors or similar layouts within the same or across different floors, our method may struggle to verify the correct transformation among candidates. To address these challenges, incorporating additional information, particularly semantic cues such as floor numbers, room numbers, and room types, could enhance its robustness and reliability.}, archive = {J_TASE}, author = {Zhijian Qiao and Haoming Huang and Chuhao Liu and Zehuan Yu and Shaojie Shen and Fumin Zhang and Huan Yin}, doi = {10.1109/TASE.2025.3549176}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22282--22295}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Speak the same language: Global LiDAR registration on BIM using pose hough transform}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Special issue on automation and artificial intelligence (AI) in construction and building. TASE, 22, 22278--22281. (www)

  @article{ , archive = {J_TASE}, author = {Jingang Yi and Dikai Liu and Wei Yan and Vineet R. Kamat and Chao Wang and Jee-Hwan Ryu}, doi = {10.1109/TASE.2025.3614764}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22278--22281}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Special issue on automation and artificial intelligence (AI) in construction and building}, volume = {22}, year = {2025}, }
• (2025). Uniformity in full-state error prescribed performance control via error-driven flexibility for input-saturated systems with external disturbances. TASE, 22, 22269--22277. (www)

  @article{ , abstract = {This paper proposes a fuzzy control scheme that enforces a unified prescribed performance for full-state errors in input-saturated systems with external disturbances. The proposed scheme is characterized by three key innovations: First, a series of functional transformations are designed to guarantee multiple performance behaviors within a unified control framework, enabling desired behaviors through parameter selection without controller redesign. Second, a novel performance function for virtual errors completely eliminates strict initial value constraints, thereby removing offline verification computations and streamlining the design/implementation process. Third, an error-driven mechanism is developed to prevent singularities induced by input saturation and disturbances. Unlike existing flexible prescribed performance control methods that rely on a control input-driven mechanism, this mechanism offers two distinct advantages: it directly adjusts only a single boundary for a more straightforward adjustment, and operates without dependency on auxiliary system integration. This design eliminates adjustment delays while minimizing performance degradation caused by boundary relaxation. Simulations confirm the scheme’s efficacy and superiority. Note to Practitioners—This work addresses a critical challenge in industrial control systems: maintaining prescribed performance under actuator physical limits (input saturation) and external disturbances (e.g., load variations, noise). Traditional prescribed performance methods often degrade or fail under such combined conditions, leading to inefficient operation or safety risks. Our proposed error-driven flexible prescribed performance control offers a practical solution by dynamically adjusting performance boundaries only during saturation or disturbance events (activated when errors exceed predefined thresholds). The unified design framework supports four configurable performance boundary types (unconstrained, upper/lower-bounded, or fully constrained initial errors) through a single control architecture, eliminating the need for case-specific controller redesign. The marriage of unified constraints and error-driven mechanism provides saturation- and disturbance-aware performance adaptation: preserving strict boundaries normally while allowing temporary, adaptive relaxation during saturation or disturbance-induced deviations (e.g., robotic arm overload or sudden external torque), all within a unified control architecture.}, archive = {J_TASE}, author = {Xuexiu Liang and Yu Xia and Xiao Gong and Imre J. Rudas and Ramesh K. Agarwal}, doi = {10.1109/TASE.2025.3618693}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22269--22277}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Uniformity in full-state error prescribed performance control via error-driven flexibility for input-saturated systems with external disturbances}, volume = {22}, year = {2025}, }
• (2025). ZSPose: Instance-level zero-shot object pose estimation with segment anything model. TASE, 22, 22251--22268. (www)

  @article{ , abstract = {Estimating the poses of new objects is a challenging problem. Although many methods have been developed for instance-level object pose estimation, they often struggle when faced with new/unfamiliar objects. In this paper, we propose a zero-shot pose estimation method for new objects called ZSPose. We leverage SAM’s zero-shot feature to segment objects in cluttered environments and acquire masks for each object. To facilitate the matching of object masks with object models and categories, we propose a novel object matching strategy that aligns masks with the corresponding object models. Subsequently, based on the derived object masks, we produce object point clouds. Utilizing the RGB images of the objects alongside the point clouds, we present a feature-weight-based method for object pose estimation, achieving accurate pose estimation by predicting the matching weights between the model features and the point cloud features. We conduct performance testing on various instance-level object pose estimation datasets, and experimental results show that our proposed method significantly enhances the accuracy of object pose estimation. It demonstrates excellent generalization, making it applicable to pose estimation for a wide range of new objects. Finally, to validate the practical applicability of ZSPose, we apply it to real-world object pose estimation tasks and robotic grasping tasks. The experimental findings indicate that ZSPose effectively estimates the poses of new objects, assisting robots in performing practical grasping tasks, thus holding considerable practical value. Note to Practitioners—Instance-level object pose estimation is a crucial approach for accurately determining the poses of objects. It enables pose estimation tasks in various scenes when the information about the object model is available and performs well even in cluttered and occluded environments. Robots rely on this pose estimation to effectively complete object grasping tasks. However, most instance-level object pose estimation methods can only estimate the poses of objects that were encountered during the training phase, rendering them incapable of incorporating new objects. This limitation poses significant challenges for practical applications. In real industrial settings, the objects that robots need to grasp can vary widely and often include new items. The lack of scalability in current instance-level object pose estimation methods is a critical issue. To tackle this problem, we propose a novel zero-shot object pose estimation method called ZSPose. This method effectively estimates the poses of new objects that were not part of the training dataset and can be successfully applied to robotic grasping tasks.}, archive = {J_TASE}, author = {Sheng Yu and Di-Hua Zhai and Jian Yin and Yuanqing Xia}, doi = {10.1109/TASE.2025.3618671}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22251--22268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ZSPose: Instance-level zero-shot object pose estimation with segment anything model}, volume = {22}, year = {2025}, }
• (2025). A data-driven evolutionary algorithm for dynamic vehicle routing problems with time windows under limited computational time. TASE, 22, 22239--22250. (www)

  @article{ , abstract = {The Dynamic Vehicle Routing Problem with Time Windows (DVRPTW) is a widespread real-world challenge, and numerous algorithms have been proposed to address it. However, in the context of an emerging logistics paradigm, namely the instant delivery, the performance of existing algorithms tailored for DVRPTW degrades significantly, as instant delivery allows only very limited computational time for solving DVRPTW instances. Owing to the periodic nature of customer orders, this paper proposes a data-driven evolutionary algorithm (DDEA) for solving DVRPTW under limited computational time. In the offline phase, a set of generalized solutions is derived from historical data via a dedicated evolutionary algorithm. These solutions are then directly employed in the online phase to construct high-quality solutions for new problem instances. By leveraging these precomputed generalized solutions, DDEA effectively operates within tight time constraints. Extensive experiments using synthetic and real-world data demonstrate that DDEA outperforms five state-of-the-art algorithms designed for DVRPTW under limited computational time, particularly under extremely short time constraints. Note to Practitioners—This paper proposes a data-driven evolutionary algorithm, DDEA, to solve Dynamic Vehicle Routing Problem with Time Windows (DVRPTW) under limited computational time, a critical challenge in instant delivery services. DDEA utilizes historical data to generate generalized solutions during an offline phase, which are then efficiently deployed to construct high-quality routes for real-time request scenarios. Experimental results demonstrate that DDEA outperforms state-of-the-art algorithms, particularly when the computational time is severely limited. Practitioners in instant delivery services can benefit from integrating DDEA into their routing optimization systems to efficiently construct high-quality delivery routes. Future research can explore further enhancements and integration with real-time data.}, archive = {J_TASE}, author = {Hao Jiang and Yongling Ye and Chao Wang and Xiaoshu Xiang and Tianhang Zhou and Xingyi Zhang}, doi = {10.1109/TASE.2025.3617650}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22239--22250}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A data-driven evolutionary algorithm for dynamic vehicle routing problems with time windows under limited computational time}, volume = {22}, year = {2025}, }
• (2025). Protocol-based reliable control for 2-D fuzzy systems subject to fading networks. TASE, 22, 22230--22238. (www)

  @article{ , abstract = {This paper investigates the fuzzy censoring control problem for two-dimensional (2-D) discrete fuzzy systems characterized by the Roesser model. A novel event-triggered random protocol based on nonhomogeneous sojourn probability is proposed to effectively alleviate communication loads in 2-D fuzzy systems. To address signal fading effects, a censoring strategy is introduced to quantify the severity of signal degradation in transmitted data. The censored control input is subsequently modeled through a Tobit framework, thereby facilitating the design of an integrated state feedback control scheme for 2-D fuzzy systems. By leveraging Lyapunov stability theory, sufficient conditions are derived to guarantee the asymptotic mean-square stability of the closed-loop system while satisfying prescribed performance specifications. Furthermore, parameters are optimized via genetic algorithms to mitigate redundant conservatism associated with linear matrix inequalities. Finally, a gas absorption, water stream heating and air drying industrial case are carried out to verify the theoretical rigor and practical effectiveness of the proposed methodology. Note to Practitioners—Numerous industrial control processes and communication intensive systems can be aptly modeled as 2-D fuzzy systems. In these scenarios, where real-time data transmission over wireless channels is of paramount importance and precise multi-dimensional control is essential, this paper centers on the fuzzy censoring control problem for 2-D discrete fuzzy systems characterized by the Roesser model. Fading networks present a significant challenge in such systems, often resulting in substantial data loss and transmission delays. To tackle these problems, a event-triggered random protocol with non-homogeneous sojourn probability is proposed. This protocol, combined with a customized communication protocol and a fuzzy logic-based control algorithm, is designed to effectively reduce communication burdens in 2-D fuzzy systems. For practitioners implementing this control strategy, it offers enhanced system reliability, improved tolerance to network fading, and more efficient utilization of network resources. This study not only provides practical implementation guidelines but also offers solid theoretical justifications. As a result, it effectively bridges the gap between theoretical research on 2-D fuzzy systems and their real-world applications in network-constrained environments.}, archive = {J_TASE}, author = {Qiongwen Zhang and Yucheng Xu and Jun Cheng and Leszek Rutkowski and Yonghong Chen}, doi = {10.1109/TASE.2025.3617454}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22230--22238}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Protocol-based reliable control for 2-D fuzzy systems subject to fading networks}, volume = {22}, year = {2025}, }
• (2025). Stability guaranteed approximation of model predictive control using unsupervised learning. TASE, 22, 22218--22229. (www)

  @article{ , abstract = {This work presents an unsupervised learning-based approach to approximate the controller of the Lyapunov-based Model Predictive Control (MPC) for a class of continuous-time nonlinear systems. The proposed design aims to produce an optimal control input by aligning the objectives and constraints of the learning problem with those of the MPC. The MPC is approximated by a deep feedforward neural network (DFNN) whose output can strictly satisfy the system constraints. A sufficient condition is provided under which the stability of the closed-loop system with the approximation DFNN control implemented in a sampling-and-hold fashion can be guaranteed. Additionally, we define a polyhedral Lyapunov function to shape the domain of attraction, allowing it to approach the shape of the state boundary defined by polyhedral state constraints, thereby providing potential for enlarging the domain of attraction. The implementation of the proposed method on a permanent magnet synchronous generator wind turbine and a continuous stirred tank reactor illustrates the effectiveness and performance of the designed approximation controller. Note to Practitioners—The approximation controller design proposed in this paper aims to reduce the computational time of Model Predictive Control (MPC) in complex, large-scale systems while meeting real-time performance requirements. The proposed method can compute control signals effectively, ensuring the satisfaction of the system’s constraints. The implementation of the proposed approximation MPC involves the following steps: first, determine the input and output constraints of the dynamic system, along with the stability constraint of the Lyapunov-based MPC, to access the feasible region; second, based on the desired approximation accuracy and the feasible region of the state, specify the sampling scale for each state dimension and construct the training dataset, construct the loss function of the neural network model as the sum of the Laypunov-based MPC performance index evaluated at all states in the training set, and optimize the network parameters with gradient ascent; finally, integrate the approximation Lyapunov-based MPC controller with the control platform. At every time step, the control signal is generated by the trained deep feedforward neural network (DFNN) with optimized network parameters. Preliminary experiments suggest that this approach is feasible.}, archive = {J_TASE}, author = {Yi Zheng and Qibo Liu and Yihua Zhu and Shaoyuan Li and Yuyan Liu}, doi = {10.1109/TASE.2025.3616115}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22218--22229}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stability guaranteed approximation of model predictive control using unsupervised learning}, volume = {22}, year = {2025}, }
• (2025). Hybrid control strategies for mitigating covert malware spread in heterogeneous networks. TASE, 22, 22198--22217. (www)

  @article{ , abstract = {In the current era of information explosion, cybersecurity challenges have become increasingly prominent. Compared to malware that causes direct attacks, covert malware poses a significant cybersecurity threat as it can remain hidden within networks for extended periods while stealthily stealing users’ private information. To address the challenge of clearing covert malware, this study proposes two defense mechanisms: the Patch Propagation Mechanism (PPM) the aSleep-aWake Mechanism (SWM). Based on these mechanisms, we develop a mean-field two-layer propagation model, SW-VIPU (aSleep-aWake-Vulnerable-Infected-Patched-Unaware), to analyze the network’s dynamic evolution. In addition to theoretical analysis, the effectiveness of PPM and SWM in curbing the spread of malware is validated through comparisons with existing models and simulations on real-world network topologies. Subsequently, to enhance the performance of SWM, we introduce a multi-pulse control mechanism and analyze the dynamic performance of SW-VIPU under this mechanism. Finally, to comprehensively explore and improve the control effectiveness, we introduce static strategies as well as an objective-function-based dynamic defense framework, leading to the development of several hybrid control strategies (HCSs). The effectiveness of these HCSs is confirmed through simulation experiments. The experimental results demonstrate that the proposed dynamic HCSs outperform static HCSs in most scenarios. Furthermore, when malware exhibits a high propagation capability, patches should be deployed at their maximum deployment output, and users should be forced to log out and be prohibited from logging back in. Conversely, when malware spread capability is weaker, although the patch output should still be maintained at its highest level, user logins and logouts should not be restricted. Note to Practitioners—This study is driven by the practical challenge of securing large-scale networked systems against covert malware that remains undetected while exfiltrating sensitive information. Owing to the stealthy nature and obscure infection sources of such threats, conventional reactive defense mechanisms often prove inadequate. To address this issue, we propose two proactive defense strategies: the PPM and the SWM. These mechanisms are designed to autonomously suppress malware spread through distributed control of patch propagation and user connectivity. They are particularly suitable for deployment in environments such as enterprise networks and cloud infrastructures, where automated control is feasible. Simulation results demonstrate that dynamic HCSs consistently outperform their static counterparts across a wide range of scenarios. Moreover, the results offer practical guidance: in the presence of highly virulent malware, maximum patch propagation combined with enforced user disconnection is essential, whereas for low-virulence threats, maintaining a high patch rate without restricting user activity is sufficient. Future work will focus on integrating the proposed strategies into real-time cybersecurity systems and evaluating their effectiveness in practical operational settings.}, archive = {J_TASE}, author = {Baihao Peng and Junfeng Liu and Jun Zeng}, doi = {10.1109/TASE.2025.3614747}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22198--22217}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hybrid control strategies for mitigating covert malware spread in heterogeneous networks}, volume = {22}, year = {2025}, }
• (2025). Intent-based autonomous network framework guided by large language model. TASE, 22, 22185--22197. (www)

  @article{ , abstract = {With the rapid development of next-generation networks, the highly heterogeneous and dynamic nature of networks poses significant challenges for automated network management. Autonomous Network (AN), as a new network paradigm, aims to provide customers with a zero-wait, zero-touch, and zero-fault experience. AN facilitates network management through intent-driven interactions and provides on-demand resource orchestration and service scheduling. However, accurately translating user intents into commands and allocating resources on demand for services remain significant challenges for AN. Therefore, this paper proposes IAN, an intent-based AN framework guided by the Large Language Model (LLM). In the intent translation phase, IAN introduces RAG to enhance command generation quality by retrieving from manuals. In the resource allocation phase, the method utilizes LLM to analyze service characteristics, thereby guiding the training and inference of the resource allocation model to effectively distribute resources uniformly across emerging services. Experimental results demonstrate that IAN improves performance by 52.66% in intent translation tasks and increases overall gain by 33.57% in resource allocation tasks compared to other models. Note to Practitioners—In this study, the goal is to design an intent-driven autonomous framework. This framework translates user-input natural language intents into network configurations and continuously updates these configurations based on computations of downstream network states to achieve on-demand resource allocation. To accomplish this, we designed two main modules: the Intent Translator and the Resource Allocator. In the Intent Translator, we propose an LLM-empowered intent translation approach. The introduction of RAG enhances the performance of LLMs in intent translation tasks, effectively mitigating hallucinations and generating higher-quality, deployable commands. In the Resource Allocator, we present an LLM-guided resource allocation method. Leveraging their strong contextual understanding, LLMs can better analyze the attributes of emerging services, thereby guiding the training and inference of resource allocation models. Through this approach, the proposed framework achieves high-quality intent translation and efficient resource allocation, enabling intent-driven network autonomy.}, archive = {J_TASE}, author = {Lingqi Guo and Lei Zhang and Jingyu Wang and Jianyu Wu and Yuhang Yan and Haifeng Sun and Bo He and Qi Qi and Jianxin Liao}, doi = {10.1109/TASE.2025.3610906}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22185--22197}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intent-based autonomous network framework guided by large language model}, volume = {22}, year = {2025}, }
• (2025). Toward indirect data-driven predictive control for heating phase of thermoforming process. TASE, 22, 22173--22184. (www)

  @article{ , abstract = {Shaping thermoplastic sheets into three-dimensional products is challenging since overheating results in failed manufactured parts and wasted material. To this end, we propose an indirect data-driven predictive control approach using Model Predictive Control capable of handling temperature constraints and heating-power saturation while delivering enhanced precision and overshoot control compared to state-of-the-art methods. We employ a Non-linear Auto-Regressive with Exogenous inputs model, which is linearized to define a linear control-oriented model at each operating point. Using a high-fidelity simulator, several simulation studies have been conducted to evaluate the proposed method’s robustness and performance under parametric uncertainty, indicating overshoot and average steady-state error less than $2^{\circ } \mathrm {C}$ and $0.9^{\circ } \mathrm {C}$ ( $7^{\circ } \mathrm {C}$ and $2^{\circ } \mathrm {C}$ ) for the nominal (worst-case) scenario. Finally, we applied the proposed method to a lab-scale thermoforming platform, resulting in a close response to the simulation analysis with overshoot and average steady-state error metrics of $5.3^{\circ } \mathrm {C}$ and $1^{\circ } \mathrm {C}$ , respectively. Note to Practitioners—This work addresses a key challenge in the thermoforming process: precisely controlling sheet temperature during the heating phase to improve product quality and reduce waste. We present a data-driven control approach that can be implemented on existing thermoforming systems to enhance temperature accuracy and consistency. The proposed method uses historical process data to build a simplified model of heating dynamics. This model is then used in a predictive control framework to optimize real-time heater settings. The key advantages of the proposed method are improved temperature precision, faster settling times than conventional methods, the ability to handle practical constraints like maximum temperatures and heating power limits, and robustness to process variations and disturbances. Complex physics-based models or extensive system modifications are not needed. To implement this approach, practitioners would need to collect temperature and heater power data from their existing process, use it to train the data-driven model and integrate the control algorithm with their PLC or control system.}, archive = {J_TASE}, author = {Hadi Hosseinionari and Mohammad Bajelani and Klaske van Heusden and Abbas S. Milani and Rudolf J. Seethaler}, doi = {10.1109/TASE.2025.3618582}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22173--22184}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward indirect data-driven predictive control for heating phase of thermoforming process}, volume = {22}, year = {2025}, }
• (2025). Maximum correntropy unscented kalman filter with unsupervised adaptive kernel scale selection. TASE, 22, 22153--22172. (www)

  @article{ , abstract = {Dynamic systems in practice are often nonlinear and subject to complex disturbances such as sensor outliers, which invalidate the Gaussian noise assumptions underlying conventional Kalman filters. To address the limitations of conventional Kalman filters under such conditions, recent studies have explored nonlinear filters based on the maximum correntropy criterion (MCC), which exploit both second-order and higher-order moments of the innovation for enhanced robustness. However, their performance is highly sensitive to the choice of kernel scale, and existing strategies—whether fixed offline or empirically adjusted—fail to adapt to unknown disturbances of varying intensities. This paper proposes a novel unsupervised adaptive maximum correntropy unscented Kalman filter (AMCUKF), which introduces three major contributions: 1) Online Kernel Scale Adaptation—the kernel scale is modeled as a latent random variable governed by an inverse Gamma distribution, and variational Bayesian inference is applied to jointly estimate the system state and kernel scale in a recursive manner; 2) Theoretical Consistency and Closed-Form Solutions—the algorithm maintains consistency with traditional MCUKF but achieves dynamic adaptability through the conjugacy of Gaussian and inverse Gamma priors, enabling efficient closed-form updates; and 3) Validated Performance under Non-Gaussian Noise—comprehensive tests on three representative scenarios, including nonlinear numerical systems with stationary/non-stationary $t$ -distributed and Cauchy measurement noise, a single-target tracking task, and real-world battery state-of-charge (SoC) estimation using the public LG 18650 HG2 dataset. Compared with UKF, EMCUKF and FMCUKF, the proposed AMCUKF achieves about 20–60% lower RMSE and 30–70% lower worst-case error across heavy-tailed noise simulations, while maintaining comparable computational cost and convergence within 1–3 iterations. In the real-world SoC estimation task, it further reduces the maximum estimation error from over 7% to below 1%, demonstrating strong practical robustness. Note to Practitioners—This work addresses the practical need for robust nonlinear state estimation under unknown disturbances and non-Gaussian noise. 1) By treating the kernel scale as a latent random variable and updating it online via variational Bayesian inference, the proposed method achieves adaptive filtering behavior without requiring manual tuning, even under time-varying interference. 2) Its ability to cope with heavy-tailed noise and corrupted measurements makes it highly suitable for real-time applications such as radar target tracking, autonomous robotics, and multi-sensor fusion. 3) The algorithm exhibits fast convergence within 1–3 iterations per step and can be integrated into standard UKF-based pipelines with minimal modification, preserving computational efficiency. 4) For practitioners dealing with nonlinear systems and uncertain sensing conditions, this approach offers a practical enhancement in both robustness and estimation accuracy. 5) Future efforts will focus on extending its applicability to multi-target tracking and more complex, cluttered environments.}, archive = {J_TASE}, author = {Guanghua Zhang and Sitong Li and Xiqian Zhang and Dou An and Feng Lian and Xinqiang Liu}, doi = {10.1109/TASE.2025.3617541}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22153--22172}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Maximum correntropy unscented kalman filter with unsupervised adaptive kernel scale selection}, volume = {22}, year = {2025}, }
• (2025). Wavelet-discrete cosine transform synergy for ship motion-based sea state estimation in autonomous ships. TASE, 22, 22140--22152. (www)

  @article{ , abstract = {Developing a robust autonomous sea state estimation (SSE) model stands as a pivotal challenge in advancing autonomous ships. Presently, deep learning (DL) methodologies have showcased remarkable efficacy in SSE tasks. Nonetheless, the dynamic nature of ship motion introduces temporal variations alongside frequency domain characteristics like periodic swinging, posing challenges for existing DL approaches. Most prevailing DL techniques, predominantly leveraging Convolutional Neural Networks or Long Short-Term Memory Networks, often fail to effectively harness frequency domain information post feature extraction. To tackle these limitations head-on, this paper introduces a pioneering SSE model. Specifically, in order to solve the frequency-domain feature extraction problem, we design a wavelet transform-based frequency domain encoder to extract relevant frequency-domain features from ship motion data by discriminating the contribution of different frequencies in the signal. Subsequently, in order to better integrate the extracted ship motion features, we designed a Feature Perception module based on discrete cosine transform. This module adeptly merges the extracted feature insights while prioritizing crucial frequency domain features. Following rigorous experimentation, our methodology exhibits superior performance compared to existing baseline techniques in SSE, a capability of profound significance for autonomous ships. Moreover, across diverse public multivariate time series classification datasets, our model outperforms current state-of-the-art approaches, underscoring its scalability across distinct domains. Note to Practitioners—Note to Practitioners (NtP): In the development of autonomous ships, precise SSE is crucial for ensuring safety and efficiency. The inherent uncertainties of open sea environments amplify the complexity of autonomous vessel navigation, presenting significant challenges. This paper introduces a DL model that leverages wavelet transform and discrete cosine transform to enhance the feature extraction of ship motion data. Unlike traditional methods that rely on manual feature extraction, our model autonomously captures both time-domain and frequency-domain features, thereby addressing the complexities of ship motion. In the discussion section of this paper, we focus on the model’s performance in handling the prevalent issue of sea state imbalance and its real-time inference capabilities. The results validate the model’s applicability and reliability, suggesting that practitioners can deploy this model on ships to better support the advancement of autonomous ships.}, archive = {J_TASE}, author = {Xiaorui Zhang and Feng Xiao and Xu Cheng and Saša V. Nikolić and Jianhua Zhang and Shengyong Chen}, doi = {10.1109/TASE.2025.3617252}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22140--22152}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Wavelet-discrete cosine transform synergy for ship motion-based sea state estimation in autonomous ships}, volume = {22}, year = {2025}, }
• (2025). Regional feature enhancement domain adaptation for automatic road damage detection in hazy and rainy weather. TASE, 22, 22124--22139. (www)

  @article{ , abstract = {Currently, most neural networks for automatic road damage detection are trained using labeled normal weather datasets, and the detection performance decreases dramatically in hazy and rainy weather, while labeling adverse weather road damage datasets is a rather difficult task. Therefore, we propose a regional feature enhancement domain adaptation method (RFEDA), which employs a two-stage strategy of inter-domain adaptation and intra-domain adaptation to alleviate the domain shift between normal weather and hazy and rainy weather road damage images. Specifically, RFEDA uses a regional feature enhancement module (RFEM) to segment the damage region. The semantic segmentation task facilitates the object detection task to locate the damage object more accurately through path sharing, thus suppressing the interference of background noise. Meanwhile, the segmented region features are utilized to construct instance-level features for instance-level feature alignment. In addition, the multi-scale features extracted by backbone are used for image-level feature alignment. In the intra-domain adaptation stage, the model after inter-domain adaptation is used to generate high-confidence pseudo-labels for the training set of the target domain, and the pseudo-labels are used for self-training of the model and thus for fine-tuning the model. Extensive experiments on the CNRDD and Japan-RDD datasets in hazy and rainy weather demonstrate the effectiveness of our method. Compared with no domain adaptation, RFEDA can improve the mAP@0.5 by up to 16.9% and 7.4% in hazy and rainy weather, respectively. Note to Practitioners—Installing automatic road damage detection systems on smart cars is a development trend, and such systems need to work properly under different weather conditions. We propose the RFEDA method to realize unsupervised domain adaptation road damage detection under hazy and rainy weather. We adopt a two-stage domain adaptation strategy to train the model. In the first stage, labeled road damage images in normal weather are used as the source domain, and unlabeled road damage images in haze or rainy weather are used as the target domain for training. In the second stage, the samples are split into easy and hard samples based on high-confidence pseudo-labels. Self-training is performed with easy samples as source domain and hard samples as target domain. Experiments show that our domain adaptation method can achieve good performance in the YOLO series object detectors.}, archive = {J_TASE}, author = {Jie Li and Zhong Qu and Xuehui Yin}, doi = {10.1109/TASE.2025.3613972}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22124--22139}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Regional feature enhancement domain adaptation for automatic road damage detection in hazy and rainy weather}, volume = {22}, year = {2025}, }
• (2025). A generalist agent learning architecture for versatile quadruped locomotion. TASE, 22, 22112--22123. (www)

  @article{ , abstract = {Quadrupeds can generate various motor behaviors with the muscle synergies activated by the central nervous system. However, versatile locomotion for quadruped robots remains challenging due to the complexity of the high-dimensional limb dynamics with many physical constraints. Current approaches typically apply a dedicated policy or controller for each motor behavior, which requires the optimization of a large number of parameters and training process is complicated. In this paper, we propose a Generalist Agent Learning Architecture (GALA) to learn diverse motor behaviors simultaneously with a single policy network for quadruped locomotion. GALA significantly decreases the number of trainable parameters while producing appropriate motor behaviors by simply reactivating the generalist policy based on different sensory feedback and commands at run time. We experimentally analyze and demonstrate the versatile locomotion delivered by GALA on both simulated and real quadruped robots in various environments. The code is available at https://github.com/vsislab/GALA. Note to Practitioners—This paper presents a generalist agent learning architecture (GALA) in which a single policy network learns various motor behaviors simultaneously for quadruped robots. GALA technically integrates multiple critics actor learning with a generalist policy to generate a diversity of motor behaviors in accordance with high-dimensional sensory feedback and commands. Our method is easy to train with a much smaller number of trainable parameters for versatile locomotion. The policy can be explored and exploited in simulated environments and directly deployed on the real robot through domain randomization technique that has been applied to many robots for sim-to-real transfer. The GALA-based quadruped robot flexibly generates versatile motor behaviors to smoothly traverse a series of real-world environments in runtime.}, archive = {J_TASE}, author = {Yanyun Chen and Ran Song and Jiapeng Sheng and Xing Fang and Wenhao Tan and Yibin Li and Wei Zhang}, doi = {10.1109/TASE.2025.3612147}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22112--22123}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A generalist agent learning architecture for versatile quadruped locomotion}, volume = {22}, year = {2025}, }
• (2025). Stability and stabilization of positive fuzzy systems via polynomial lyapunov functions with application to tank level control. TASE, 22, 22102--22111. (www)

  @article{ , abstract = {In this paper, new stability conditions are obtained by designing a Lyapunov function that contains polynomials of the system states and membership functions. An iterative algorithm is designed to solve the stabilization problem. Since more free variables are introduced by using the square root of the system states and square matrix representation, the proposed conditions are less conservative than existing ones. Furthermore, the time derivative of the membership functions is dealt with by applying polynomial switching mechanism, while the average dwell time approach is employed to address the switching signal. The effectiveness of the proposed method is demonstrated through three numerical examples, including a practical tank level control system, confirming its applicability to real-world scenarios. This paper shows that for positive fuzzy systems, better results can be achieved by utilizing the positivity of the system states and the information of the membership functions. Note to Practitioners—Positive fuzzy systems have found wide applications in engineering domains such as chemical process control, biological systems, and water resource management, where system variables are inherently non-negative. However, traditional control methods often fail to fully exploit the structural characteristics of such systems, resulting in conservative designs and limited performance. Motivated by these challenges, this paper proposes a new stabilization approach that utilizes the positivity of the system states and the information of the membership functions. A Lyapunov function is constructed by incorporating the square roots of state polynomials and polynomial terms of the membership functions, and an iterative algorithm is developed to solve the associated stabilization conditions. Furthermore, the relationship between the stability region and the degree of the polynomials is investigated, which has important implications for enhancing overall system performance.}, archive = {J_TASE}, author = {Yinghong Zhao and Likui Wang and Qianling Wang and Hak-Keung Lam}, doi = {10.1109/TASE.2025.3618681}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22102--22111}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stability and stabilization of positive fuzzy systems via polynomial lyapunov functions with application to tank level control}, volume = {22}, year = {2025}, }
• (2025). Multimodal shared control of a fully wearable prosthetic Hand/Wrist system. TASE, 22, 22085--22101. (www)

  @article{ , abstract = {Expanding the input bandwidth of the human-machine interface to capture more control intentions from the human is key to achieving dexterous control of multi-degree-of-freedom prosthetic hands/wrists. This paper presents a wearable intelligent prosthesis system based on multimodal fusion, which integrates voice interaction, myoelectric control, limb movement decoding, and computer vision-based environmental perception. The system supports intention estimation throughout the entire process from grasping to operation, enabling synchronized control of 4 grasp gestures and 2 wrist DOFs. Moreover, all these decisions are made automatically during the user’s natural and continuous body movements. Six healthy subjects and two amputee subjects participated in a comparative experiment involving multi-object grasping and operation in a cluttered environment. Compared to myoelectric pattern recognition, our method demonstrated significant advantages in improving grasp and operation efficiency (reducing total time, grasp time, and operation time by 67.62%, 67.89%, and 67.40% for healthy subjects, and by 73.93%, 69.82%, and 76.67% for amputees). It also reduced control burden, with gesture and wrist myocontrol times decreasing by 71.77% and 100% in healthy subjects, and by 80.59% and 100% in amputees. These advantages are not limited to the grasp object, grasp part, operation type, or the subject involved. Compared to other semi-autonomous control methods, our method achieved a higher reduction in time metrics, with reduction rates ranging from 22.61% to 84.82% higher, resulting in a more significant performance improvement. Furthermore, the questionnaire showed that the system was well-recognized by the subjects in terms of operational robustness, wearing comfort, and user experience. Note to Practitioners—This paper aims to enhance the rehabilitation performance and user acceptance of prosthetic devices by proposing a truly practical wearable intelligent prosthetic system. We developed a prototype that can be worn on a single arm and proposed a comprehensive control framework. This framework ensures users have full control authority (via electromyography and voice) while leveraging the system’s autonomous decision-making capabilities (through local visual perception and limb movement decoding), effectively reducing the user’s control burden. By incorporating voice feedback and a coordinated control module, users can supervise, correct, and fine-tune automated decisions, further enhancing system stability. The reliance on datasets is a limitation inherent to machine learning, as reflected in this paper. However, this issue can be effectively addressed through dataset expansion and transfer learning. In future research, we will focus on achieving “hand-eye integration” in hardware design and developing more advanced vision algorithms. We also aim to tailor the system according to individual user habits, providing a more user-friendly and stable control experience.}, archive = {J_TASE}, author = {Chunhao Peng and Dapeng Yang and Deyu Zhao and Jinghui Dai and Li Jiang and Hong Liu}, doi = {10.1109/TASE.2025.3615137}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22085--22101}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multimodal shared control of a fully wearable prosthetic Hand/Wrist system}, volume = {22}, year = {2025}, }
• (2025). Integral action in variable impedance control of articulated-soft robots. TASE, 22, 22072--22084. (www)

  @article{ , abstract = {This paper presents a novel control framework for enhancing the tracking and interaction performance of articulated soft robots (ASRs), via an amalgamation of variable impedance control (VIC) and integral action. Although VIC is generally known to yield a certain degree of improvement in terms of interaction and/or motion control performance, the incorporation of integral terms can further ameliorate robustness to unknown disturbances. To this end, the proposed methodology exploits power-shaping-signal control (PSC) as a means of suppressing the energy injections pertaining to VIC, thereby yielding an overall (globally asymptotically) stable closed-loop system, even in the presence of integral action. Further, its implementation requires knowledge of only the (linear) motor dynamics and the gravitational torque vector, thus obviating the need for calculation of the Coriolis and link inertia matrices; this in turn renders the usage of link jerk feedback redundant, thereby leading to considerable improvements w.r.t physical interaction, as is known from the existing literature. A series of physical experiments involving an ASR, namely the Rethink Robotics Baxter, corroborates the proposed scheme’s ability to practically conduce to enhanced position/motion and interaction control performance, in addition to improved robustness to unknown disturbances. Note to Practitioners—The work presented in this paper is motivated by the emergent requirement for more efficacious and adaptable control schemes that will pave the way for widespread adoption of ASRs in practical settings. Although ASRs are naturally backdrivable, most existing algorithms do not fully exploit their immanent compliance to achieve improved interaction performance. Moreover, their inherent elasticity can lead to motion control degradation, as compared to their rigid-joint counterparts. To this end, the proposed control methodology (namely UMIVIC) combines improved interaction performance – via VIC – with augmented trajectory-tracking accuracy through link-side integral action. As compared to other, existing ASR tracking controllers with stability guarantees, the proposed scheme is more straightforward to implement, since it obviates the need for computation of link inertia and Coriolis elements, whose algorithmic complexity can preclude real-time implementation on physical systems; this also renders link jerk feedback redundant, which considerably improves interaction performance.}, archive = {J_TASE}, author = {Emmanouil Spyrakos-Papastavridis and Kun Wang and Jian S. Dai}, doi = {10.1109/TASE.2025.3612734}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22072--22084}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integral action in variable impedance control of articulated-soft robots}, volume = {22}, year = {2025}, }
• (2025). Adaptive prescribed-time target capture control for marine vehicles with game-based navigation risk field. TASE, 22, 22059--22071. (www)

  @article{ , abstract = {This paper proposes an adaptive prescribed-time target capture control algorithm for marine vehicles with collision avoidance, incorporating intelligent guidance and control design. A novel game-based navigation risk field is established to capture the escaping target and avoid collisions with any obstacles. Particularly, an attractive potential field centered on the Nash equilibrium is established to guide the capture activity. In addition, a velocity-related navigation risk field is designed to describe the different risk regions around the dynamic obstacle, ensuring that collision avoidance maneuvers comply with the International Regulations for Preventing Collisions at Sea (COLREGs). A novel feedback-dependent adaptive prescribed-time control algorithm is introduced to guarantee the predefined control performance within the prescribed time. The error feedback-based scaling function is adjusted online by balancing control performance and engineering requirements. That guarantees the adaptation of the convergence rate and prescribed time. Furthermore, a fuzzy prescribed-time observer is constructed to compensate for ocean disturbances and model uncertainties. All the signals of the closed-loop system are uniformly bounded, and the tracking errors converge to an acceptable accuracy within the prescribed time. The experiment results demonstrate the effectiveness and rapid convergence of the proposed algorithm. Note to Practitioners—This paper is motivated by the problem of target capture and navigation safety for marine vehicles in dynamic maritime environments. The target capture operations demand precision, timeliness, and strict compliance with safety regulations such as COLREGs. The proposed scheme integrates an adaptive prescribed-time control algorithm with a game-based guidance strategy to address these demands. In this scheme, an attractive potential field centered on the Nash equilibrium is established to guide capture activities. A velocity-dependent risk field is constructed to determine collision avoidance strategies in accordance with COLREGs. By utilizing a scaling function to transform the system states, the prescribed-time control scheme ensures system convergence within a user-specified time. Traditional prescribed-time schemes, which rely on fixed scaling functions, lack adaptability to varying system dynamics. For this limitation, a feedback-dependent adaptive scaling function is proposed to dynamically adjust the control gain based on real-time error feedback. This design enhances the controllers flexibility and responsiveness in dynamic environments. Furthermore, a fuzzy prescribed-time observer is incorporated to estimate and compensate for the adverse impact of ocean disturbances and model uncertainties, further improving robustness. By balancing rapid convergence, robust performance, and operational safety, the proposed algorithm is suitable for real-time target capture tasks in unpredictable maritime environments. Beyond marine applications, this scheme can be extended to other nonlinear systems for high-timeliness tasks, such as intruder interception.}, archive = {J_TASE}, author = {Shengjia Chu and Ning Li}, doi = {10.1109/TASE.2025.3612430}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22059--22071}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive prescribed-time target capture control for marine vehicles with game-based navigation risk field}, volume = {22}, year = {2025}, }
• (2025). Activation control of laguerre neural network. TASE, 22, 22044--22058. (www)

  @article{ , abstract = {The Laguerre Neural Network (LaNN) employs Laguerre polynomials (LPs) as the activation functions of its hidden neurons (HNs) to approximate the nonlinear functions (NFs). From the perspective of statistics, the more LPs are activated, the more LP-based activation functions (LPAFs) of the HNs will operate, and thus the smaller approximation error (APE) will be suffered by the LaNN, while leading to more energy consumption since activating any LP has to consume energy. Therefore, how to control the activation of LPs to make the optimal tradeoff between approximation accuracy and energy consumption arises as an interesting issue. To address this issue, this paper first establishes an energy-constraint probability-based approximation-accuracy optimization-theoretical (EPAO) framework, considering the probability of obtaining the minimum APE (MAPE) as the objective while the energy consumption of the LaNN as the constraint, then reveals the optimal number of HNs and thus their optimal LPAFs, and finally discloses the optimal activation probability for each LP. Based on our discovered theoretical results, a novel algorithm for LaNN to maximize the probability of obtaining the MAPE at the acceptable level of energy consumption is proposed. Theoretical analysis verifies its performance, and field experiments verify that, compared with existing methods, it improves the probability of obtaining the MAPE by 6.00% to 20.26% at the similar cost of energy. To our best knowledge, this is the initial work for LaNN towards joint optimization of its energy efficiency and approximation accuracy. Note to Practitioners—For practitioners implementing our optimized LaNN approach, here is a straightforward step-by-step process: First, quantify the energy budget of your system based on operational constraints like battery life or power limits. Then, identify the optimal number of HNs and their optimal LPAFs in terms of best balancing approximation accuracy and energy consumption for LaNN according to our design criterion regarding HNs and LPAFs. Next, determine the optimal activation probability for each LP by using our activation criterion for LPs. Finally, perform activation for each LP with the previous optimal activation probability. When deploying, configure the LaNN to dynamically adjust its LP activation for maximizing the probability of obtaining the MAPE while staying within energy limits. Field tests demonstrate that careful activation probability tuning can yield significant energy savings with minimal accuracy loss, making this approach ideal for battery-powered applications requiring precise neural network performance. The complete mathematical form of EPAO, theoretical results, and experimental validation can be found in the technical sections of this paper for those requiring deeper details. As LaNN is an effective model for accurately emulating complex real-world systems in AI, our proposed solution can enable their high-accurate and energy-efficient operation, thereby advancing low-power AI application.}, archive = {J_TASE}, author = {Chen Hou}, doi = {10.1109/TASE.2025.3617969}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22044--22058}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Activation control of laguerre neural network}, volume = {22}, year = {2025}, }
• (2025). Energy-optimal control of generalized laguerre neural network. TASE, 22, 22028--22043. (www)

  @article{ , abstract = {Generalized Laguerre neural network (GLNN) is a specialized artificial neural network that employs generalized Laguerre polynomials (GLPs) as activation functions for hidden neurons, offering distinct advantages in function approximation tasks. Statistically, the larger number of activated GLPs, the more hidden neurons utilizing these GLPs will be engaged, thereby enhancing approximation accuracy. However, this comes with a tradeoff: Each arithmetic operation inherent to GLP computations consumes energy, leading to higher energy consumption. Consequently, optimizing the activation patterns of GLPs to strike an optimal tradeoff between approximation accuracy and energy efficiency emerges as a compelling research issue. To address this issue, this paper establishes an energy-constrained approximation accuracy optimization framework. Within this framework, we reveal the optimal number of candidate GLPs serving as activation functions for hidden neurons, the optimal number of hidden neurons that should be set in the GLNN, and the optimal activation criterion for candidate GLPs, under energy constraint. We also establish the root-exponentially decay convergence of the minimum approximation error under energy constraint. An algorithm based on our discovered results is proposed for the GLNN to minimize the approximation error under the acceptable level of energy consumption. Theoretical analysis and experiments jointly demonstrate its performance. Note to Practitioners—In the realm of artificial neural networks, the GLNN stands out with its use of GLPs as hidden neuron activation functions, providing notable advantages for function approximation tasks. However, in practical applications, the practitioners may encounter a common tradeoff: While activating a larger number of GLPs can enhance approximation accuracy, it simultaneously increases energy consumption due to the energy-intensive arithmetic operations involved in GLP computations. Our research addresses this critical issue directly. We have developed an energy-constrained approximation error optimization framework that offers clear and actionable insights for this work. By using this framework, we have systematically determined the optimal number of candidate GLPs serving as activation functions for hidden neurons, the ideal number of hidden neurons that should be set in the GLNN, and the most effective GLP activation criterion, all while staying within energy budget. The discovery of the root-exponentially decay convergence of the minimum approximation error under energy constraint provides a theoretical basis for practitioners to predict GLNN’s performance. We have also created an algorithm based on our research findings. This algorithm allows GLNN to minimize approximation error while keeping energy consumption at an acceptable level. Whether the practitioners are working on signal processing, data analysis, or other applications that rely on GLNN, implementing this algorithm can facilitate them to achieve better approximation without sacrificing energy efficiency beyond predetermined budget. To sum up, this work fills a critical gap in energy-efficient neural networks and AI research: the joint optimization of approximation accuracy and energy efficiency for GLNNs. It represents a credible and specialized advancement in the field of energy-efficient GLNNs, offering practical guidance for low-power data processing within the broader category of orthogonal polynomial neural networks. This enables practitioners to fully harness the capabilities of GLNNs while managing the constrained energy resources effectively.}, archive = {J_TASE}, author = {Chen Hou}, doi = {10.1109/TASE.2025.3617782}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22028--22043}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy-optimal control of generalized laguerre neural network}, volume = {22}, year = {2025}, }
• (2025). A novel adaptive prescribed-time controller for uncertain systems with unknown non-smooth nonlinearities. TASE, 22, 22018--22027. (www)

  @article{ , abstract = {This paper studies the adaptive prescribed-time control problem for the strict-feedback systems with unknown non-smooth nonlinearities. Different from the existing adaptive prescribed-time designs where the nonlinearities are required to be known and smooth, a novel design is developed, for the first time, in the presence of the unknown non-smooth nonlinearities eliminating the need for smoothness or knowledge. Moreover, in the design we suggest a new scaled quadratic Lyapunov function depending on a tuning parameter, which together with the characteristics of the introduced scaling function constructs a state-feedback controller allowing to balance the tradeoff between the transient performance and control effort. As a byproduct of this development, we also provide solutions to prescribed-time control of the strict-feedback nonlinear systems. It should be noted that the designed controller has smaller gains thus with reduced control effort while ensuring the prescribed-time stabilization. Finally, three examples are provided to illustrate the efficiency of the proposed method. Note to Practitioners—Many practical engineering systems, such as robotic arms and ship rudders, exhibit unknown non-smooth nonlinearities. The characteristic may cause the conventional prescribed-time controllers to be not applicable. This paper proposes a novel adaptive prescribed-time control strategy, which achieves the prescribed-time stability of the uncertain non-smooth nonlinear system. Since a new scaled quadratic Lyapunov function depending on a tuning parameter is proposed, the merit of the proposed strategy involves smaller gains thus with less control effort compared with existing methods, which offers efficient solutions that hold immense promise in practical applications.}, archive = {J_TASE}, author = {Jin Zhang and Mo Shi and Chen Peng and Songlin Hu}, doi = {10.1109/TASE.2025.3617084}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22018--22027}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel adaptive prescribed-time controller for uncertain systems with unknown non-smooth nonlinearities}, volume = {22}, year = {2025}, }
• (2025). One-shot cross-domain instance detection with universal representation. TASE, 22, 22003--22017. (www)

  @article{ , abstract = {In this work, we pay research efforts to address the challenging research problem of One-shot Cross-domain Instance Detection (OCID). Particularly, given a one-shot exemplar instance (e.g., from the visible spectrum), OCID aims to detect the instance’s specific counterpart within another domain (e.g., the infrared spectrum). It can be applied to robotics grasping, aerospace navigation, etc. To address this task, one key issue is to extract features of strong instance discrimination and domain adaptation capability. To this end, we propose to characterize the instance with universal representation from multiple deep networks pretrained on different datasets of variational representation biases. Particularly, they are fused with different backbone attention weights via Transformer to yield a universal representation for adaptively fitting the one-shot instance. Meanwhile, inter and intra domain contrastive learning is conducted to drive universal representation learning with hard sample mining, to facilitate domain adaptation capability and discriminative power jointly. For test, we build a new dataset that contains 14 data sources with over 6K instances of diverse types with dramatic appearance variation. Experiments on this dataset verify OCID’s challenges and the superiority of our proposed method. The source code and dataset have been released at https://github.com/OCIDwUR/OCID Note to Practitioners—Given a one-shot exemplar instance, OCID aims to advance the research of general instance detectors, while considering three practical challenges simultaneously: data scarcity, domain shift, and instance-level discrimination. The “general” denotes that the types of target and task scenarios in OCID are diversified so that the detection methods are challenged with complex pattern distribution and are encouraged to possess open-set recognition capability, with the aim to meet different detection requirements such as autonomous driving and robotics. We also consider different settings of domain shift, such as multiple spectra and image degradation, to make it more applicable. To address the challenging OCID task, we propose a Transformer-based universal representation method that learns to dynamically ensemble a group of pretrained feature backbones with various representation biases and adapts to new even unseen OCID tasks without fine-tuning. The domain adaptation capability and instance discrimination are jointly facilitated, in a meta-learning and hence transferable manner. The experimental results on our carefully designed diversified dataset validate the generality of our method.}, archive = {J_TASE}, author = {Chen Feng and Jian Cheng and Yang Xiao and Zhiguo Cao}, doi = {10.1109/TASE.2025.3616634}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {22003--22017}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {One-shot cross-domain instance detection with universal representation}, volume = {22}, year = {2025}, }
• (2025). Adaptive predefined-time safety learning control for switched multi-agent systems: An advanced encryption self-triggered algorithm. TASE, 22, 21990--22002. (www)

  @article{ , abstract = {This study develops an advanced self-triggered predefined-time safety learning control algorithm for switched multi-agent systems with full-state mask. To strengthen encryption while reducing the impact to system performance, an improved settling time privacy preservation mechanism based on the full-state mask function is designed, which encrypts the true information of the system and enhances the privacy of information delivery. Unlike traditional learning control schemes, a novel actor-critic weight update law is designed to guarantee that the system energy cost is minimized resulting in predefined time optimization. Besides, an improved self-triggered condition with a compensation term is developed to overcome the complex challenges posed by full-state privacy preservation mechanism. It not only eliminates the necessity to continually monitor the triggered state of the system but also saves communication resources. Finally, the validity of the designed control scheme can be proven by a simulation experiment. Note to Practitioners—This paper is motivated by designing a self-triggered predefined-time optimal control method for switched multi-agent systems with full-state mask. As artificial intelligence continues to develop, information security and resource waste have become key problems in practical applications, such as smart grids and autonomous vehicles. In this study, we develop privacy preservation mechanism and optimal control strategy that minimize energy cost to overcome the above challenges. It is worth noting that the UAV swarm switches between different modes to adapt to different environments during the mission. With the help of a self-triggered algorithm, the photovoltaic inverter calculates the next control moment in advance to avoid continuous sampling during sudden changes in light thereby saving communication resources. Additionally, the predefined time control method is characterized by its fast and accurate convergence, which makes it widely used in fields such as missile guidance and rehabilitation robots.}, archive = {J_TASE}, author = {Shiyu Xie and Wei Sun}, doi = {10.1109/TASE.2025.3615563}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21990--22002}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive predefined-time safety learning control for switched multi-agent systems: An advanced encryption self-triggered algorithm}, volume = {22}, year = {2025}, }
• (2025). Finite-time consensus of input-constrained networked fractional-order systems and its application to ship power systems. TASE, 22, 21975--21989. (www)

  @article{ , abstract = {In this paper, finite-time consensus is investigated for networked fractional-order systems with input constraints. Unlike existing studies where only the upper bound of settling time is considered, both explicit upper and lower bounds are established for the settling time range. This topic is challenged by the inapplicability of integer-order system properties to fractional-order systems, the complexities caused by asymmetric interactions in directed networks, and the inherent contradictions between finite-time stabilization and input constraints. To resolve these issues, two scenarios of consensus in leaderless directed networks and containment control in leader-following directed networks are considered. A novel fractional-order derivative property for composite functions is developed to handle the limitations of traditional integer-order calculus. Integral-type Lyapunov functions are constructed through graph theory to address the challenges posed by asymmetric Laplacian matrices. State evolution is analyzed comprehensively for both truncated and free input cases, with explicit bounds derived for the settling time. Finally, numerical simulation and practical demonstration of ship power system are provided to verify the effectiveness of the proposed methods. Note to Practitioners— Due to the fact that the fractional-order dynamics can more accurately describe the characteristics of actual industrial systems than integer-order dynamics, the cooperative control of networked fractional-order systems is studied, which can be applied to the ship power systems. Our study proposes a novel fractional-order method that is endowed with the properties of input-constrained satisfactions, zero-error finite-time convergence and chatter avoidance. The common existing finite-time methods neglect the lower bound and contradict input constraints, hinder their implementations in time-sensitive tasks. We address these limitations and explicitly give the range of settling time. By comparing the simulation results, the fractional-order controller of this paper has superior control performances over the conventional controllers, such as smaller control input, faster convergence speed and higher accuracy. We also apply the proposed method to control multiple fractional-order ship power systems, to demonstrate its practical engineering value.}, archive = {J_TASE}, author = {Jin Ke and Ying Li and Tao Xie}, doi = {10.1109/TASE.2025.3604426}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21975--21989}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time consensus of input-constrained networked fractional-order systems and its application to ship power systems}, volume = {22}, year = {2025}, }
• (2025). Fully distributed finite-time formation tracking of heterogeneous multi-agent systems under directed networks. TASE, 22, 21960--21974. (www)

  @article{ , abstract = {This paper considers the finite-time formation tracking of high-order heterogeneous multi-agent systems (HMASs), where the communication network among agents is directed and the leader’s input is unknown to all followers. An innovative fully distributed finite-time control strategy, independent of global topological information, is proposed. First, a fully distributed observer is constructed to estimate the leader’s state in finite time, despite leader’s unknown input. It is noted that an adaptive gain is wisely designed for the observer such that global topological information is no longer necessary. However, due to the asymmetry of the Laplacian matrix and the fact that the adaptive gain may not converge to its ideal value, the commonly used finite-time convergence criteria are not applicable to our case, and achieving the finite-time convergence of the observer becomes rather challenging. To overcome technical obstacles, a suitable Lyapunov function is specifically constructed, and then the desired finite-time convergence is established with the aid of a contradiction argument. Further, a disturbance observer is designed by utilizing the state transformation, which is able to estimate the composite disturbances consisting of both external disturbances and the leader’s unknown input. Finally, an observer-based formation controller is proposed to achieve the desired configuration. It is guaranteed that zero formation error, rather than just a neighbor of zero, can be realized in finite time through chattering-free control signals. Simulations are performed to verify the effectiveness of the proposed algorithm. Note to Practitioners—The formation tracking of HMASs is investigated in this paper, which has wide applications such as unmanned aerial vehicle (UAV)-unmanned ground vehicle (UGV) collaborative reconnaissance and multi-type UAV coordinated attack. Furthermore, the convergence time is very important for some urgent tasks, such as multi-UAV collaborative firefighting. Hence, the finite-time control approaches have been developed for multi-agent systems (MASs). However, most existing finite-time methods for general linear MASs require the global topological information, such as the eigenvalues of the Laplacian matrix or the number of agents, to determine the parameters. Obtaining such global information is difficult especially in large-scale networks. Therefore, the need for the global topological information reduces both the practicality of the algorithm and the scalability of the system. To address this problem, a fully distributed finite-time formation approach is proposed under directed networks, where the global topological information is successfully avoided through adaptive strategy. It is guaranteed that the formation error converges to zero in finite time, despite disturbances and the leader’s unknown input. In addition, instead of using a controller that generates high-frequency chattering to suppress uncertainties, our proposed robust controller produces a chattering-free output with the aid of a novel disturbance observer, thereby reducing actuator wear and enhancing the algorithm’s practicality.}, archive = {J_TASE}, author = {Binghe An and Huijin Fan and Bo Wang and Lei Liu}, doi = {10.1109/TASE.2025.3598150}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21960--21974}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully distributed finite-time formation tracking of heterogeneous multi-agent systems under directed networks}, volume = {22}, year = {2025}, }
• (2025). Event-triggered distributed MPC for cyber-physical systems: An adaptive dual-horizon mechanism. TASE, 22, 21948--21959. (www)

  @article{ , abstract = {This paper aims to propose a new event-triggered distributed model predictive control framework based on an adaptive dual-horizon (prediction and control horizon) mechanism for multi-agent cyber-physical systems (CPS) with additive disturbances. Firstly, the control horizon is introduced into the constrained optimal control problem (OCP) of multi-agent CPS to reduce the number of independent variables and improve the rapidity of the algorithm. Secondly, a periodic event-triggered mechanism with Zeno-free behavior is designed based on state error information, which can more efficiently reduce the frequency of solving OCP. Then, adaptive prediction horizon and control horizon contraction mechanisms are designed, effectively reducing the computational complexity of solving OCP for the controller at the triggering moment. In addition, sufficient conditions are provided through theoretical analysis to guarantee the recursive feasibility of algorithm and the stability of the closed-loop system. Finally, the effectiveness of the proposed algorithm is verified through simulation and experimentation example of networked multi-robot system, and the results showed that the algorithm can effectively reduce communication resource consumption and computational complexity in solving OCP while ensuring the expected cooperative control effect. Note to Practitioners—The feature of this paper is to propose a new distributed model predictive control method for control engineers of CPS, such as aircraft formation, vehicle formation, and multi-robot systems. The aim is to effectively reduce the online computational complexity and the impact of high communication rates on control performance. Specifically, we achieved rapid solution of OCP and efficient utilization of communication resources through time-varying prediction and control horizons, and lowering the sampling and communication frequency through event-triggered mechanisms. It should be emphasized that the proposed mechanisms may require the multicore computation unit to implement, as the computation of the OCP and the verification of the event triggered condition need to be performed simultaneously. To facilitate the application of the method, practical experiments on multi-robot systems have also been conducted, providing practical implementation guidelines.}, archive = {J_TASE}, author = {Ning He and Xiaofei Feng and Zhongxian Xu and Dongyuan Tian and Fuan Cheng and Huiping Li}, doi = {10.1109/TASE.2025.3617127}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21948--21959}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered distributed MPC for cyber-physical systems: An adaptive dual-horizon mechanism}, volume = {22}, year = {2025}, }
• (2025). Design and experimental evaluation of a novel robot-assisted surgical system for pelvic fracture closed reduction. TASE, 22, 21936--21947. (www)

  @article{ , abstract = {This paper designs a novel clinical-oriented robot-assisted surgical system (RASS) for pelvic fracture closed reduction, including three core parts: preoperative planning system, intraoperative navigation system and reduction robot. In preoperative planning system, a virtual reduction method based on the Statistical Shape Model (SSM) and pelvic partial surface data is adopted, providing personalized and detailed preoperative plans with high-efficiency and enhanced adaptability. Then, a serial-parallel hybrid reduction robot with robot autonomous and main-secondary reduction mode switching control strategy is designed, combined with the intuitive and real-time guidance of the intraoperative navigation system, achieving flexible and precise reduction operations. Additionally, clinical solution research based on the designed RASS is developed and evaluated by ten pelvic models and five animal experiments. In the pelvic model experiments, the root mean square error (RMSE) of the distance between the reduced and intact pelvic point cloud is 1.0mm, and the three points error of 10 cases are all within 2.0mm, achieving an excellent reduction effect. In the animal experiments, the four point-pairs distance decreases from $26.7\pm 1.04$ mm to $5.86\pm 0.82$ mm, and the mean distance between the two point clouds of the “Pig-081” before modeling and after reduction is 2.81mm, which further confirmed the effectiveness of the system, as well as the feasibility and safety of the clinical solution. The designed novel RASS provides a promising solution in the treatment of pelvic fractures, and the experimental results lay the foundation for further clinical application. Note to Practitioners—RASS is crucial for pelvic fracture closed reduction. However, the clinical application of RASS for pelvic fracture closed reduction still face some challenges, such as the poor adaptability of preoperative virtual reduction, the limited mechanical structure and control strategy of reduction robot, the lack of integration all technical modules of RASS into a unified clinical workflow. To address these challenges, this paper designs a novel clinical-oriented RASS, including a more adaptive preoperative planning system, an intuitive intraoperative navigation system, as well as a flexible and accurate serial-parallel hybrid reduction robot. Based on the designed RASS, a comprehensive solution for RASS-integrated clinical workflow is proposed and evaluated through pelvic model and animal experiments. Experimental results show that our designed RASS can effectively address these challenges and demonstrates excellent reduction effects. The system’s integration design and adaptive workflow also offer insights for advancing robot-assisted solutions in broader orthopedic and trauma surgery applications.}, archive = {J_TASE}, author = {Xiao Cheng and Wei Kou and Xinyu Chen and Dongya Li and Haixia Wang and Shaolong Kuang}, doi = {10.1109/TASE.2025.3615799}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21936--21947}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and experimental evaluation of a novel robot-assisted surgical system for pelvic fracture closed reduction}, volume = {22}, year = {2025}, }
• (2025). Resilient asynchronous sampled-data control for markov jump systems under DoS attacks: A discontinuous interval-dependent lyapunov functional. TASE, 22, 21925--21935. (www)

  @article{ , abstract = {This article investigates the security control problem of Markov jump systems (MJSs) under denial of service (DoS) attacks by constructing a class of discontinuous interval-dependent functional. Motivated by 1) the enhanced practicality and flexibility of multi-sensor sampled-data control with asynchronous packet arrivals and 2) the destabilizing impact of DoS-induced data loss, a resilient asynchronous sampled-data control (RASDC) scheme is designed. RASDC can send data once immediately after the attack, thereby alleviating the performance degradation caused by long-term no input. Furthermore, by utilizing the stability analysis idea of switched systems, a discontinuous interval-dependent functional is constructed, which combines the information of the resilient sampling interval and the attack interval. To overcome the analytical challenges posed by its discontinuities, the discrete-time Lyapunov stability theory, convex combination techniques, and some estimation methods are used to derive the sufficient condition for mean square asymptotic stability. In light of the stability criterion, a design algorithm for the secure controller is provided. Finally, simulation results demonstrate the effectiveness of the proposed approach. Note to Practitioners—In smart grid and power distribution systems, communication between remote sensors (e.g., voltage, current, and fault detectors) and controllers is often vulnerable to cyber threats such as DoS attacks. These attacks disrupt the timely of critical data, posing serious risks to system stability and performance. To address this challenge, a resilient asynchronous sampled-data control scheme is proposed. The scheme provides a practical solution to ensure stability in MJSs, even in the presence of data loss caused by DoS attacks. Theoretical tools such as discontinuous interval-dependent functionals and switched system stability analysis are employed to guide controller design, and the final outcome is an implementable control algorithm. The proposed method can be applied to enhance the security of industrial control systems, and it is well-suited for integration into modern power automation infrastructures to improve cyber-physical resilience.}, archive = {J_TASE}, author = {Lan Yao and Xia Huang and Zhen Wang and Min Xiao}, doi = {10.1109/TASE.2025.3615233}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21925--21935}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient asynchronous sampled-data control for markov jump systems under DoS attacks: A discontinuous interval-dependent lyapunov functional}, volume = {22}, year = {2025}, }
• (2025). Observer-based compensation control for nonlinear interconnected power systems with load disturbances and actuator faults. TASE, 22, 21913--21924. (www)

  @article{ , abstract = {In this article, in the occurrence of load disturbances and internal actuator faults, exact performance compensation problem for nonlinear interconnected power systems (NIPSs) integrated with wind farms is solved by presenting an observer-based sliding mode control strategy. Concretely, a dynamic event condition in the sensor channel is designed by virtue of triggered output error and a dynamic threshold. With the help of event outputs and neural networks (NNs) technology, a composite observer is proposed to perform state, disturbance, and fault estimations of the power system simultaneously. By utilizing the variable estimations, a novel sliding mode manifold is designed and an observer-based controller is derived directly. Intuitively, the compensation action is triggered once the unknown variables of the NIPSs are estimated, thereby the disturbance rejection and fault tolerance can be achieved. A prominent benefit of the presented strategy is that the estimation convergence ability of the observer is proved independently, then the relationship between the observer and controller for the NIPSs can be relaxed. The control ability is analyzed within a $H_{\infty } $ performance through the Lyapunov theory. Simulation results for the two area NIPSs and comparisons on the recent achievements of the NIPSs are made, the results show that more than ten percent improvements can be realized compared with the recent control methods, which in return demonstrate the compensation performance of the presented control strategy. Note to Practitioners—This study is motivated by achieving the safety compensation control for nonlinear interconnected power systems (NIPSs) with actuator faults and unknown load disturbances. With the improvement of automation, the safety compensation control scheme has been applied to many critical fields, such as large scale power system, reactor process, and so on. However, the measured value of the instrument may not be fully achieved, thus causing some variables cannot be obtained directly. Furthermore, modern industrial systems are inherently vulnerable to unexpected load variations and actuator faults, which could degrade control precision and operational integrity. Compounding these challenges, bandwidth-constrained communication architectures persistently impose critical bottlenecks on control signal transmission efficiency. In view of the unknown variables, the threat of actuator faults and disturbances for the NIPSs, an observer-event-based compensation control strategy is proposed to deal with the above challenges.}, archive = {J_TASE}, author = {Bin Guo and Songyi Dian and Ben Niu and Yuqi Zhu and Tao Zhao}, doi = {10.1109/TASE.2025.3615193}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21913--21924}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based compensation control for nonlinear interconnected power systems with load disturbances and actuator faults}, volume = {22}, year = {2025}, }
• (2025). STGN: A spatio-temporal graph network for real-time and generalizable trajectory planning. TASE, 22, 21897--21912. (www)

  @article{ , abstract = {In dynamic and unstructured environments, mobile robots need to generate safe and efficient trajectories in real time, which poses significant challenges due to the uncertainty of surrounding obstacles. To address this, this article presents a real-time obstacle avoidance trajectory planning method, built upon a spatio-temporal graph network that integrates temporal modeling with graph attention mechanisms. The proposed network captures both temporal dynamics and spatial structural dependencies in dynamic environments by integrating a temporal information module based on long short-term memory (LSTM) and a spatial module based on relational graph attention networks (RGAT). On the whole, the approach follows a two-phase pipeline. In the offline phase, a high-quality trajectory dataset is constructed to represent the heterogeneous state graph of the robot and surrounding obstacles. Then the dataset is used to train the spatio-temporal network, which learns to map environment-state graphs to optimal control commands. In the online phase, the trained network is deployed on the robot to perform real-time perception, decision-making, and control, forming a closed-loop trajectory optimization process. Extensive experiments in both simulated and real-world scenarios demonstrate that the proposed method achieves high-quality trajectory planning, robust obstacle avoidance, and fast generalization under multi-obstacle and sudden disturbance conditions, while maintaining low computational overhead. Note to Practitioners—This article addresses the generalization challenges commonly encountered in traditional supervised learning-based obstacle avoidance methods. We propose a trajectory planning framework that leverages a spatiotemporal graph network to model dynamic interactions between a robot and its surrounding obstacles. This approach enables robust and adaptable behavior in complex, changing environments by explicitly capturing both spatial and temporal dependencies. The system is implemented on a four-wheel steering (4WS) robot for experimental validation. However, its modular design ensures straightforward transferability to a wide range of mobility platforms. The proposed method requires only basic obstacle position information, readily available from standard onboard sensors such as LiDAR or radar, and does not depend on raw sensor inputs or semantic maps, making it highly suitable for real-world deployment. It can be easily integrated into existing perception–planning–control pipelines, and future extensions may focus on incorporating richer semantic information and expanding to more diverse obstacle types.}, archive = {J_TASE}, author = {Runjiao Bao and Yongkang Xu and Chenhao Wang and Tianwei Niu and Junzheng Wang and Shoukun Wang}, doi = {10.1109/TASE.2025.3614472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21897--21912}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {STGN: A spatio-temporal graph network for real-time and generalizable trajectory planning}, volume = {22}, year = {2025}, }
• (2025). Dual-layer-based active disturbance rejection decoupling control for lateral stability of power-decoupled agricultural mobile platforms under complex operating conditions. TASE, 22, 21880--21896. (www)

  @article{ , abstract = {Due to zero emissions of pollutant gases, rapid response capability, and multifunctionality, electric agricultural mobile platforms have gradually become a significant research focus in the field of facility agriculture, especially power-decoupled agricultural mobile platforms with excellent mobility and flexibility. In order to enhance the adaptability of mobile platforms to confined facility spaces and meet high standards of agricultural practices, a four-wheel power-decoupled agricultural mobile platform (FPD-AMP) with independent drive and independent steering is designed in this paper. Furthermore, a dual-layer-based active disturbance rejection decoupling control strategy is proposed to address the lateral instability encountered by FPD-AMPs under complex operating conditions, particularly in turning scenarios. Firstly, the dynamic characteristics of the FPD-AMP, considering both facility agricultural requirements and complex operating conditions, are analyzed. And the corresponding lateral stability state equations are established. Additionally, an S-curve trajectory experiment is conducted to fit the key parameters of the tire model. Then, to tackle the challenges posed by strong nonlinear coupling, time-varying parameters, and external disturbances under complex operating conditions, an upper-layer active disturbance rejection decoupling control method is proposed. This method not only satisfies the requirements for global lateral stability control but transmits clear drive force distribution commands to the lower-layer control system. In the lower-layer control, a torque distribution controller integrating the unscented Kalman filter algorithm is designed to address the multiple constraints under varying soil-road conditions and to further enhance the lateral stability margin. Finally, Experimental results demonstrate that the proposed strategy effectively enhances the lateral stability of the FPD-AMP. Note to Practitioners—In greenhouses, limited space and dense crop planting severely constrain turning areas, while traditional mobile platforms, with poor maneuverability and large turning radii, typically rely on discontinuous and complex forward-reverse maneuvers, significantly reducing efficiency. This paper proposes an FPD-AMP with independent drive and independent steering. Yet, the four wheels still need to undergo large steering angles to complete a continuous U-turn maneuver, which inevitably increases the risk of lateral instability. Furthermore, wide-ranging velocity variations for agronomic requirements, coupled with diverse farmland soil-road conditions, further exacerbate lateral instability during the U-turn maneuver of the FPD-AMP. To address the lateral instability issues, a dual-layer active disturbance rejection decoupling control strategy for lateral stability of the FPD-AMP is proposed. The upper-layer active disturbance rejection decoupling control consists of nonlinear dynamic model decoupling and the design of multi-channel active disturbance rejection controllers for lateral stability. The lower-layer torque allocation control, based on the unscented Kalman filter (UKF), involves farmland soil-road identification using a UKF observer and torque allocation optimization through quadratic programming. The experimental results demonstrate that the proposed control strategy effectively estimates the varying tire-road adhesion coefficients and significantly enhances the lateral stability of the FPD-AMP during U-turn maneuvers and plowing operations under variable farmland conditions.}, archive = {J_TASE}, author = {Yang Wang and Xiaoyong Zhu and Yudan Cai and Lei Xu and Yue Shen and Lizhang Xu and Wen-Hua Chen}, doi = {10.1109/TASE.2025.3615903}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21880--21896}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-layer-based active disturbance rejection decoupling control for lateral stability of power-decoupled agricultural mobile platforms under complex operating conditions}, volume = {22}, year = {2025}, }
• (2025). Design and control of a novel multi-mode aerial ground robot with variable configuration. TASE, 22, 21864--21879. (www)

  @article{ , abstract = {The mobile robots possess immense application potential in planetary exploration, field investigation and other related fields. Adopting suitable movement modes in multiplex task scenarios and flexible modes transformations significantly enhance the flexibility and mobility efficiency of robots. Therefore, a novel multi-modal mobile robot with various ground and aerial movement modes is proposed. Notably, different modes transformation and aerial manipulation can be smoothly performed during aerial maneuvers, thereby improving the deployment efficiency of robots. Firstly, the proposed multi-modal robot achieves high-integration and high-mobility efficiency in all modes through the iterative optimization design. Subsequently, a comprehensive hybrid dynamic model of the robot is established. Based on the dynamic model, the leg motion trajectory during the aerial transformation is planned by optimizing the coupling torque, minimizing the motion’s interference on the flight system. Additionally, controllers for various movement modes are designed, including a novel flight controller based on the trajectory linearization control (TLC) method with an extended state observer (ESO), compensating for the time-varying inertial parameters of the robot and coupling or external disturbances, improving the agility and stability of the system in aerial. Finally, the practicability and effectiveness of robot’s multi-modal mobility, modes transformations and aerial manipulation are validated through simulations and real-world experiments. Note to Practitioners—Mobile robots are playing an increasingly significant role in industrial production, as demonstrated by their use in inspection and transportation. However, industrial environments often feature three-dimensional obstacles and uneven surfaces, which place high demands on robot adaptability. Although multi-modal mobile robots can enhance environmental adaptability, current solutions are limited by inflexible mode transitions and reduced functionality (e.g., manipulation). In response, we propose a novel multi-modal mobile robot that integrates both ground and aerial modes. To mitigate the adverse effects of dynamic variations and coupling force disturbances, comprehensive system dynamics model is established. A trajectory optimization-based approach combined with a novel nonlinear flight controller is then employed to compensate for these disturbances, thereby enabling seamless aerial transitions and agile aerial manipulation through leg-arm reuse.}, archive = {J_TASE}, author = {Kun Xu and Haoyuan Liu and Pengfei Lu and Jiawei Chen and Tao Zhang and Xilun Ding}, doi = {10.1109/TASE.2025.3615470}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21864--21879}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and control of a novel multi-mode aerial ground robot with variable configuration}, volume = {22}, year = {2025}, }
• (2025). Securing grid-connected packed E-cell multilevel inverter: A LSTM-AE approach to hybrid attack mitigation. TASE, 22, 21853--21863. (www)

  @article{ , abstract = {The deployment of open communication infrastructure into power systems has drawn much attention due to its significant benefits, such as real-time monitoring, diagnostics, and regulatory purposes. But utilizing such technologies poses security challenges to the cyber-physical power systems (CPPS), which can highly degrade their operation. In this paper, a defense mechanism is adopted to tackle the hybrid attacks, including the Denial-of-Service (DoS) and false data injection (FDI) attacks in the grid-connected multilevel inverters from a systematic point of view. The proposed protection mechanism for the grid-connected multilevel inverter is realized in two stages. $(i)$ A Long-Short Term Memory based on autoencoder (LSTM-AE) is developed to detect DoS attacks, and an event-trigger mechanism based on Lyapunov theory is implemented to eliminate the effect of false data. $(ii)$ A sliding mode observer is adopted to recognize FDI threats, where the false data is eliminated by injecting the negative value of the identified false data. A prototype of a grid-connected nine-level Packed E-Cell (PEC9) topology as a targeted multilevel inverter is constructed to experimentally validate the feasibility of the proposed cyber resilience scheme for CPPS in microgrid applications. Note to Practitioners—The motivation of this work comes from the issue that gird-connected multilevel inverters, which can realize the large-scale application of sustainable generation units, are susceptible to cyber threats. The cyber threats will introduce security problems to the cyber-physical power systems (CPPS) with the photovoltaic (PV) modules. False data injection (FDI) and denial-of-service (DoS) attacks, as the most common cyber-attacks, can seriously compromise CPPSs’ performance. In this regard, the current work develops a two-stage defense mechanism to tackle cyber-attacks. The proposed cyber resilient framework is designed to identify and mitigate hybrid attacks, including DoS and FDI attacks. In particular, the Long-Short Term Memory based on autoencoder (LSTM-AE) is utilized to identify the DoS attack, while the FDI attack is detected by the sliding mode observer (SMO). An event-triggered mechanism is also developed in the proposed defense algorithm to block the signal falsified by the DoS attack and submit the signal predicted by LSTM-AE to the system’s controller. The SMO estimates the FDI disruption to the system and injects it into the system measurement signal to eliminate the FDI attack’s impact on the system dynamic performance. For this approach, the critical challenge is to develop the proposed protection mechanism in practical applications. To address this difficulty, the experimental examinations are carried out by building a prototype of the PEC9 inverter.}, archive = {J_TASE}, author = {Soroush Oshnoei and Meysam Gheisarnejad and Mohammad Sharifzadeh and Eric Laurendeau and Kamal Al-Haddad}, doi = {10.1109/TASE.2025.3614757}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21853--21863}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Securing grid-connected packed E-cell multilevel inverter: A LSTM-AE approach to hybrid attack mitigation}, volume = {22}, year = {2025}, }
• (2025). Estimation and detection of intermittent faults for nonlinear systems disturbed by noises with uncertain covariances. TASE, 22, 21843--21852. (www)

  @article{ , abstract = {In this paper, the problems of intermittent fault (IF) estimation and detection are investigated for stochastic nonlinear systems disturbed by noises with uncertain covariances. The research methodology begins with the transformation of conventional nonlinear systems into descriptor systems through state and fault augmentation. For IF estimation, an innovative moving horizon estimator is introduced by utilizing the variational Bayesian technique. This approach enables the simultaneous and iterative estimation of both IFs and noise covariance matrices through the strategic selection of appropriate conjugate prior distributions. Furthermore, a rigorous analysis is conducted on the statistical properties of the estimation outcomes, leading to the development of an IF detection framework based on Hotelling’s $T^{2}$ statistic. To validate the effectiveness of the proposed methodology, extensive experimental evaluations are performed on a rotary steerable drilling tool system, demonstrating the superior performance of our algorithm in practical applications. Note to Practitioners—In recent decades, intermittent faults (IFs) have emerged as a critical safety challenge across multiple industrial sectors, with particular prominence in aerospace systems and electronic infrastructure. The precise detection of appearance and disappearance time of IFs constitutes a crucial requirement for ensuring industrial safety. Conventional approaches dependent on truncated residuals frequently exhibit significant limitations in practical implementations, primarily attributable to their insufficient capacity in dealing with nonlinear dynamic behaviors and statistically ambiguous noise profiles. To address these challenges, an adaptive moving horizon estimation framework is designed for nonlinear systems with uncertain noise covariance matrices. Leveraging the estimation results, a novel detection methodology is developed to accurately identify the appearance and disappearance of IFs. Experimental validation through comparative case studies demonstrates the operational viability and detection efficacy of the proposed methodology, providing industry practitioners with a reliable tool for enhancing operational reliability in IF-affected systems.}, archive = {J_TASE}, author = {Li Sheng and Yifan Liu and Ming Gao and Donghua Zhou}, doi = {10.1109/TASE.2025.3614648}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21843--21852}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Estimation and detection of intermittent faults for nonlinear systems disturbed by noises with uncertain covariances}, volume = {22}, year = {2025}, }
• (2025). Safe trajectory generation for nonholonomic multi-robot systems: A compensation-based MPC approach. TASE, 22, 21831--21842. (www)

  @article{ , abstract = {This article investigates the collaborative trajectory generation problem for nonholonomic multi-robot systems using a compensation-based nonlinear model predictive control (MPC), where the system is subject to unknown uncertainties. A dynamic uncertainty estimator is designed for each robot to estimate the discrepancy between the predictive model and the actual system, enabling adaptive model corrections that enhance the robustness and adaptability of the MPC. By incorporating control Barrier function constraints, physical constraints, and stability constraints, the nonlinear MPC generates feasible, collision-free trajectories. These trajectories are further optimized using piecewise Bézier curves, yielding smoother and more efficient paths. Additionally, a dynamic safety-stability gain is introduced, allowing the MPC to adaptively balance safety and stability based on the system state and obstacle positions. The theoretical results are validated through simulations and experiments, demonstrating the effectiveness of the proposed approach. Note to Practitioners—The motivation of this article is to address the collaborative trajectory generation problem for multi-robot systems in environments with unknown uncertainties. Existing nonlinear MPC methods have two major limitations: 1) some schemes struggle to handle dynamic model uncertainties effectively, and 2) there may be adverse interactions between safety and stability components. To overcome these limitations, we propose a compensation-based nonlinear MPC framework that incorporates a dynamic uncertainty estimator for each robot. The estimator continuously measures the discrepancy between the predictive model and the actual system, adaptively updating the model to improve control accuracy. By incorporating control Barrier functions, physical, and stability constraints, the method generates feasible, collision-free trajectories. These trajectories are optimized using piecewise Bézier curves for smoother, more efficient paths. A dynamic safety-stability balancer adjusts constraints based on the system’s state and detected obstacles, relaxing stability when safety is critical and prioritizing progress when less urgent, thereby ensuring both safety and efficiency.}, archive = {J_TASE}, author = {Zhixu Du and Hao Zhang and Peiyu Cui and Zhuping Wang and Huaicheng Yan}, doi = {10.1109/TASE.2025.3613403}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21831--21842}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe trajectory generation for nonholonomic multi-robot systems: A compensation-based MPC approach}, volume = {22}, year = {2025}, }
• (2025). Second-order sliding mode control of flying-wing aircraft based on feedforward neural networks. TASE, 22, 21811--21830. (www)

  @article{ , abstract = {The flying-wing aircraft control problem is a major concern. In this paper, a new control strategy is introduced. First, a Feedforward neural network (FNN) modeling is introduced. Then, a second-order sliding mode control is applied, with the parameters generated from Deep Deterministic Policy Gradient (DDPG) reinforcement learning. To study the disturbance rejection performance, wind disturbance is applied to the aircraft using a deep neural network as an disturbance observer for different types of winds. Finally, All three simulations: Simulink, Software In The Loop, and Hardware In the Loop are applied to show the effectiveness of the proposed strategy. The simulation results show that the proposed method demonstrates good robustness in various conditions. Note to Practitioners—This paper is motivated by the traditional linearized flying-wing aircraft controller with the effectiveness of the FNN and reinforcement learning on UAV applications. The controller can be designed for various situations without changing the parameters by modeling the aircraft through the FNNs. The theoretical framework proposed in this paper combines the multiple-input multiple-output (MIMO) sliding mode strategy with the reinforcement learning with higher accuracy modeling. This method reduces the settling time, overshoot and steady error. More realistic effects should be considered for deployment into real aircraft. The parameters of the neural networks should also be adjusted in real applications.The FNN networks and DDPG have low computational footprints and are readily deployable on embedded systems like Pixhawk. The DNN observer may require model compression for the smallest processors. While the current framework requires per-aircraft training to achieve optimal performance, this process is conducted offline. The resulting controller gains are then fixed for reliable real-time operation, providing a clear pathway for implementation on specific UAV platforms.}, archive = {J_TASE}, author = {Yuecheng Song and Zhenbao Liu and Junwei Han and Jinbiao Yuan and Wen Zhao and Qingqing Dang}, doi = {10.1109/TASE.2025.3613383}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21811--21830}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Second-order sliding mode control of flying-wing aircraft based on feedforward neural networks}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered-based quantized consensus for fractional-order MASs with asymmetric time-varying state constraints. TASE, 22, 21798--21810. (www)

  @article{ , abstract = {In this article, we contribute to dynamic event-triggered-based leader-follower quantized consensus design for fractional-order (FO) nonlinear multi-agent systems (MASs) under asymmetric time-varying state constraints. Firstly, we propose a novel dynamic event-triggered mechanism (ETM) to fully utilize the channel resources between the controller and actuator. Subsequently, by means of the characteristics of quantization nonlinearities and the architecture of FO nonlinear MASs, the influence of quantized inputs is eliminated. In the consensus protocol design procedure, the radial basis function neural networks (RBF-NNs) technology is employed to dynamically estimate the uncertain functions existing in the system. Then, by employing a bivariate FO derivative lemma and some convex time-varying barrier Lyapunov functions (BLFs) with two variables, a dynamic event-triggered-based leader-follower quantized consensus protocol is constructed. Exhaustive theoretical analysis manifests that the tracking errors of the FO nonlinear MASs converge to a small region, all signals in the FO nonlinear MASs are bounded, and the asymmetric time-varying state constraints can be achieved regardless of the presence of the quantized inputs and dynamic event-triggered communication. Finally, a chaotic Duffing FO nonlinear MASs is employed to demonstrate the efficacy of the developed consensus protocol. Note to Practitioners—Quantized input technology refers to the situation that the input signals of a system are restricted to a finite set of discrete values. Compared to continuous input, discrete input signals can be processed and implemented more efficiently, but they can also cause distortion of the control signals. Meanwhile, in industrial production processes, system state constraints may change due to many factors such as raw material quality, equipment failures, and environmental conditions. These factors can introduce uncertainties and disturbances that affect the system’s behavior and performance. Therefore, in the procedure of control design for industrial systems, it is necessary to detect the change of state constraints and adjust the control parameters to ensure the safe operation of the production process. Based on this, we attempt to construct a leader-follower quantized consensus protocol specifically designed for FO nonlinear MASs. This protocol aims not only to eliminate the influence of quantized inputs, which can cause errors and distortions, but also to constrain all system states within predetermined ranges.}, archive = {J_TASE}, author = {Xiaoan Wang and Xiaobing Nie and Jinde Cao and Liang Hua}, doi = {10.1109/TASE.2025.3612186}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21798--21810}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered-based quantized consensus for fractional-order MASs with asymmetric time-varying state constraints}, volume = {22}, year = {2025}, }
• (2025). Quasi-projective synchronization of discrete-time fractional-order T-S fuzzy complex-valued neural networks with hybrid delays. TASE, 22, 21784--21797. (www)

  @article{ , abstract = {This article delves into quasi-projective synchronization (Q-PS) problem for a class of discrete-time fractional-order T-S fuzzy complex-valued neural networks (DFTSFCNNs) with leakage and time-varying delays. First of all, according to the theory of discrete fractional calculus and properties of Mittag-Leffler function, an innovative property of discrete Mittag-Leffler function is strictly proved, and then an inequality for dealing with mixed time delays is rigorously derived. Next, by utilizing Caputo fractional difference theory and combining with property and the inequality offered in this article, some easily verifiable Q-PS criteria are derived under complex-valued fuzzy linear controller. Eventually, a numerical example is presented to demonstrate availability of the derived results. Note to Practitioners—Leakage delay and time-varying delay are core factors influencing the dynamic behavior of complex systems, as they directly impact synchronization control, controllability, and stability. Ignoring them may lead to model distortion or control failure. Therefore, this paper proposes a new inequality for handling mixed delays in the quasi-projective synchronization of DFTSFCNNs with leakage delay and time-varying delay. Based on this inequality, delay-dependent criteria for the quasi-projective synchronization are established, which breaks the traditional limitation of using delayed controllers to handle time-delay terms. In addition, numerical examples are used to verify the effectiveness of the quasi-projective synchronization criteria.}, archive = {J_TASE}, author = {Rong Chen and Hong-Li Li and Long Zhang and Tingwen Huang and Jinde Cao}, doi = {10.1109/TASE.2025.3610123}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21784--21797}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quasi-projective synchronization of discrete-time fractional-order T-S fuzzy complex-valued neural networks with hybrid delays}, volume = {22}, year = {2025}, }
• (2025). Modular acoustic graph SLAM for underwater monitoring with autonomous underwater vehicles. TASE, 22, 21770--21783. (www)

  @article{ , abstract = {This work was developed under the need for an acoustic localization system to monitor marine protected areas (MPAs) with the help of autonomous underwater vehicles (AUVs). Although the use of acoustic signals for underwater localization has been previously studied, most of the solutions rely on filter-based optimization, which is prone to linearization problems in long-term applications. Instead, we implemented a Modular Acoustic Graph Simultaneous Localization and Mapping (SLAM) algorithm that, using a factor graph framework, tracks acoustic beacons with either ranges or bearings. In addition, we developed several novel methods, like a delayed-position update for ultra-short baseline (USBL) position factor integration process, an initialization algorithm for acoustic landmarks, and the creation of a new 3D bearing factor that combines two angles. After developing the algorithm, field experiments were carried out in different areas on the coast of Catalonia. Besides the localization, some monitoring tasks were also tested, such as visual mapping of localized landmarks or optical transmission of data with seafloor stations, which helped validate the accuracy of the acoustic localization system. The results of such experiments are presented and discussed. Note to Practitioners—Autonomous robots can benefit underwater monitoring of deep areas. Our work presents a direct application of such robots, which use acoustic signals to localize underwater elements, like monitoring stations in the seafloor equipped with acoustic modems, and interact with them. The navigation is computed using a graph-based approach and can use either range or bearing information for the acoustic localization. The experiments in real scenarios are executed using preplanned trajectories, and the results prove that this system is able to localize acoustic targets within low uncertainty. The next step would be the implementation of path-planning strategies to optimize the localization process.}, archive = {J_TASE}, author = {Marta Real and Pau Vial and Roger Pi and Narcís Palomeras and Marc Carreras}, doi = {10.1109/TASE.2025.3615829}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21770--21783}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modular acoustic graph SLAM for underwater monitoring with autonomous underwater vehicles}, volume = {22}, year = {2025}, }
• (2025). Cross-modal commonality graph matching frame: A fault diagnosis method for multimode process. TASE, 22, 21760--21769. (www)

  @article{ , abstract = {Fault diagnosis in complex industrial systems, such as large-scale chemical plants and advanced manufacturing lines, is critically challenged by data acquired in multiple and varying operating modes. These modal shifts, driven by different production demands and process parameters, often obscure fault signatures and undermine the reliability of conventional diagnostic models. Despite these operational variations, it is observed that variables within such multimode processes often exhibit consistent correlations. Consequently, the same fault can manifest itself as a shared propagation characteristic across different modes. To leverage these invariant features for robust fault identification, this paper proposes a novel Cross-Modal Commonality Graph Matching (CMCGM) framework. Our approach first extracts the shared causal structures of a specific fault from data across multiple modalities to construct a modality-independent “commonality fault dictionary,” which captures the essential fault signature. By transforming the diagnosis task into a local subgraph matching problem against this dictionary, the CMCGM method achieves robust and accurate fault identification while circumventing the computational costs of online feature propagation. The effectiveness and superior performance of the proposed framework are validated through extensive experiments on the Tennessee-Eastman (TE) process, a widely-recognized benchmark for complex chemical systems. Note to Practitioners—This work addresses the challenge of diagnosing faults in industrial processes where data comes from multiple working modes. In such systems, faults often propagate in similar ways across different modalities, but traditional methods struggle to leverage these shared patterns effectively. Our solution, the Cross-Modal Commonality Graph Matching (CMCGM) framework, identifies and matches these common fault signatures across modalities, enabling faster and more reliable fault diagnosis without requiring extensive computational resources. The method works by analyzing how faults propagate across different data sources, extracting their shared structural features, and building a reusable fault reference library. Instead of treating each data stream separately, it compares faults in a way that reduces redundant computations. This makes the diagnosis process more efficient while maintaining reliability. One current limitation is that it requires sufficient historical fault records to build an effective reference model. Future improvements could focus on making the system more responsive to newly emerging faults and integrating it with real-time monitoring systems. Potential applications include predictive maintenance in factories, power plants, and other complex industrial systems where early and accurate fault detection is critical.}, archive = {J_TASE}, author = {Shuai Tan and Zhiyun Chen and Qingchao Jiang and Weimin Zhong}, doi = {10.1109/TASE.2025.3613749}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21760--21769}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cross-modal commonality graph matching frame: A fault diagnosis method for multimode process}, volume = {22}, year = {2025}, }
• (2025). Adaptive resilient event-triggered control for T-S fuzzy multi-area power system with pumped storage hydropower under deception attacks. TASE, 22, 21748--21759. (www)

  @article{ , abstract = {This paper investigates the load frequency control problem for nonlinear multi-area power system that incorporates pumped storage hydropower. First, a unified Takagi-Sugeno fuzzy model is developed to represent the nonlinear dynamics of power exchange progress in pumped storage hydropower, including uncertainties. Subsequently, a fuzzy logic algorithm is applied by incorporating both system error and its derivative to dynamically adjust the trigger threshold, an adaptive resilient event-triggered mechanism is proposed to reduce communication burden while improving control performance. Furthermore, a Lyapunov functional is constructed to ensure that the system is stable and strictly $\left ({{\mathcal {D}_{1},\mathcal {D}_{2},\mathcal {D}_{3}}}\right)-\lambda -dissipative$ . Finally, a three-area power system is used to show the effectiveness of the proposed mechanism. Note to Practitioners—With the increasing application of pumped storage technology in power system, the problem of frequency regulation is investigated in this paper. Takagi-Sugeno fuzzy model is employed due to its linear substructure that enables direct application of Lyapunov-based methods for stability verification. Meanwhile, an adaptive resilient event-triggered control mechanism with fuzzy logic rules is proposed to balance data transmission requirements with control performance. Qualitative and quantitative analyses verify the proposed control strategy.}, archive = {J_TASE}, author = {Congzhe Mu and Lin Gan and Tao Deng and Zhenzhen Zhang and Yaonan Shan and Hao Chen and Shouming Zhong}, doi = {10.1109/TASE.2025.3613303}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21748--21759}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive resilient event-triggered control for T-S fuzzy multi-area power system with pumped storage hydropower under deception attacks}, volume = {22}, year = {2025}, }
• (2025). Joint optimization of order sequencing and temporary rack shelving for separated bin-picking systems. TASE, 22, 21728--21747. (www)

  @article{ , abstract = {As the adoption of warehouse automation continues its upward trajectory, “parts-to-picker” order picking systems, epitomized by the Robotic Mobile Fulfillment System (RMFS), are increasingly deployed across diverse enterprises. Considering that current automation systems are economically unfriendly to the traditional warehouse transformation, and other shortcomings such as low space utilization, a separated bin-picking system has been developed and adopted in companies. The key improvement compared to RMFS is the adoption of a novel Automated Guided Vehicle (AGV) equipped with temporary racks, which can pick material bins from traditional racks through a grabbing device without redesigning and alterations of the fixed racks. However, research on related decision-making and optimization is still lacking. To fill this gap, we explore the joint optimization of order sequencing and temporary rack shelving and formulate a mixed integer programming model, taking into account the SKU-based (Stock Keeping Unit) workload balancing among multiple picking stations. To solve the model, we propose an interactive order driven heuristic combined neighborhood search algorithm. Based on the real data from an auto-parts distribution center, the case study is conducted to demonstrate the effectiveness of the proposed method. The proposed method can reduce the number of rack movements by 15.39% and improve the utilization rate of temporary racks by 27.49% compared to the two-step method typically used in practice. Sensitivity analysis is also performed to provide valuable managerial insights for the operational improvement of the separated bin-picking system. Note to Practitioners—This paper is inspired by an emerging “parts-to-picker” order picking system, which can more effectively reduce retrofit costs of warehouses and improve efficiency compared to currently used Automated Storage and Retrieval System (AS/RS) and Robotic Mobile Fulfillment System (RMFS). However, although the system has been applied in practice, there lacks research on related decision-making and optimization at the operational level. In this paper, we investigate the joint optimization problem of order sequencing and temporary rack shelving, considering the SKU-based workload balancing among multiple picking stations. A mixed integer programming model is formulated. Then, to deal with large-scale cases, an interactive order driven heuristic combined neighborhood search algorithm is proposed to improve the computation efficiency. Practitioners can implement the proposed method and algorithm for routine order processing of automated warehouses that use separated bin-picking systems. Based on the sensitivity analysis, we also present some managerial insights, which can provide guidance for practitioners when implementing the operational improvement of the system. For instance, practitioners could prioritize increasing the capacity of temporary racks over that of picking stations if they would like to improve the system’s efficiency.}, archive = {J_TASE}, author = {Meimei Zheng and Zhenqi Xu and Edward Huang and Tangbin Xia and Kan Wu}, doi = {10.1109/TASE.2025.3613008}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21728--21747}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Joint optimization of order sequencing and temporary rack shelving for separated bin-picking systems}, volume = {22}, year = {2025}, }
• (2025). Actuator–Fault–Tolerant adaptive tracking control of delayed fuzzy systems using reinforcement learning. TASE, 22, 21715--21727. (www)

  @article{ , abstract = {In nonlinear systems, monitoring control behavior, fault occurrence, and latency factor continue to be major obstacles. Traditional control models frequently handle edge–case situations inaccurately and are unable to adjust to dynamic changes. Reinforcement learning control is a novel intelligent framework that incorporates sophisticated modules for nonlinear models in order to overcome these constraints. First, an innovative approach to solving the adaptive tracking algorithm for a type of T–S fuzzy plant is presented in this research. Furthermore, the authors examine time–varying delay with the actuator failure for complex systems that are generally represented by the T–S fuzzy plant. Second, it has been demonstrated that the chosen performance index solves the delayed algebraic Riccati formula, which can be resolved by scheme of repetition methods, when the resultant plant with the tracking signal is constructed under the decoupling delayed T–S fuzzy framework. Third, the coupled delayed algebraic Riccati formulas are then solved using a Reinforcement Learning (RL) technique that makes use of complex model dynamics knowledge. Furthermore, the fundamental RL iteration technique’s convergence is demonstrated. Lastly, the power sector and a F–16 airplane flight examples are provided to demonstrate how the online repetition method outperforms the offline one in terms of tracking accuracy and effectiveness. Note to Practitioners—This work bridges the gap between theoretical reinforcement learning techniques and practical control applications in dynamic and uncertain environments, providing a scalable, adaptive solution for high–reliability systems. Contemporary industrial systems like those in the power sector and manned aircraft (F–16)–increasingly face challenges involving nonlinear factors, actuator faults, and communication delays. This research introduces a reinforcement learning (RL)–based control approach tailored for systems modeling using Takagi–Sugeno (T–S) fuzzy logic, which is well–suited for capturing nonlinear actions. The proposed method integrates an adaptive tracking mechanism with delay compensation and fault dealing with making it highly relevant to real–world scenarios involving uncertain and time–varying circumstances.}, archive = {J_TASE}, author = {Muhammad Shamrooz Aslam and Affaq Qamar and Hazrat Bilal and Athanasios V. Vasilakos}, doi = {10.1109/TASE.2025.3611686}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21715--21727}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Actuator–Fault–Tolerant adaptive tracking control of delayed fuzzy systems using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Distributed affine formation maneuvering for networked marine surface vessels under radom packet loss: A knowledge-based localization solution. TASE, 22, 21701--21714. (www)

  @article{ , abstract = {Formation maneuvering (FM) of networked marine surface vehicles (MSVs) encounters challenges in preserving system performance under unreliable inter-vehicle interactions and ensuring real-time operation flexibility in dynamic environments. To solve these issues, a knowledge-based affine formation maneuver (AFM) scheme is proposed, which is composed of a virtual leader trajectory generator (VLTG), distributed hybrid localizer (DHL) and knowledge-based event-triggered communication mechanism (KETCM). Specifically, the VLTG generates the real-time guidance trajectories for networked MSVs with intuitively defined affine transformations. Based on this, the DHL allows local vehicles to realize real-time resilient AFM trajectory observations under distributed neighboring interactions. Notably, to mitigate the performance deterioration caused by packet losses, a pseudo feedback neighbor estimator (PFNE) is integrated into the DHL, which provides accurate neighbor data estimations between successive packet receptions. Based on these components, KETCM schedules efficient communication triggers while ensuring DHL performance, which is derived from evaluations of multiple historical communication knowledge to avoid unnecessary communications under random packet losses. Compared to existing methods, the key advantage of the proposed method lies in the simultaneous achievement of real-time maneuverability, AFM performance, and communication efficiency under unreliable neighboring communications. Finally, theoretical analysis and semi-physical experiments validate the effectiveness of the proposed solution. Note to Practitioners—This article is motivated to achieve the real-time FM control for networked MSVs under unreliable wireless network communications, which are characterized by limited network bandwidth and random packet loss. The practical significance of this work contains three parts. On one hand, we propose a novel VLTG and DHL framework to enable multiple MSVs real-time maneuvering their formation pattern in complex water terrains. Under this protocol, operators only need to plan a single formation center velocity and geometric commands for three virtual leaders, thereby generating complex AFM trajectories for networked MSVs with arbitrary cluster scale. On the other hand, inter-vehicle packet transmission loss results in degraded AFM performance and inefficient utilization of limited communication resources. To prevent these risks, a novel PFNE and KETCM are established. Specifically, PFNE ensures accurate estimations of neighbor information, thereby guaranteeing target formation localization accuracy, which is significant in ensuring operation safety. Meanwhile, KETCM optimizes the utilization of limited network resources in the presence of random packet losses.}, archive = {J_TASE}, author = {Bin Zhou and Zihan Xia and Jiasheng Ren and Bing Huang}, doi = {10.1109/TASE.2025.3609933}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21701--21714}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed affine formation maneuvering for networked marine surface vessels under radom packet loss: A knowledge-based localization solution}, volume = {22}, year = {2025}, }
• (2025). Overshoot suppression of hydraulic actuation based on the integrating flow compensation: Applied to lift tracking control of an engine variable valve actuator. TASE, 22, 21686--21700. (www)

  @article{ , abstract = {Overshoot is prevalent in hydraulic actuators due to the inertia of motion. Especially for high-speed position servo hydraulic actuators, overshooting is almost unavoidable in order to achieve high response position tracking. This paper addresses overshoot suppression in high-frequency electro-hydraulic systems (HESs). Traditional overshoot mitigation strategies—over-damping and reduced feedback coefficients—compromise dynamic performance, failing to meet high-frequency response requirements. And nonlinear factors, including friction, dead zones, and flow variations, are identified as critical contributors to overshoot. To suppress overshoot, a novel backstepping controller with integrated flow compensation is proposed and applied to hydraulic engine variable valve actuator. In the controller, the nonlinear friction force, system delay, dead zone, etc. are synthesized into the controller firstly. And the feedback gains are kept at a low level to reduce overshoot. Secondly, proportional valve response error is converted as flow error, and the integrating flow error is added to the controller. Finally, some comparative experiments are implemented on a hydraulic engine variable valve actuator (VVA). The experiment results show that the overshoot and tracking error of VVA are both reduced significantly with the guaranteed dynamic performance. When the desired maximum valve lift is 12mm and 10mm, the valve lift overshoot at opening stage is reduced from 1.2mm, 0.72mm to 0.03mm, 0.01mm respectively. And the overshoot at closing stage is reduced from 0.45mm, 0.11mm to 0.02mm, 0.02mm respectively. Note to Practitioners—High-frequency electro-hydraulic valve systems face critical overshoot-induced failures, including piston-valve collisions and valve seat damage, which increase maintenance costs by 18-25% in typical engine production lines. Traditional PID-based methods require trade-offs between overshoot suppression ( $\ge 5$ % residual) and response speed (+ 35% settling time). Our integrated backstepping controller resolves this dilemma through three field-ready steps: (1) Compensate nonlinearities (dead zones, asymmetric friction) via embedded LuGre modeling and inverse functions; (2) Convert valve displacement errors to hydraulic flow errors in real-time, preserving dynamics; (3) Implement delay compensation through automated lift retuning. Validated on VVA prototype, under the conditions of 12mm and 10mm maximum lift, the valve lift overshoot is restrained most significantly, and the overshoot is less than 0.05mm in both the opening and closing stages.}, archive = {J_TASE}, author = {Jian Li and Yong Lu and Jiarui He and Pengyong Xie and Yawen Li}, doi = {10.1109/TASE.2025.3608156}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21686--21700}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Overshoot suppression of hydraulic actuation based on the integrating flow compensation: Applied to lift tracking control of an engine variable valve actuator}, volume = {22}, year = {2025}, }
• (2025). Dual-agent-based robotic ultrasound path planning and interaction control in external-vision-independent environments. TASE, 22, 21674--21685. (www)

  @article{ , abstract = {Ultrasound imaging has emerged as a crucial tool for the diagnosis and navigation of spinal diseases. However, high-quality image acquisition heavily relies on experienced sonographers, which restricts its further popularization. In this paper, a robotic system designed for automated spinal ultrasound scanning is proposed. Drawing inspiration from the spinal anatomy and the actions of seasoned sonographers, the system integrates both a deep learning agent and a reinforcement learning agent to collaboratively guide the adjustment of the ultrasound probe in external-vision-independent environments, relying on real-time ultrasound images and contact force. Then, a hybrid force-to-velocity control framework is proposed to ensure proper ultrasound coupling during the scanning process. Experimental results on a phantom and human participants demonstrated that this system can accurately track spinal features (mean error: less than 1 mm) and maintain normal probe orientation (out-of-plane angular error: $1.61~\pm ~1.1^{\circ }$ , in-plane angular error: $1.27~\pm ~0.9^{\circ }$ ), resulting in high-quality and reproducible ultrasound images. Overall, our system shows great potential for clinical applications. Note to Practitioners—This paper is motivated by the increasing needs of human-robot interaction in medical applications, with a specific emphasis on robotic ultrasound imaging. Clinical sonographers suffer from repetitive workload during the diagnostic process, highlighting the significance of automated scanning solutions. In this work, we propose a modular control framework for ultrasound probe positioning that supports a multi-modal autonomous ultrasound scanning system. The system operates independently of external optics or prior geometric knowledge of the scanning object. Comprehensive experimental results demonstrate that the system can effectively cover the region of interest of the spine, facilitating high-quality ultrasound image acquisition and related disease assessment. This advancement is expected to enhance the efficiency of human-robot interaction in healthcare settings and holds promising clinical applications. Furthermore, our research offers valuable insights for the implementation of robotic ultrasound scanning systems applicable to other human tissues.}, archive = {J_TASE}, author = {Xinye Wang and Zhiyuan He and Peng Chen and Zhe Wang and Tao Sun}, doi = {10.1109/TASE.2025.3614996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21674--21685}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-agent-based robotic ultrasound path planning and interaction control in external-vision-independent environments}, volume = {22}, year = {2025}, }
• (2025). Vision-based autonomous robotic arc welding: State-of-the-art review and perspectives. TASE, 22, 21651--21673. (www)

  @article{ , abstract = {Robotic welding is an essential technology in industrial production. Various sensors are integrated into robotic welding systems to enhance welding efficiency and quality. Due to their advantages of non-contact measurement, high information capacity, and high accuracy, vision sensors play an increasingly important role in robotic autonomous welding. Therefore, the latest vision-based autonomous robotic arc welding technologies are comprehensively summarized, providing valuable reference and support for researchers. First, the research progress of visual sensing technology for welding robots is outlined, and its advantages and disadvantages are analyzed. Subsequently, the key technologies of robotic autonomous welding are introduced, covering weld type identification, initial point guidance, welding path generation, welding parameter planning, feature point extraction, seam tracking, welding posture adjustment, and welding quality control. Finally, the limitations existing in current research are summarized, and the future research directions of autonomous robotic welding are prospected. Note to Practitioners—Most current programming methods for welding robots rely on manual teaching and offline programming, limiting their adaptability to small batches and diverse categories in production. With the development of artificial intelligence and deep learning, vision-based technologies are propelling welding robots toward greater intelligence and autonomy. Autonomous welding robot technology is continuously advancing in areas such as weld type identification, initial point guidance, welding path generation, welding parameter planning, feature point extraction, seam tracking, welding posture adjustment, and welding quality control. This paper systematically reviews relevant research to date and outlines the future development directions of autonomous robotic arc welding.}, archive = {J_TASE}, author = {Yunkai Ma and Junfeng Fan and Jun Hou and Yichen Fu and Rui Tao and Shuo Wang and Min Tan and Fengshui Jing}, doi = {10.1109/TASE.2025.3614992}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21651--21673}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Vision-based autonomous robotic arc welding: State-of-the-art review and perspectives}, volume = {22}, year = {2025}, }
• (2025). Enhancing battery electric vehicles safety optimization with adaptive MOEA/D based on dynamic grid and multiple dominance. TASE, 22, 21635--21650. (www)

  @article{ , abstract = {With the rapid development of electric vehicles (EVs), the safety of battery systems has become an increasing concern. However, the application of multi-objective optimization algorithms in the safety domain of EVs is rare, particularly in optimizing complex, non-linear multi-objective problems (MOPs) such as crashworthiness and thermal management. This work proposes an improved multi-objective evolutionary algorithm based on decomposition (MOEA/D), which introduces a dynamic grid system for building stability matrices to determine whether the current population reaches a stable convergence state with less computational demands and designs a novel sparsity level evaluation to measure the weight vectors. Additionally, we propose a multiple dominance to address the Pareto dominance limitation. These aim to optimize the key issues in EV safety. To further demonstrate the performance of our proposed algorithm, we compare it with 10 state-of-the-art algorithms on 26 benchmark problems. The experimental results based on multiple performance metrics show that the proposed algorithm have outstanding performance. Note to Practitioners—Ensuring the safety of the battery system in EVs under collision conditions is important. This paper proposes an improved MOEA/D, aiming to optimize the maximum energy absorption of battery protection materials, enhance crash force efficiency, and improve the temperature difference of the battery system coolant. In proposed improved MOEA/D, the dynamic grid system and sparsity level evaluation ensure the acquisition of feasible solutions that align with users’ diverse preferences. Additionally, the multiple dominance strategy further enhances the quality of the solutions, leading to more superior optimization results. Experimental results demonstrate that the proposed algorithm shows outstanding optimization performance. We also further analyze the influence of each DAH structural parameter on performance, which enables users to make a trade-off according to their specific application requirements. Future research will explore the interdisciplinary integration of battery protection optimization to achieve more refined and precise optimization.}, archive = {J_TASE}, author = {Mingran Li and Li Huang and Hua Han and Chunyuan Wang}, doi = {10.1109/TASE.2025.3614750}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21635--21650}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing battery electric vehicles safety optimization with adaptive MOEA/D based on dynamic grid and multiple dominance}, volume = {22}, year = {2025}, }
• (2025). A platform-centric framework for intelligent parking traffic prediction and resource optimization in shared AVPC systems. TASE, 22, 21621--21634. (www)

  @article{ , abstract = {To address the challenges posed by uncertainties in the parking behaviors of private owners and temporary users, as well as the complexities involved in integrating shared parking with Electric Vehicle (EV)-charging, this paper proposes a novel platform-centric intelligent framework for shared Automated Valet Parking and Charging (AVPC) systems. The framework leverages Long Short-Term Memory (LSTM) and Deep Reinforcement Learning (DRL) to optimize both parking traffic prediction and resource allocation, respectively. Specifically, to mitigate uncertainties in vehicle parking and EV-charging demand and supply, we utilize an LSTM prediction model to forecast the average day-ahead arrival times, departure times, and service pricing for parking space owners (O-users) and temporary users (R-users). Then, we design an improved Proximal Policy Optimization (PPO)-based algorithm with large warm-up training steps that integrates the LSTM prediction results with real-time supply and demand information from O-users and R-users to determine optimal shared AVPC resource allocation. Extensive simulations using real-world parking datasets demonstrate that the LSTM model achieves an average Mean Absolute Percentage Error (MAPE) of 1.71% and 0.08% for O-users and R-users’ parking traffic predictions, respectively. Additionally, the proposed LSTM-PPO-based approach improves platform profit and parking resource utilization by at least 9% and 15%, respectively, compared with state-of-the-art. Note to Practitioners—Intelligent Transportation Systems (ITS) are increasingly facing critical challenges in urban planning, especially in managing limited parking and EV-charging resources for autonomous vehicles and shared AVPC systems. This paper addresses these challenges by proposing an intelligent framework that combines LSTM-based prediction and DRL-based resource allocation for dynamic parking demand and EV-charging coordination in shared AVPC systems. Unlike traditional rule-based methods, our solution adapts to real-time parking space supply-demand fluctuations while optimizing for both operational efficiency and user convenience, from the perspective of the parking management platform. Validated using real-world datasets, the framework is readily deployable in smart cities, logistics hubs, and commercial complexes to alleviate congestion, maximize parking revenue, and enable the seamless integration of EV-charging infrastructure. Its automated decision-making capability minimizes operational overhead, offering a scalable solution for modern urban parking and mobility challenges. Practitioners managing smart parking systems, mobility platforms, and infrastructures can leverage this framework to automate parking and EV-charging resource allocation, optimize parking resource pricing, and improve parking resource utilization with minimal manual oversight.}, archive = {J_TASE}, author = {Gordon Owusu Boateng and Huang Xia and Haonan Si and Xiansheng Guo and Cheng Chen and Nirwan Ansari}, doi = {10.1109/TASE.2025.3614684}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21621--21634}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A platform-centric framework for intelligent parking traffic prediction and resource optimization in shared AVPC systems}, volume = {22}, year = {2025}, }
• (2025). Linear-parameterization-based model free adaptive predictive control. TASE, 22, 21609--21620. (www)

  @article{ , abstract = {A novel data-driven predictive control scheme is proposed for unknown nonlinear discrete-time systems. The control increment vector is linearly parameterized through the dynamic linearization (DL) on the ideal controllers along the prediction horizon, leading to a predictive control law. The control gain matrix is adaptively tuned through the recursive least square method. The distinctive predictive control scheme is purely data-driven, because the predictive control law does not rely on underlying system dynamics and the control gain matrix is tuned by using only the input/output data of the controlled system. The asymptotic convergence of the designed scheme is rigorously guaranteed under reasonable conditions. One numerical example and one welding example are simulated to verify the validity and applicability of the designed scheme.Note to Practitioners—This paper was motivated by the industrial problem of sending what control input to the machine to follow the given signals N steps in advance. This problem is applied in various fields such as autonomous driving vehicles and welding process. Existing methods require the system dynamics to be known, which is challenging. This paper suggests a new control method by using only the data generated from the industrial process without knowing the system dynamics. In this paper, we mathematically characterize the mechanics of the new method and its reliability. Preliminary mathematical simulation and welding process simulation suggest that this method is feasible but this method has not been incorporated with disturbances and constraints. In future research, we will consider the disturbances and constraints in the discussed method.}, archive = {J_TASE}, author = {Juanping Zhu and Qiuyan Wei and Xian Yu and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3613545}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21609--21620}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Linear-parameterization-based model free adaptive predictive control}, volume = {22}, year = {2025}, }
• (2025). Koopman-based fractional predefined-time control for wearable exoskeleton system. TASE, 22, 21595--21608. (www)

  @article{ , abstract = {In this paper, an innovative fractional predefined-time control scheme based on the Koopman operator is proposed for the wearable exoskeleton system (WES). Leveraging the nonlinear mapping of the Koopman operator and the prior information of the exoskeleton system, a prior Koopman model (PIKOM) is devised. As a data-driven model, the PIKOM can be constructed through measuring input/output (I/O) data, avoiding complex dynamic modeling. Besides, the integer-order PIKOM is broadened into a fractional-order form, and the tanh function is adopted to handle the problem of saturation constraint. The fractional predefined-time controller is meticulously designed to enhance the tracking performance of the WES. The tracking error converges to a narrow range near the origin within a predefined time. The experimental results verify the effectiveness and superiority of the proposed approach. Compared with state-of-the-art (SOTA) methods (Schemes 1-4), the proposed method achieves improvements of 12.2%, 18.2%, 8.8%, and 20.7%, respectively. Note to Practitioners—In recent years, the trajectory tracking of the WES has garnered significant and escalating research interests. Owing to the complicated structure of the WES, dynamic model-based control methods generally fail to realize satisfactory control performance. A PIKOM that only embraces I/O data is constructed to avoid this problem. Meanwhile, the PIKOM is converted into a fractional form because of the outstanding advantages of fractional-order control (FOC). With the rapid development of automatic control technology, there is a growing demand for the convergence rate of the WES. The asymptotic FOC method is no longer applicable. In this paper, a PIKOM-based fractional predefined-time control is presented to achieve trajectory tracking. In addition to inheriting the merits of fixed-time control strategies, this method exhibits a characteristic that the upper bound of convergence time is solely determined by a single parameter. The control performance of the proposed scheme is demonstrated by experimental results.}, archive = {J_TASE}, author = {Xianlei Zhang and Yan Zhang and Qing Hu and Zongyu Han and Yixin Yang and Xilong Yu}, doi = {10.1109/TASE.2025.3612902}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21595--21608}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Koopman-based fractional predefined-time control for wearable exoskeleton system}, volume = {22}, year = {2025}, }
• (2025). ADP: Adaptive diffusion policy energizes robots thinking in both learning and practice. TASE, 22, 21585--21594. (www)

  @article{ , abstract = {Adaptive control policies for robots often require balancing generalization from large offline datasets with efficient adaptation to specific deployment conditions. In this paper, we propose Adaptive Diffusion Policy (ADP), a two-stage framework that integrates temporal-aware diffusion models with parameter-efficient LoRA adaptation. First, in the learning stage, ADP imitates and generates actions based on image and video signals from a meager amount of expert demonstrations, considering both spatial and temporal information. This component contrasts with most existing works, which focus solely on spatial information. Second, in the practice stage, ADP incorporates a low-rank adaptation module into the policy, subsequently training it using residual reinforcement learning with minimal environment interaction. Experiments conducted on Meta-World benchmark demonstrate the efficiency of each ADP component and the superiority of ADP over representative baseline methods. Note to Practitioners—This work introduces Adaptive Diffusion Policy (ADP), a two-stage visuomotor framework that first learns from just a few image-and-video demonstrations by modeling both spatial and temporal cues, then rapidly refines its behavior via a lightweight low-rank adapter and residual reinforcement learning. The ADP enables swift skill acquisition on new tasks with minimal expert data and limited environment trials, making it ideal for industrial or household robots where extensive data collection is impractical. To apply ADP, collect a small set of demonstration clips, train the diffusion-based policy offline, and deploy the adapter online for in situ fine-tuning. The proposed Meta-World results show the ADP’s consistent gains over standard imitation and residual RL baselines, which is very easy for practitioners in multiple real-world robot scenarios.}, archive = {J_TASE}, author = {Dechao Chen and Zhengwen Chen and Xiangyan Zheng and Weiling Xu and Chencong Ma and Chentao Mao}, doi = {10.1109/TASE.2025.3612396}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21585--21594}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ADP: Adaptive diffusion policy energizes robots thinking in both learning and practice}, volume = {22}, year = {2025}, }
• (2025). Continuum-model-based security control of multi-tiered re-entrant manufacturing networks under cyber-attacks. TASE, 22, 21571--21584. (www)

  @article{ , abstract = {In this paper, the security control problem for a class of multi-tiered re-entrant manufacturing networks (RMNs) under cyber-attacks is investigated. First, a linear hyperbolic partial differential equation (PDE) continuum model is employed to model the dynamics of each single manufacturing line node and the overall RMN system is then characterized by a dual-layer coupling structure consisting of a production line network and a workshop network, both of which adhere to the global mass conservation law. Second, to mitigate the adverse effects of malicious cyber-attacks on RMNs, a node-dependent control approach and an edge-dependent control approach are developed, both of which guarantee the global exponential stability of the multi-tiered RMNs under cyber-attacks. Numerical simulations demonstrate the effectiveness of the proposed network architecture and control schemes. Note to Practitioners—The motivation of this research stems from practical challenges faced by contemporary manufacturing systems characterized by intricate, multi-stage processes, such as semiconductor wafer fabrication, multiphase chemical reactors, and automotive assembly lines. Given the increasing complexity and vulnerability of these RMNs to cyber threats, there is an urgent industrial need to develop robust and scalable network architectures alongside effective security control strategies. In contrast to traditional discrete-event models, the proposed continuum-model-based approach uses hyperbolic PDEs to provide a scalable and efficient framework that captures complex cross-scale interactions between production lines and workshops while rigorously adhering to mass conservation principles. Furthermore, this research introduces node-dependent and edge-dependent control strategies specifically designed to ensure global exponential stability against cyber-attacks. The practicality and versatility of the developed methodologies make them particularly suitable for implementation in a broad spectrum of real-world manufacturing scenarios exhibiting significant re-entrant characteristics.}, archive = {J_TASE}, author = {Chenguang Liu and Qing Gao and Michael V. Basin and Jinhu Lü}, doi = {10.1109/TASE.2025.3611938}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21571--21584}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Continuum-model-based security control of multi-tiered re-entrant manufacturing networks under cyber-attacks}, volume = {22}, year = {2025}, }
• (2025). Computationally efficient bayesian model predictive control for 4-D flight trajectory tracking under windy conditions. TASE, 22, 21552--21570. (www)

  @article{ , abstract = {Accurate 4-D trajectory tracking can improve trajectory predictability and further enhance the efficiency of air traffic management (ATM). However, the tracking accuracy is inevitably affected by uncertainties arising from wind. To maintain tracking performance, a novel Bayesian model predictive control (MPC) framework is proposed, within which a Bayesian recurrent neural network-based probabilistic wind prediction module is integrated into the control design process, yielding a stochastic optimal control problem (OCP) with predefined trajectory violation probabilities. On the other hand, computational cost from solving the complex OCP hinders the real-time implementation of the control strategy on resource-limited flight management systems. To alleviate this burden, state-related inequality constraints are projected from the state space into the input space. Rigorous theoretical guarantees for the proposed Bayesian MPC scheme are provided, including proof of asymptotic stability and derivation of necessary conditions to ensure recursive feasibility. The effectiveness of the proposed method is validated through simulation tests on 4-D flight trajectory tracking problems using reliable datasets. Note to Practitioners—Trajectory-based operations in ATM typically require high-precision 4-D trajectory tracking of aircraft. However, conventional MPC struggles to handle precise trajectory tracking tasks under external uncertainties from wind. This work provides new insights into MPC by integrating a Bayesian recurrent neural network that predicts the unknown wind, thereby formulating a Bayesian MPC scheme which can remove the reliance on prior knowledge of the wind. To enhance computational efficiency, a constraint projection technique is employed to address the constrained optimization problem arising from Bayesian MPC. The practicality of this approach is further improved by pre-adjusting the boundaries of the approximated constraints using empirical data. A theoretical analysis is conducted to verify the recursive feasibility of the proposed Bayesian MPC scheme. Simulation tests are performed on Boeing aircraft by using reliable datasets to validate the effectiveness of the proposed method.}, archive = {J_TASE}, author = {Yuhang Wang and Kaiquan Cai and Yanbo Zhu and Jingyao Zhang and Deyuan Meng}, doi = {10.1109/TASE.2025.3609449}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21552--21570}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Computationally efficient bayesian model predictive control for 4-D flight trajectory tracking under windy conditions}, volume = {22}, year = {2025}, }
• (2025). A learning-based two-stage bidirectional packing framework for 3D packing problems. TASE, 22, 21541--21551. (www)

  @article{ , abstract = {The increasing demands of modern logistics have driven the need for the development of efficient packing methods capable of addressing the 3D Packing Problem (3D-PP). While Deep Reinforcement Learning (DRL) has emerged as a promising solution, the conventional three-stage scheme used in existing DRL-based methods still faces challenges, particularly in coordinating the behaviors of its constituent sub-networks and managing the large action space for item placement on instances involving large-sized bins. This work proposes a two-stage scheme to integrate item index and orientation selections into a single sub-stage, thereby simplifying behavioral coordination. To mitigate the issue of excessive memory usage associated with the selection integration, a Set Transformer with Induced Set Attention Block (ISAB) is employed to encode the rotated item state, thus keeping relatively light computation. Additionally, we propose a bidirectional packing method that compresses the placement action space while encouraging the agent to explore reasonable placement positions. Experimental results demonstrate that the Two-Stage Bidirectional Packing (TS-BP) framework, formed by the above components, improves space utilization of 2.8%~4.6% on high-difficulty packing instances compared to current state-of-the-art methods. The code is available at https://github.com/Ashenone511/Two-Stage-Bidirectional-Packing-Framework Note to Practitioners—This work proposes an effective learning-based Two-Stage Bidirectional Packing (TS-BP) framework to address the 3D Packing Problem (3D-PP), responding to the critical need to maximize warehouse storage density and reduce manufacturing logistics costs. The two-stage scheme equipped with the Induced Set Attention Block (ISAB) simplifies behavioral coordination among sub-networks while maintaining a lightweight computation. The bidirectional packing method reduces the action space size while ensuring the exploration of reasonable placement positions. The above components help to mitigate the shortcomings existing in current technologies. Simulation results demonstrate that the TS-BP framework outperforms state-of-the-art methods on standard and high-difficulty datasets. This work establishes a methodological framework for research on real-world packing solutions and contributes to improving warehouse storage density and reducing manufacturing logistics costs.}, archive = {J_TASE}, author = {Hao Yin and Fan Chen and Hongjie He}, doi = {10.1109/TASE.2025.3607410}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21541--21551}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A learning-based two-stage bidirectional packing framework for 3D packing problems}, volume = {22}, year = {2025}, }
• (2025). Privacy-preserving distributed fault diagnosis for multiple wind farms using a federated feature fusion method. TASE, 22, 21525--21540. (www)

  @article{ , abstract = {While data-driven methods have gained prominence in wind turbine fault diagnosis, their effectiveness is increasingly constrained by the proliferation of data silos. This phenomenon primarily arises from the reluctance toward data sharing, a trend exacerbated by intensifying commercial competition and mounting privacy concerns. Such limitations fundamentally challenge centralized data processing paradigms in this field. Federated learning provides a viable technical solution to mitigate data silos while preserving data sovereignty. Nevertheless, two critical challenges that substantially hinder practical implementation remain unresolved: 1) pronounced data heterogeneity among geographically distributed wind farms, and 2) prohibitive communication overhead stemming from sluggish convergence rates in distributed optimization. To bridge these gaps, we propose the Federated Feature Fusion (Fed-FF), a novel cross-farm fault diagnosis framework that synergistically integrates feature-level federation with convergence acceleration. Furthermore, the theoretical guarantees of the proposed method are studied in this paper, where the convergence risk bounds for both convex and non-convex settings are derived. The proposed method is then validated against three real-world datasets collected in Yunnan Province, Jiangsu Province and Shanghai, China. The experimental results show that the proposed method achieves an average accuracy of 93.83%, outperforming several state-of-the-art fault diagnosis methods, while communication costs are reduced by up to 94.71%. Note to Practitioners—As competition intensifies and data privacy regulations tighten, wind farm operators are increasingly withholding operational data, creating persistent data silos. Compounding this challenge, the inherent heterogeneity of SCADA measurements across geographically dispersed farms exacerbates communication overhead in conventional federated learning. To overcome these limitations, we propose a novel approach that ensures data privacy while significantly accelerating convergence and reducing communication costs, thereby lowering operational expenses for wind farms. Supported by rigorous theoretical guarantees, our method enhances both reliability and practical applicability. As a distributed fault diagnosis framework, it facilitates efficient collaboration and statistical information sharing among multiple wind farms without exchanging raw data, enabling precise fault localization. Beyond wind farms, this method is also applicable to diverse domains, including healthcare, intelligent connected vehicles, photovoltaic power plants, and smart manufacturing, offering a versatile and privacy-preserving solution for distributed learning.}, archive = {J_TASE}, author = {Zhijun Wang and Yanting Li and Zijun Zhang and Ershun Pan}, doi = {10.1109/TASE.2025.3613954}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21525--21540}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Privacy-preserving distributed fault diagnosis for multiple wind farms using a federated feature fusion method}, volume = {22}, year = {2025}, }
• (2025). Process optimization of multi-stage continuous production system based on feature fusion modeling. TASE, 22, 21512--21524. (www)

  @article{ , abstract = {Modeling and process optimization of Multi-stage Continuous Production System (MCPS) is an important research topic in the field of intelligent manufacturing today. However, due to the inherent properties of MCPS, existing researches face difficulties in 1) coupling optimization across multiple stages under diverse constraints and objectives, and 2) accurately mapping controllable process variables to critical production outputs. In this work, we propose a novel multi-objective optimization method for MCPS based on feature fusion modeling, with main innovations including: 1) For MCPS modeling, we propose a Transformer network structure with a parallel dual-branch attention mechanism for linking process variables to production outputs. A multi-stage feature fusion prediction model based on DenseNet is developed, which not only improves prediction accuracy but also provides intermediate stage prediction results. 2) For process optimization of MCPS, we design a multi-constraint and multi-objective optimization model. Subsequently, we propose a dynamic multi-objective optimization algorithm framework to enhance the performance of the algorithm and improve the quality of solutions. Furthermore, we conducted experiments with a real Coke-Chem integrated production dataset, and the results show that our predictive model achieves an average MAE of 0.11, MSE of 0.04, and RMSE of 0.20, outperforming state-of-the-art methods and boosts solution coverage (up to 80%) and hypervolume (up to $2.5\times $ ) when integrated with classical algorithms like NSGA-II and GDE3. Note to Practitioners—The PO-MCPS commonly found in modern industrial production is a complex issue, especially when some intermediate stages also produce end products. For example, optimizing a single stage may deteriorate other stages, causing a global loss; simply increasing production volume may also increase raw material costs and reduce overall profitability. The strategy proposed in this work can help practitioners build predictive models similar to Coke-Chem integrated production, thereby enabling the prediction of outputs at various stages of the production and supporting the overall optimization of process variables. This strategy has a high value for engineering applications.}, archive = {J_TASE}, author = {Xiaojie Li and Runlong Yu and Lei Chen and Shengjun Liu and Enhong Chen}, doi = {10.1109/TASE.2025.3612822}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21512--21524}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Process optimization of multi-stage continuous production system based on feature fusion modeling}, volume = {22}, year = {2025}, }
• (2025). Decentralized control of crop growth conditions in vertical farms under dynamic energy markets. TASE, 22, 21498--21511. (www)

  @article{ , abstract = {The growing global population and the increasing scarcity of arable land highlight the urgent need for reliable and efficient food production systems. With their controlled environments, vertical farms (VFs) offer a promising solution for sustainable food security. Nevertheless, their high energy demands call for innovative approaches to optimize energy consumption while maintaining optimal growing conditions. This paper introduces a novel control-oriented model for VFs, capturing the interactions between crop growth conditions and energy consumption. To address the high energy demand of VFs, the model is integrated into a dynamic energy market characterized by time-varying energy prices and a demand response scheme, which includes a discrete reward to encourage flexible energy consumption. Then, centralized and decentralized receding horizon control approaches are proposed to minimize the energy cost of the VF while ensuring optimal crop growth. Experimental evaluations on real systems of varying scales demonstrate the effectiveness of the proposed approaches in reducing costs and ensuring sustainable agricultural practices. Note to Practitioners–This work addresses a growing challenge in operating vertical farms: reducing energy costs while maintaining optimal conditions for crop growth. We introduce a control system that helps vertical farms schedule energy-intensive activities to take advantage of dynamic electricity prices or incentives from grid operators. In particular, we focus on a binary reward structure reflecting real-world demand response programs, where financial incentives are granted only if strict consumption targets are fully met. The approach relies on forecasting and optimization techniques already compatible with standard industrial automation systems. Two control systems are proposed: a centralized controller that manages the entire facility from a single decision point and a decentralized version that allows each unit (e.g., a room or a growing tray) to make decisions independently. The decentralized version offers better scalability and can more easily adapt to farm layout or crop type changes. This framework could also be applied to greenhouses, food storage systems, or other indoor environments with high energy demand.}, archive = {J_TASE}, author = {Kirill Zhukovskii and Paolo Scarabaggio and Polina Ovsiannikova and Pranay Jhunjhunwala and Raffaele Carli and Mariagrazia Dotoli and Valeriy Vyatkin}, doi = {10.1109/TASE.2025.3609694}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21498--21511}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Decentralized control of crop growth conditions in vertical farms under dynamic energy markets}, volume = {22}, year = {2025}, }
• (2025). Multi-agents cooperative target tracking under physical attacks with environment-aware dynamic constraints. TASE, 22, 21484--21497. (www)

  @article{ , abstract = {In this work, we consider a cooperative target tracking problem by multi-agents in a 3D space, under disruptions from multiple physical attackers, where the target and physical attackers are intelligent, that is, they adjust velocities based on relative coordinates with agents and agents’ velocities. Due to the presence of physical attackers, some safety and performance constraint requirements cannot be merely constant or time-varying. Specifically, safety constraint requirements on inter-agent distances and performance constraint requirements on formation tracking errors are environment-aware and dynamic, which depend on distances between agents and physical attackers. We propose a neural network-based adaptive cooperative control framework for the agents, which incorporates a universal barrier function to handle safety and performance constraints. We show that formation tracking errors are uniformly ultimately bounded, while all safety and performance constraints are met. A comparative simulation study further illustrates the efficacy of the proposed framework. Note to Practitioners—Cooperative target tracking control for multi-agent systems has gained significant attention due to its real-world applications. In practice, target-tracking tasks are often carried out in complex environments, where multiple physical attackers can fly nearby to disrupt the agents. Moreover, the target and attackers often possess a certain level of intelligence, making the target more difficult to track, complicating an already difficult problem. Thus, most control frameworks in the literature, which assume targets and attackers have constant or time-varying velocities, are not effective to address real-world operational complexities. Additionally, in complex environments, certain constraint requirements need to be environment-aware and dynamic, instead of being mere constants or time-varying. To address these practical challenges, this work proposes a neural network-based adaptive cooperative control framework that integrates a universal barrier function to ensure that all safety and performance constraints are met. Finally, we consider two scenarios in the simulation study to further validate the efficacy of the proposed control framework.}, archive = {J_TASE}, author = {Zhongjun Hu and Jesse B. Hoagg and Xu Jin}, doi = {10.1109/TASE.2025.3607874}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21484--21497}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-agents cooperative target tracking under physical attacks with environment-aware dynamic constraints}, volume = {22}, year = {2025}, }
• (2025). Rapidly antibiotic susceptibility prediction via deep learning from bacterial fluorescence microscopy images. TASE, 22, 21469--21483. (www)

  @article{ , abstract = {The incidence of multi-drug resistant bacterial species is rapidly increasing. To avoid antibiotic misuse and further exacerbation of this health crisis, clinicians should obtain a swift and accurate diagnosis of bacterial resistance. Existing laboratory-based antibiotic susceptibility testing (AST) methods are slow and laborious. Our study puts forward an efficient end-to-end image classification network using deep learning for predicting antibiotic resistance from fluorescence microscopy images of bacteria. Our resulting PAS-Net model integrates two parallel branches: the convolution branch (C-branch), employing ConvNeXt, and the Transformer branch (T-branch), utilizing Vision Transformer (ViT). To bridge these branches effectively, we introduce the feature interaction unit (FIU), which facilitates the integration of local features generated by the C-branch and global representations by the T-branch interactively. Additionally, we present a novel attention mechanism which reduces computational costs while maintaining the global representation capability. Experimental results, conducted on a fluorescence image dataset and public open datasets, suggest the promising potential of our method for rapidly diagnosing antibiotic susceptibility in bacteria and achieving high prediction performance. The code is publicly available at: https://github.com/Td270/PAS-Net Note to Practitioners—The rise of multi-drug resistant bacteria poses a significant challenge to public health, requiring faster and more accurate methods to predict bacterial antibiotic susceptibility. Current laboratory methods are slow and often impractical for time-sensitive clinical decisions. This work introduces PAS-Net, a deep learning framework designed to predict antibiotic resistance directly from fluorescence microscopy images of bacteria. Practitioners can apply this method in clinical and laboratory settings to significantly reduce diagnosis time, potentially enabling real-time decision-making. By combining convolutional and Transformer-based architectures, PAS-Net captures both detailed local patterns and global structural information, leading to improved predictive accuracy. This method is particularly suited for labs with access to fluorescence imaging equipment but limited by the slow throughput of traditional assays. However, the approach requires high-quality imaging and computational resources, which could limit its immediate application in low-resource settings. Future research should focus on adapting the model for broader imaging modalities and reducing its dependency on computational infrastructure. Beyond clinical diagnostics, the approach could be extended to other microscopy-based applications, such as microbial ecology or industrial bioprocess monitoring, where rapid and accurate phenotypic assessments are required.}, archive = {J_TASE}, author = {Baiying Lei and Tengda Zhang and Jiashu Li and Junlong Qu and Wei Xiong and Kaiwei Yu and Liang Yang}, doi = {10.1109/TASE.2025.3607484}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21469--21483}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rapidly antibiotic susceptibility prediction via deep learning from bacterial fluorescence microscopy images}, volume = {22}, year = {2025}, }
• (2025). Reactive and safety-aware path replanning for collaborative applications. TASE, 22, 21456--21468. (www)

  @article{ , abstract = {This paper addresses motion replanning in human-robot collaborative scenarios, with an emphasis on reactivity and safety-compliant efficiency. While existing human-aware motion planners perform well in structured environments, they often struggle with unpredictable human behavior. This can result in safety measures that hinder the robot’s performance and overall throughput. This study combines reactive path replanning and a safety-aware cost function, enabling the robot to adapt its path to the changes in the scene in real-time. This solution reduces the execution time and trajectory slowdowns while ensuring safety. Simulations and real-world experiments show the method’s effectiveness compared to standard human-robot cooperation approaches, with efficiency enhancements of up to 60%. Note to Practitioners—The ISO/TS 15066 outlines collaborative operations designed to uphold worker safety in human-robot collaboration. While these safety measures lead to robot deceleration to prevent collisions, they can also cause inefficiencies due to robot inactivity. This work introduces a dynamic real-time adaptation of the robot’s path, aiming to minimize the execution time while considering safety interventions. In contrast to methods relying on pre-computed trajectories, this approach modifies the robot’s path in real-time based on the continuous monitoring of the human state, thereby decreasing safety interventions and downtime, especially in dynamic environments.}, archive = {J_TASE}, author = {Cesare Tonola and Marco Faroni and Saeed Abdolshah and Mazin Hamad and Sami Haddadin and Nicola Pedrocchi and Manuel Beschi}, doi = {10.1109/TASE.2025.3607037}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21456--21468}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reactive and safety-aware path replanning for collaborative applications}, volume = {22}, year = {2025}, }
• (2025). Selective densification for rapid motion planning in high dimensions with narrow passages. TASE, 22, 21443--21455. (www)

  @article{ , abstract = {Sampling-based algorithms are widely used for motion planning in high-dimensional configuration spaces. However, due to low sampling efficiency, their performance often diminishes in complex configuration spaces with narrow corridors. Existing approaches address this issue using handcrafted or learned heuristics to guide sampling toward useful regions. Unfortunately, these strategies often lack generalizability to various problems or require extensive prior training. In this paper, we propose a simple yet efficient sampling-based planning framework along with its bidirectional version that overcomes these issues by integrating different levels of planning granularity. Our approach probes configuration spaces with uniform random samples at varying resolutions and explores these multi-resolution samples online with a bias towards sparse samples when traveling large free configuration spaces. By seamlessly transitioning between sparse and dense samples, our approach can navigate complex configuration spaces while maintaining planning speed and completeness. The simulation results demonstrate that our approach outperforms several state-of-the-art sampling-based planners in $\mathbb {SE}(2)$ , $\mathbb {SE}(3)$ , and $\mathbb {R}^{14}$ with challenging terrains. Furthermore, experiments conducted with the Franka Emika Panda robot operating in a constrained workspace provide additional evidence of the superiority of the proposed method. Note to Practitioners—Motion planning for robots with a high degree of freedom (DoF), such as humanoid robots and robotic arms, presents significant challenges due to the vast configuration space involved. Each joint adds a dimension, leading to an exponentially larger and more complex configuration space that is difficult to navigate. To address these challenges, sampling-based techniques are commonly employed. These methods sample points in the configuration space to construct a graph or tree of potential motion paths. However, their computational complexity increases dramatically with environmental complexity. Obstacles can create narrow passages in the configuration space, requiring a large number of samples to ensure the connectivity of the underlying graph or tree and, therefore, the existence of solutions. This paper proposes a novel sampling-based planner along with its bidirectional version that reduces search complexity by simultaneously exploring samples at various resolutions. The proposed planners efficiently find solutions in large free configuration spaces by utilizing sparse random samples. When navigating narrow passages, they transition to denser random samples to locate viable paths. By integrating samples of different resolutions, the proposed planners effectively navigate high-dimensional configuration spaces with narrow passages. Simulation results demonstrate that our approach outperforms several state-of-the-art sampling-based planners across various scenarios. The proposed planners, with their adaptability, can be easily implemented as a plug-and-play solution for diverse robotic systems, enabling them to perform autonomous navigation tasks in unstructured environments.}, archive = {J_TASE}, author = {Lu Huang and Lingxiao Meng and Xingjian Jing and Jiankun Wang}, doi = {10.1109/TASE.2025.3606845}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21443--21455}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Selective densification for rapid motion planning in high dimensions with narrow passages}, volume = {22}, year = {2025}, }
• (2025). Hierarchical defense decision-making for MSV-guided USVs under complex obstacle constraints. TASE, 22, 21430--21442. (www)

  @article{ , abstract = {Given the unique cognitive advantages of humans in complex maritime missions, this study proposes an intelligent collaborative decision-making framework for a hybrid system consisting of one crewed surface vehicle and multiple uncrewed surface vehicles (MSV-USV). We introduce a regularized polygonal obstacle modeling method based on geometric feature simplification to reduce computational complexity. A distributed task negotiation framework with dynamic priority-based message-passing enables adaptive task decomposition through multi-agent interactions. For maneuver optimization, a tripartite game model captures multidirectional strategy interactions while ensuring MSV safety requirements. Monte Carlo simulations validate the algorithm’s stability and reliability, and prototype tests demonstrate engineering feasibility. Note to Practitioners—With the increasing application of uncrewed systems in maritime missions, effectively integrating the cognitive advantages of human operators with the computational capabilities of uncrewed systems has become a key challenge in enhancing mission performance. The proposed distributed collaborative decision-making framework is specifically designed for mixed fleets consisting of crewed surface vessels and multiple uncrewed surface vessels, and can be applied to complex and dynamic mission scenarios such as collaborative defense and maritime search and rescue. The framework ensures real-time decision-making through geometrically simplified obstacle modeling and dynamic priority-based distributed task negotiation, while a tripartite game optimization guarantees the superiority. Monte Carlo simulations and prototype tests have demonstrated the framework’s excellent stability, reliability, and engineering applicability. However, the study is based on ideal sensor assumptions and does not yet account for complex noise interference in real-world maritime environments. Future work may extend the framework to cross-domain mixed systems involving underwater, surface, and aerial platforms, while incorporating more robust perception algorithms to improve decision-making reliability in noisy real-world conditions.}, archive = {J_TASE}, author = {Cong Chen and Xiao Liang and Zhao Zhang and Pengcheng Gao}, doi = {10.1109/TASE.2025.3612198}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21430--21442}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical defense decision-making for MSV-guided USVs under complex obstacle constraints}, volume = {22}, year = {2025}, }
• (2025). Autonomous dental surgery for root canal treatment: Compensating for robot-patient misalignment and file deflection. TASE, 22, 21412--21429. (www)

  @article{ , abstract = {Robotic technologies are increasingly used in dentistry for their precision in delicate procedures. While most dental robots focus on implant surgery, automating root canal treatment (RCT) remains challenging due to the need to guide a thin, flexible endodontic file through a narrow, curved root canal without causing ledging or file fracture. Patient movements—particularly those that induce additional file bending during insertion—further complicate robot-assisted procedures. This study presents an autonomous approach for root canal cleaning and shaping by combining force admittance and position tracking. A novel Patient Tracking Module, which connects the patient’s dental brace to the robot end-effector via string potentiometers, is developed to estimate real-time robot-patient pose. Additionally, a file flexibility model is proposed to predict and compensate for file deflection during insertion. A hybrid position/force control strategy, which integrates these estimations, autonomously guides file manipulation, minimizes misalignment, and therefore reduces the risk of file fracture. Experimental validation demonstrates the system’s feasibility and potential for clinical application in precision endodontic procedures. Note to Practitioners—This paper addresses the challenge of automating root canal treatment in robot-assisted endodontics, which is complicated by unpredictable patient movements and bending of endodontic files. The proposed system follows a structured workflow: (1) the dentist first diagnoses the case, labels the working length of the canal on preoperative cone-beam computed tomography (CBCT) scans, and performs access opening to the pulp chamber; (2) a custom-designed tooth brace is then installed; (3) the endodontic file held by the robot is aligned with the root canal entrance; (4) the robot is connected to the patient via a string-based Patient Tracking Module (PTM); and (5) the robot autonomously performs canal cleaning and shaping until the working length is reached. The PTM estimates patient movement in real time and compensates for it. Combined with force sensing and file deflection compensation, the system dynamically adjusts the surgical path to maintain alignment and reduce the risk of file fracture. This improves procedural safety, precision, and efficiency. Preclinical tests on acrylic models and 3D-printed teeth show promising results, but further validation in live patient settings is needed. Future work will explore automating additional steps, such as access opening and obturation, to enhance the overall treatment workflow.}, archive = {J_TASE}, author = {Hao-Fang Cheng and Yi-Ching Ho and Cheng-Wei Chen}, doi = {10.1109/TASE.2025.3611997}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21412--21429}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous dental surgery for root canal treatment: Compensating for robot-patient misalignment and file deflection}, volume = {22}, year = {2025}, }
• (2025). A quality prediction approach for compressor blade manufacturing with various geometric dimensions. TASE, 22, 21400--21411. (www)

  @article{ , abstract = {As the core component of aero engines, compressor blades in turbine engines have complex surface geometry requirements. With the rapid iteration of the turbine power system, the compressor blades used are constantly updated. An accurate prediction model that can quickly adapt to new blade production on manufacturing processes is urgently needed. However, most of the existing technologies cannot meet the high precision requirements of blade manufacturing, and the lack of data for newer blades makes training a predictive model in a new machining task a high time and economic cost. In this paper, we propose a domain generalization-based quality prediction framework for high-precision compressor blade manufacturing processes. The framework can make full use of the relevant manufacturing process data of historical blades to build a domain generalization model for adapting to the prediction of processing accuracy of new blades that have not been processed before. In particular, this study proposes a non-invasive generalization solution that meets accuracy requirements. Coordinate measuring machine (CMM) data is used as input features for prediction, rather than retrofitting sensors onto machinery after installation. To evaluate the effectiveness of the proposed framework, experiments were conducted on four series of aerospace compressor blades from Wuxi Turbine Blade Co., Ltd., China. CMM data from key stages such as blade root milling and integrated precision milling are used as input to predict final geometric errors. Comparison results verify that the proposed framework achieves the smallest prediction error at 0.037 mm, 0.013 mm, 0.021 mm, and 0.016 mm for four blades respectively. Note to Practitioners—This study is motivated by the necessity to develop a high-precision quality prediction model for compressor blade manufacturing, aiming to enhance production efficiency and mitigate economic and temporal costs associated with manufacturing losses. The rapid advancement of turbine engine power systems necessitates continuous dimensional modifications of compressor blades, which are high-precision components with complex curved surfaces. However, when blades with dimensional changes are introduced into production, existing quality prediction models often fail to align with the new specifications, necessitating model retraining. This model retraining process not only incurs significant time delays, adversely affecting production efficiency, but also results in economic losses due to the consumption of test-cut products. To address this issue, we propose a domain generalization-based quality prediction model that leverages historical machining data from similar blades. This approach enables the direct application of the model to blades with dimensional adjustments, ensuring high precision, improving processing efficiency, and reducing costs.}, archive = {J_TASE}, author = {Qianjun Liu and Zhanpeng Yang and Xinxin Dong and Yongjian Dai and Zhexuan Zeng and Xuechun Qiao and Cheng Cheng}, doi = {10.1109/TASE.2025.3611990}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21400--21411}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A quality prediction approach for compressor blade manufacturing with various geometric dimensions}, volume = {22}, year = {2025}, }
• (2025). Innovative cost allocation and energy management in off-grid microgrids via stackelberg game. TASE, 22, 21387--21399. (www)

  @article{ , abstract = {This paper proposes a novel energy interaction management strategy for off-grid microgrids, addressing the challenges posed by the uncertainty in renewable energy generation and the variability of consumer willingness to participate in demand response (DR). This study introduces an adaptive load type conversion (ALTC) model to quantify consumer willingness and a dynamic coefficient of responsibility (DCR) mechanism for the fair allocation of costs arising from renewable energy prediction errors. A consumer-free interaction algorithm based on the Stackelberg game is further proposed to manage energy scheduling to balance supply-side (SS) and demand-side (DS) objectives. The proposed method is validated using real-world data from an off-grid microgrid in Guizhou, China. Experimental results demonstrate a 26.2% improvement in the profit-to-cost ratio (PCR) compared to the baseline method, and a 37% reduction in peak-to-valley load ratio (PVR). By integrating privacy-preserving load scheduling and adaptive tariff policies, our approach enhances the adaptability of hourly scheduling operations and promotes equitable DR participation, offering a robust solution for modern microgrids. Note to Practitioners—Off-grid microgrids, which power remote communities, industrial sites, or emergency facilities, often struggle with two key challenges: unreliable renewable energy generation and unpredictable consumer electricity usage. These uncertainties can lead to wasted energy, higher costs, and unfair pricing for users. Our work provides a practical solution by introducing an automated system that estimates how willing consumers are to adjust their electricity use without violating privacy, fairly assigns extra costs when renewable energy forecast errors affect the balance between supply and demand, and optimizes energy pricing and supply without requiring direct communication between producers and consumers. In real-world testing, this approach reduced peak energy demand by 37% and improved cost efficiency by 26%, making microgrids more stable and economical. While the system works best with historical usage data and may need adjustments for larger or more complex grids, it can be applied today in remote villages. Future improvements could expand it to networks of multiple microgrids or adapt to rapidly changing weather conditions. This method offers a smarter, fairer way to manage off-grid energy systems while keeping costs and complexity low.}, archive = {J_TASE}, author = {Fei Teng and Yunpeng Gao and Jiangzhao Wang and Wei Zhang and Miao Wu and Keyue Zhuo}, doi = {10.1109/TASE.2025.3611935}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21387--21399}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Innovative cost allocation and energy management in off-grid microgrids via stackelberg game}, volume = {22}, year = {2025}, }
• (2025). Control performance orientated region stabilization and h∞ control of discrete stochastic systems: A data-driven approach. TASE, 22, 21376--21386. (www)

  @article{ , abstract = {A data-driven $H_{\infty }$ control method with dynamic performance constraints is proposed for the problem of region stabilization control in the case of unknown system matrix. By establishing a system characterization model based on measured data, the generalized pole configuration theory is ingeniously combined with the data-driven control framework. Firstly, the real-time data generated from the system operation is used to construct a data-driven dynamic characterization. A generalized pole constraint-based regional constraints condition is established to accurately regulate the state convergence rate and damping response characteristics while ensuring the stability of the closed-loop system. The classical $H_{\infty }$ control framework is further integrated to construct a novel data-driven $H_{\infty }$ performance criterion to achieve the synergistic optimization of dynamic performance constraints and disturbance suppression capability. The method gets rid of the dependence on the a priori model of the system and solves the co-optimization problem of unknown system dynamic performance regulation and robustness enhancement under a unified framework. Numerical simulations and power system control case validations show that the integrated control performance of uncertain complex systems is significantly improved. Note to Practitioners—Optimizing dynamic system performance is critical in practical engineering applications. However, inherent uncertainties in real-world systems pose significant challenges to precise mathematical modeling. Consequently, achieving dynamic performance enhancement without reliance on a priori models remains a core problem that embodies both theoretical value and engineering challenges. To address this challenge, this paper introduces a novel data-driven precision control methodology that utilizes real-time operational data generated by systems to construct data-driven regional constraints. This approach not only ensures system stability but also enables precise regulation of convergence rate, damping response, and other dynamic characteristics. Furthermore, the proposed methodology guarantees both steady-state and dynamic performance for discrete-time stochastic system under disturbances. Simulation experiments validated the effectiveness of this strategy in power systems.}, archive = {J_TASE}, author = {Xinran Li and Huasheng Zhang and Shun-Feng Su and Wei Sun}, doi = {10.1109/TASE.2025.3610951}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21376--21386}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Control performance orientated region stabilization and h∞ control of discrete stochastic systems: A data-driven approach}, volume = {22}, year = {2025}, }
• (2025). Optimal secure control for cyber-physical systems under false data injection attacks via incremental iterative Q-learning algorithm. TASE, 22, 21365--21375. (www)

  @article{ , abstract = {An incremental iterative Q-learning algorithm (IIQLA) is proposed to tackle the optimal secure control problem for cyber-physical systems under false data injection attacks. Within a zero-sum game framework, the secure control problem is transformed into solving an iterative algebraic Riccati equation. To derive the optimal secure control policy from the equation, the IIQLA is developed, which does not require prior knowledge of the system dynamics. This algorithm utilizes two auxiliary variables to separate the behavior policy and the target policy, thereby improving the exploration of data. As a result, the devised control policies are more conservative, leading to solutions that are closer to the optimal policy. Moreover, by introducing an adaptive learning rate, the proposed IIQLA can accelerate the convergence speed and decrease the number of required iterations, thus alleviating the computation burden. The convergence of the proposed IIQLA with different learning rates is also analyzed. In addition, the closed-loop system is guaranteed to be asymptotically stable, and the exploration noise does not introduce bias into the optimal policies. Finally, a simulation example is provided to demonstrate the effectiveness of the proposed approach. Note to Practitioners—This study addresses the secure control challenges posed by cyber-physical systems in the presence of false data injection attacks. Traditional secure control methods mainly compensate for the systems after they are attacked, which is passive rather than active resistance to attacks. In the zero-sum game framework, an IIQLA is proposed to obtain the optimal secure control policy without the system dynamics, which integrates an improved adaptive relaxation function and more history information. The IIQLA helps accelerate the convergence speed, decrease the number of required iterations, and improve the exploration of data. The designed method is well suited for real systems that face unpredictable events, such as cyber-attacks, disturbances, and faults.}, archive = {J_TASE}, author = {Jinyan Li and Xiao-Jie Peng and Yan Lei and Guangdeng Chen and Hongyi Li}, doi = {10.1109/TASE.2025.3610922}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21365--21375}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal secure control for cyber-physical systems under false data injection attacks via incremental iterative Q-learning algorithm}, volume = {22}, year = {2025}, }
• (2025). Practical finite-time contraction of impulsive systems with application to design of impulsive control for PFTC. TASE, 22, 21353--21364. (www)

  @article{ , abstract = {This paper investigates practical finite-time contraction (PFTC) for impulsive systems with application to design of impulsive control for PFTC. The PFTC notions including practical finite-time incremental stability (PFTIS) and practical finite-time stability (PFTS) are defined. By using the variation method, the equivalence of PFTC and PFTIS is established for an impulsive system via PFTS of its virtual dynamics. Based on the equivalence, the PFTC criteria and the settling time estimates are obtained for impulsive systems. Then, the results are used to design impulsive control for PFTC. Three types of impulsive control: time-triggered impulsive control (T-IC), state-triggered impulsive control (S-IC), and time-state-mixed-triggered impulsive control (TS-IC), are designed, respectively, for PFTC. The theoretical results as well as numerical simulations with application to the precision control of missile dynamics are shown that TS-IC is the optimal impulsive control with the shortest settling time, i.e., the fastest entry and stay in the target area, the least number of impulsive control, the lowest impulse frequency, and strongest robustness compared to T-IC and S-IC. And TS-IC also improves the mechanism of impulsive control in the literature, which is either purely time or state triggered. Note to Practitioners—In practical applications, many systems often struggle to achieve complete consistency with the target state, and even if achievable, the conditions to be met will be very strict. The study introduces “practical finite-time contraction” (PFTC), which ensures systems quickly reach and stay near a target state without requiring exact convergence. This is useful in real-world systems where minor fluctuations are tolerable (e.g., voltage regulation in power grids). This approach is more feasible in practical implementation and holds significant applied value. This study focuses on the stability control problem of dynamical systems (e.g., power systems or robotic systems) and proposes a solution to ensure that the system state enters and remains within a predefined acceptable range within a finite time without requiring exact convergence to equilibrium. By using the Lyapunov-like function metric, the PFTC criteria and the PFTC settling time are obtained. Three control strategies are designed for finite-time stabilization: time-triggered impulsive control (T-IC), state-triggered impulsive control (S-IC), and time-state-mixed-triggered (TS-IC). Both theoretical analysis and numerical simulations with application to the precision control of missile dynamics demonstrate that TS-IC achieves faster stabilization with fewer control actions and lower impulse frequency compared to T-IC and S-IC, and also stronger robustness w.r.t. the perturbation, making it the optimal strategy for practical applications.}, archive = {J_TASE}, author = {Bin Liu and Ming-Han Jiang and Ling Li and Dong-Nan Liu}, doi = {10.1109/TASE.2025.3610668}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21353--21364}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical finite-time contraction of impulsive systems with application to design of impulsive control for PFTC}, volume = {22}, year = {2025}, }
• (2025). Leader-following formation of heterogeneous multi-agent systems with time-varying topology: A virtual neighbor framework. TASE, 22, 21341--21352. (www)

  @article{ , abstract = {This paper presents a set-membership control approach for formation tracking in heterogeneous leader-following multi-agent systems. In order to improve the adaptability of the control strategy to changes in topology, a virtual neighbor is constructed for each follower, which represents the combined influence of all its neighbors. With the help of virtual neighbors, the interaction relationships between agents can be hidden, thereby mitigating the impact of topology changes on controller design. Then, a novel time-varying topology with adaptive adjacent weights is proposed, such that the coupling strength between agents can be actively adjusted to improve the formation performance. Subsequently, the set-membership technique is utilized to establish the optimization problems to ensure that the one-step forward tracking errors can be constrained within the ellipsoid sets. Further, an online algorithm is developed to continuously obtain the optimal controllers by solving the optimization problems recursively. Finally, the effectiveness and practicality of the proposed approach are demonstrated via a numerical simulation and its application on a robotic platform with UAVs and UGVs. Note to Practitioners—Formation tracking in multi-agent systems, such as those involving UAVs and UGVs, often faces challenges brought by dynamic interaction topology and limited local information. To better adapt to the dynamics of the topology, the original interaction relationships of each agent will be concealed by introducing a virtual neighbor. Then, each agent only needs to consider the interactions with virtual neighbor, which simplifies the design of the controller and is beneficial for scaling up the system size. Another practical limitation in such scenarios is that agents often only have access to local information, making it unclear how to determine the best direction to move. To solve this problem, a time-varying topology with adaptive neighbor weights is proposed, which allows the agent to dynamically adjust its coupling strength according to the state information of its neighbors, so as to make more appropriate movement decisions in real time. Experimental results show that the proposed methods can adapt to continuous changes in topology, while effectively reducing formation errors. Moreover, considering the possible existence of obstacles in the actual environment, collision avoidance strategies will hopefully be incorporated to enrich the proposed virtual neighbor framework with time-varying topology.}, archive = {J_TASE}, author = {Zehui Xiao and Hongyi Xu and Jie Tao and Renquan Lu and Peng Shi}, doi = {10.1109/TASE.2025.3606489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21341--21352}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Leader-following formation of heterogeneous multi-agent systems with time-varying topology: A virtual neighbor framework}, volume = {22}, year = {2025}, }
• (2025). Point-to-point reference trajectories generation using frequency-aware B-Splines/NURBS. TASE, 22, 21325--21340. (www)

  @article{ , abstract = {Implementing Non-Uniform Rational B-Splines in motion controllers involves challenges like computational complexity, parameterization issues, high-fidelity interpolation needs, real-time processing requirements, hardware limitations, system integration, error management, and the demand for user-friendly design tools, all of which necessitate advanced algorithms, robust hardware, and effective interfaces for successful implementation. Expanding the standard sine profile allows for the creation of a generalized, symmetric, frequency-sensitive basis function for generating point-to-point motion trajectories, while other functions like polynomials, sigmoids, and harmonic models can also be utilized. This study employs the basis function at the jerk level to meet the motion system’s constraints, and time shifts create B-spline-like and NURBS-like profiles for CAD and motion control, eliminating the need for interpolation and curve fitting. Utilizing vector notation and algorithms reduces computational demands on both the CAD and the motion controller sides, enabling real-time implementation. Frequency-sensitive B-spline profiles enhance both repetitive and random walk motions, facilitate data collection, and help segment the workspace into regions with unique frequency characteristics. A model-free approach addresses position-dependent errors, and experimental results validate the effectiveness of these techniques. Robust statistical models based on Multivariate Analysis of Variance are developed to both characterize and assess the performance of a precision motion system under the proposed framework. For example, the mean positioning error of the motion system improved from $2.29\,\mu \mathrm {m}$ to $0.717\,\mu \mathrm {m}$ at the maximum operating frequency of $90\,\mathrm {rad/s}$ under the model-free approach. Note to Practitioners—This paper introduces novel definitions for B-spline and NURBS entities, which are widely used in computer-aided design, manufacturing, and robotics applications. Upon implementation, these new definitions enable practitioners to create computer-based parts, such as paths or artworks, while adhering to the intended machine capabilities for physical realization. By considering machine limitations, practitioners can anticipate how they will impact product quality. Modifications to parts or assigned machines can then be made, potentially reducing rework. However, it is important to note that this study focuses solely on the kinematic constraints of machines, without considering the computational overhead balance between computers and machines. Additionally, the goal is not to perfectly match the geometry obtained using standard B-splines and NURBS entities. Future investigations will explore issues related to contouring, tracking motions, and other aspects.}, archive = {J_TASE}, author = {Yazan M. Al-Rawashdeh and Marcel F. Heertjes and Mohammad Al Janaideh}, doi = {10.1109/TASE.2025.3605568}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21325--21340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Point-to-point reference trajectories generation using frequency-aware B-Splines/NURBS}, volume = {22}, year = {2025}, }
• (2025). Visual and textual commonsense-enhanced layout learning for vision-and-language navigation. TASE, 22, 21311--21324. (www)

  @article{ , abstract = {In the Vision-and-Language Navigation (VLN) task, an agent must comprehend natural language instructions and execute precise navigation in complex environments. While significant progress has been made in the VLN field, the limited availability of navigation data hinders existing methods from fully learning the commonsense relationships between rooms and landmarks, which are crucial for environmental understanding and successful navigation. To address this issue, this work proposes a Visual and Textual Commonsense-Enhanced Layout Learning Model (ViTeC). We leverage the open-world knowledge embedded in large models by utilizing ChatGPT and BLIP-2 to provide commonsense information about environments. Specifically, BLIP-2 analyzes the room type corresponding to each panoramic image, while ChatGPT infers and provides knowledge about the most common landmarks within each room type. Moreover, to compensate for the agent’s lack of commonsense at the visual level, we employ Stable Diffusion to generate commonsense-based visual images, enhancing the agent’s visual perception. To ensure the agent effectively learns commonsense about the environment, we designed a Text Commonsense Layout Learning Module and a Visual Commonsense Layout Learning Module. These modules help the agent acquire environmental commonsense from both linguistic and visual perspectives, enabling it to utilize commonsense information effectively during navigation, thereby improving its environmental understanding and reasoning capabilities. Experimental results demonstrate that ViTeC achieves strong performance on REVERIE, R2R, and SOON datasets, exhibiting good generalization ability in complex environments. This validates the effectiveness of ViTeC in enhancing the agent’s environmental understanding and navigation capabilities. Note to Practitioners—In real-world applications, service robots often lack commonsense knowledge of environmental layouts, leading to inefficient navigation in complex scenarios, particularly in smart homes, medical assistance, and retail guidance. To address this issue, our research integrates knowledge from large models and visual generation techniques, enabling navigation agents to leverage commonsense information about room layouts and the distribution of common objects to enhance path planning. The advantage of this approach lies in improving the agent’s understanding of the environment, thereby reducing unnecessary exploration and increasing navigation accuracy. However, its effectiveness depends on the comprehensiveness of the knowledge base and the agent’s visual reasoning capabilities, which may pose adaptability challenges in highly dynamic or non-standard environments. In the future, this method can be further enhanced through online knowledge updates, adaptive learning, and multimodal data fusion, equipping robots with stronger navigation capabilities in a wider range of real-world scenarios.}, archive = {J_TASE}, author = {Fang Gao and Lei Shi and Jingfeng Tang and Jiabao Wang and Shaodong Li and Shengheng Ma and Jun Yu}, doi = {10.1109/TASE.2025.3604018}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21311--21324}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Visual and textual commonsense-enhanced layout learning for vision-and-language navigation}, volume = {22}, year = {2025}, }
• (2025). Adaptive safety-critical control for high-order systems: A real-time gaussian process approach. TASE, 22, 21296--21310. (www)

  @article{ , abstract = {This paper proposes a novel adaptive fast variational sparse Gaussian process (AFVSGP) framework to ensure real-time safety for high-order systems under model uncertainties and dynamic obstacle environments. The framework effectively addresses the challenge of maintaining real-time safety guarantees during unknown trajectory transitions in nonstationary environments. To achieve this, the proposed framework incorporates three key innovations. First, a specialized kernel function is embedded within the VSGP algorithm to decouple control inputs from uncertainties while preserving the convexity of posterior-based safety constraints. Second, an adaptive online incremental learning mechanism is introduced, integrating forgetting capabilities with dynamic reconstruction rules for training datasets and inducing sets, thereby accelerating inference convergence and enabling compact uncertainty prediction with reduced computational complexity. Third, a high-order control barrier function (HOCBF)-based safety filter is developed to synthesize safe control inputs by leveraging the proposed learning model, thereby establishing rigorous probabilistic bounds on the satisfaction of safety specifications. The effectiveness of the proposed framework is validated through both simulation and real-world obstacle avoidance experiments on a 7-DOF Franka robot. The video is available at: https://www.youtube.com/watch?v=2tCKYM_79S8 Note to Practitioners—This paper addresses a critical challenge in autonomous robotics: ensuring real-time safety guarantees when robot dynamics are unknown and operating environments change rapidly. In industrial settings like robotic manufacturing, collaborative robots, or service robots navigating human spaces, maintaining safety while fast adapting to unexpected changes is essential. Our proposed framework helps solve this through a specialized learning method that continuously adapts to new situations while maintaining safety boundaries. Unlike previous approaches, which either become too conservative or require extensive offline training, our method provides three practical advantages. First, it efficiently separates control inputs from uncertainties, making implementation straightforward for high-dimensional robotic systems. Second, it features a computationally efficient adaptive learning algorithm that operates in real-time. Third, it provides probabilistic safety guarantees that can be tuned to balance performance with safety margins. The framework has been validated on a 7-DOF Franka robot avoiding dynamic obstacles, demonstrating its practicality in real-world settings. Future work will focus on proposing an obstacle perception module to enable more accurate environment observations across diverse obstacle types.}, archive = {J_TASE}, author = {Yu Zhang and Long Wen and Zhenshan Bing and Xiangtong Yao and Linghuan Kong and Wei He and Alois Knoll}, doi = {10.1109/TASE.2025.3611987}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21296--21310}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive safety-critical control for high-order systems: A real-time gaussian process approach}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based optimal formation control for multiple WMRs with visual servoing. TASE, 22, 21284--21295. (www)

  @article{ , abstract = {In this paper, a reinforcement learning (RL) control method is developed for the formation control of multiple wheeled mobile robots (WMRs) with visual servoing. First, a multi-robot system model is constructed based on the kinematic models of mobile robots, the camera model, and the multiple-view geometry principles. The leader-follower structure is then applied to derive the distributed error system. Next, the error term is separated from the optimal performance index function, and the Bellman residual error is obtained based on the Hamilton-Jacobi-Bellman equation (HJBE). Subsequently, the gradient descent method is employed to design the weight update rate, which is implemented in an actor-critic neural network (NN) architecture. The proposed RL control method achieves formation tracking and performance optimization simultaneously, which previous approaches have not accomplished. Furthermore, under the Lyapunov stability theory, it is proven that the follower robots can track the leader in a predefined formation, and the tracking error converges ultimately. The simulation outcomes verify the effectiveness of the developed approach. Note to Practitioners—The primary motivation of this study is to establish a multiple WMRs system model based on visual servoing and to develop an RL control strategy. While visual servoing enhances the flexibility and accuracy of multi-robot systems, there remains a significant gap in research on implementing RL-based control for such models. To address this gap, the paper construct an error dynamics model that integrated multi-view geometry principles with robot kinematic constraints. Based on this model, the paper proposes a novel RL approach implemented through an actor-critic NN framework. Unlike conventional adaptive NNs, the parameters of the actor network are updated through the critic network, which enables more stable and efficient policy optimization by providing continuous performance feedback.}, archive = {J_TASE}, author = {Biao Luo and Yuhao Zhou and Lin Liu and Jialin Xiao and Bing-Chuan Wang and Chunhua Yang}, doi = {10.1109/TASE.2025.3608352}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21284--21295}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based optimal formation control for multiple WMRs with visual servoing}, volume = {22}, year = {2025}, }
• (2025). Targeted path optimization using RRT-MWOAII: A hybrid approach to enhanced smoothness in robotic path planning. TASE, 22, 21269--21283. (www)

  @article{ , abstract = {Path planning in robotics and automation often demands solutions that effectively balance efficiency and path smoothness, particularly in diverse and large-scale environments. To address this, this paper presents the Rapidly Exploring Random Tree-Modified Whale Optimization Algorithm II (RRT-MWOAII), an enhanced version of the previous Rapidly Exploring Random Tree-Modified Whale Optimization Algorithm (RRT-MWOA) approach. The RRT-MWOAII is a hybrid technique that combines the strengths of the RRT path planning method and the MWOAII. It utilizes the RRT path as the initial solution and then applies the MWOAII to improve the smoothness of the RRT path. The key advancement in the RRT-MWOAII algorithm is the incorporation of a targeted global search, where the weakest solution is targeted instead of a random one during the global search phase. This targeted approach helps to efficiently identify and refine the weakest link in the solution, leading to improved overall path optimization. Additionally, the algorithm further refines its local solution by employing a neighborhood search, which allows for fine-tuning of the path. Simulations were conducted and evaluated on seven benchmarks, including RRT, RRT*, Bidirectional RRT (BiRRT), and state-of-the-art optimization methods such as the original Whale Optimization Algorithm (WOA), Improved Whale Optimization Algorithm (IWOA) and Sparrow Search Algorithm (SpSA). Experimental results show that MWOAII consistently produced shorter and smoother paths, outperforming its predecessor with a 31.6% improvement in the initial RRT path smoothness and a 21.05% better average smooth cost across all nine maps. Note to Practitioners—Path planning in robotics and automation often requires solutions that balance efficiency and path smoothness across diverse and sometimes large-scale environments. This work introduces the Rapidly Exploring Random Tree-Modified Whale Optimization Algorithm II (RRT-MWOAII), an advanced path planning technique that addresses these challenges by combining the Rapidly Exploring Random Tree (RRT) method with the Modified Whale Optimization Algorithm II (MWOAII). The algorithm first uses RRT to identify a feasible initial path, then employs MWOA to improve path smoothness in a targeted manner. The RRT-MWOAII brings two significant enhancements over prior approaches: 1) a targeted global search that refines the weakest segment of the current path, reducing redundant computations, and 2) a neighborhood search for local fine-tuning, improving path smoothness and continuity. These improvements result in a substantial gain in path quality, with a 31.6% improvement in the initial RRT path smoothness and a 21.05% better average smooth cost across nine test environments compared to its predecessor.}, archive = {J_TASE}, author = {Izzati Saleh and Mohamad Hazwan Mohd Ghazali and Wan Rahiman}, doi = {10.1109/TASE.2025.3607959}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21269--21283}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Targeted path optimization using RRT-MWOAII: A hybrid approach to enhanced smoothness in robotic path planning}, volume = {22}, year = {2025}, }
• (2025). Efficient alignment of unconditioned action prior for language-conditioned pick and place in clutter. TASE, 22, 21256--21268. (www)

  @article{ , abstract = {We study the task of language-conditioned pick and place in clutter, where a robot should grasp a target object in open clutter and move it to a specified place. Some approaches learn end-to-end policies with features from vision foundation models, requiring large datasets. Others combine foundation models in a zero-shot setting, suffering from cascading errors. In addition, they primarily leverage vision and language foundation models, focusing less on action priors. In this paper, we aim to develop an effective policy by integrating foundation priors from vision, language, and action. We propose A2, an action prior alignment method that aligns unconditioned action priors with 3D vision-language priors by learning one attention layer. The alignment formulation enables our policy to train with less data and preserve zero-shot generalization capabilities. We show that a shared policy for both pick and place actions enhances the performance for each task, and introduce a policy adaptation scheme to accommodate the multi-modal nature of actions. Extensive experiments in simulation and the real-world show that our policy achieves higher task success rates with fewer steps for both pick and place tasks in clutter, effectively generalizing to unseen objects and language instructions. Videos and codes are available at https://xukechun.github.io/papers/A2 Note to Practitioners—This research is motivated by the challenge of generalizable policy learning of language-conditioned pick and place in clutter. Solving such a challenge could significantly improve the robot’s level of automation and intelligence in household and industrial pick and place tasks. Existing methods struggle with large data requirements, poor generalization to unseen scenarios, and cascading errors across individual components. To overcome these limitations, we propose to integrate priors from vision, language, and action foundation models by learning-based alignment. Our policy aligns action priors with 3D vision-language priors by learning one attention layer, requiring less data and preserving zero-shot generalization capabilities from foundation models. Experiments show that our method can improve both task success rate and generalization for pick and place tasks in simulation and the real world. In future work, we will incorporate more action foundation models to extend our approach of action prior alignment to a wider range of tasks, offering a promising direction for general manipulation.}, archive = {J_TASE}, author = {Kechun Xu and Xunlong Xia and Kaixuan Wang and Yifei Yang and Yunxuan Mao and Bing Deng and Jieping Ye and Rong Xiong and Yue Wang}, doi = {10.1109/TASE.2025.3606549}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21256--21268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient alignment of unconditioned action prior for language-conditioned pick and place in clutter}, volume = {22}, year = {2025}, }
• (2025). URMCNet: An uncertainty region modeling and calibration network for power line inspection. TASE, 22, 21242--21255. (www)

  @article{ , abstract = {Accurate power line detection is critical for uncrewed aerial vehicle (UAV)-based inspection systems. However, the detection accuracy is severely compromised by visually complex backgrounds characterized by similar textures and environmental interference, which often lead to substantial false-positives and false-negatives in detection results. To address these limitations, we propose a novel Uncertainty Region Modeling and Calibration Network (URMCNet), which incorporates an Uncertainty Region Modeling and Calibration Strategy that explicitly models and calibrates regions prone to false positives and false negatives, so as to enhance the accuracy of power line detection. The core component of URMCNet is the Primary Feature Calibration Mechanism, which incorporates the False-Positive Region Suppression Module and the False-Negative Region Compensation Module. These two modules adaptively calibrate uncertain regions, thereby making the feature representation more accurate and comprehensive. Furthermore, we introduce a High-Frequency Aware Fusion Decoder, which effectively restores the fine-grained details, thereby guaranteeing the continuity of the prediction results. The comprehensive experimental results demonstrate that URMCNet significantly outperforms state-of-the-art methods on several key performance metrics on the publicly available VITLD and TTPLA datasets. Note to Practitioners—UAV-based inspection has become a crucial method for maintaining power grid safety and developing smart grids, especially in remote or hard-to-reach areas. However, existing detection systems often perform poorly in visually complex environments, where power lines may be obscured by trees, buildings, or other obstacles, leading to significant errors that compromise the reliability of inspections and increase maintenance costs. To address these challenges, we propose a new approach that enhances power line detection accuracy by focusing on areas where prediction errors are most likely to occur. Our method employs an adaptive calibration process that specifically targets regions susceptible to false-positives and false-negatives. This approach enhances detection accuracy and reliability in complex environments, thereby improving the efficiency of UAV-based inspections. Another key contribution of this research is the ability to restore fine details in the detection results, which is crucial for ensuring the continuous detection of power lines in complex or noisy environments. In addition to power line inspection, the techniques proposed can be applied to other fields that also require object detection in complex visual conditions, such as environmental monitoring, infrastructure inspection, and agricultural analysis.}, archive = {J_TASE}, author = {Yichen Liu and Zhiqiang Qu and Guo Zheng and Junjie Ye}, doi = {10.1109/TASE.2025.3606206}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21242--21255}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {URMCNet: An uncertainty region modeling and calibration network for power line inspection}, volume = {22}, year = {2025}, }
• (2025). Information lagrangian of multi-agent systems and its application in fault detection. TASE, 22, 21228--21241. (www)

  @article{ , abstract = {In multi-agent systems (MASs), the information connections between agents present significant challenges for the fault detection of the system. This paper studies MASs from the perspective of information, constructs and explores the significance and connotation of the information Lagrangian, and relates it to the operational status of MAS as an index for detecting whether a fault has occurred in MAS. In this process, the distribution of the potential function on MAS information manifold is analyzed, the relationship between geodesics, Fisher information distance (FID), and information Lagrangian on the information manifold is studied. The distribution of the MAS information Lagrangian is explained in depth from a perspective of information amount and least action principle. The simulation application verified the effectiveness of the proposed method. As an operating status and fault detection index of MAS, information Lagrangian has clear physical significance, and has scalability potential. Note to Practitioners—This work was motivated by the problem of real-time assessment and fault detection for multi-agent systems. A new quantity: information Lagrangian is proposed as the status assessment and fault detection index of multi-agent systems. Which integrates the operational information of multiple agents and can characterize the overall operational status of MAS. With a clear physical meaning, information Lagrangian can evaluate the overall operating status, achieve fault detection and localization of multi-agent systems without being affected by specific control strategies of the systems. The index proposed in our work can be applied to any configuration of multi-agent systems and have good scalability potential. Single type robots, unmanned aerial vehicles, aerial, or their hybrid multi-agent systems can all apply this index to achieve status assessment and fault detection.}, archive = {J_TASE}, author = {Ruotong Qu and Bin Jiang and Yuehua Cheng}, doi = {10.1109/TASE.2025.3603879}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21228--21241}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Information lagrangian of multi-agent systems and its application in fault detection}, volume = {22}, year = {2025}, }
• (2025). RE-formation: Resilient and efficient formation planning in large-scale distributed aerial swarms. TASE, 22, 21212--21227. (www)

  @article{ , abstract = {Due to the limited online computational resources and the inherent probability of hardware and software failures of real-world robots, large-scale formation planning faces two common challenges: computational intractability and agent failures. Based on the theory of sparse graphs and the maximum clique, we achieve a resilient and efficient formation planning (RE-Formation) to address these issues. To improve the computational efficiency of trajectory planning while ensuring flexible formation maneuvers, we introduce sparse graphs to describe connection relationships and present a sparse graph construction method with closed-form solutions. The sparse graphs ensure the $\underline {G}$ lobal $\underline {R}$ igidity for uniquely corresponding to a geometric shape and $\underline {P}$ reserve the main $\underline {F}$ eatures of complete graphs, denoted as the GRPF sparse graph. To prevent the impact of abnormal agents, the problem of eliminating abnormal agents is transformed into an outlier rejection problem that can be solved by computing the maximum clique. We approximate the maximum clique by periodically triggering the calculation of the maximum $k$ - $core$ to meet the real-time computational demands of large-scale swarms. We validate the performance through real-world experiments and implement formation planning with 100 drones in simulation. Benchmark comparisons and ablation experiments demonstrate the effectiveness of our method. Note to Practitioners—In this paper, we address the challenges of large-scale aerial swarm formation planning, focusing on computational efficiency and resilience to agent failures. Real-world swarm robots often face limitations in computational resources and are prone to hardware and software failures, which can severely impact formation coordination. Our proposed method uses sparse graphs that are globally rigid to reduce computational complexity while ensuring flexible formation maneuvers, and employs outliers rejection to eliminate abnormal agents, ensuring resilient collaboration.}, archive = {J_TASE}, author = {Yuan Zhou and Lun Quan and Chao Xu and Guangtong Xu and Fei Gao}, doi = {10.1109/TASE.2025.3603614}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21212--21227}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RE-formation: Resilient and efficient formation planning in large-scale distributed aerial swarms}, volume = {22}, year = {2025}, }
• (2025). Learning-based heterogeneous autonomous vehicles scheduling for on-demand last-mile transportation. TASE, 22, 21195--21211. (www)

  @article{ , abstract = {Last-mile transportation, a critical component of public transit within integrated urban mobility systems, plays a pivotal role in promoting sustainability and requires immediate attention. However, due to the high real-time variability and uncertainties of last-mile travel demand, high passenger concurrency, and dispersed destinations, existing last-mile transportation systems often encounter significant challenges. These challenges include resource shortages and congestion during peak hours, as well as high operational costs and excessive passenger waiting times during off-peak periods. With the rapid development and widespread adoption of autonomous vehicles, which are characterized by centralized control and flexible scheduling, this study proposes leveraging heterogeneous autonomous vehicles to address these challenges. Specifically, a mixed-integer programming model is developed to maximize the service provider’s profit by considering fare profit, passenger waiting time penalties, and operating costs. To enable real-time decision-making, then an attention-based deep reinforcement learning algorithm is introduced. This algorithm incorporates two decoder mechanisms for vehicle selection and passenger allocation in the scheduling of heterogeneous autonomous vehicles. This involves dynamically selecting vehicles from a heterogeneous fleet based on passenger demand using an attention mechanism, optimizing efficiency in serving last-mile travelers. Extensive numerical experiments and a real-world case study across various datasets demonstrate that the proposed service model and algorithm effectively solve the scheduling problem while meeting the demands of on-demand last-mile transportation. Furthermore, these innovations contribute to reducing fleet carbon emissions and advancing sustainable urban transportation. Note to Practitioners—Last-mile transportation systems face critical challenges: overcrowded vehicles during peak hours, high costs due to underused resources in off-peak periods, and passenger dissatisfaction from long wait times. These issues hinder sustainable urban mobility and strain transit operators’ budgets. This work addresses these problems by deploying heterogeneous autonomous vehicle fleets with varying sizes and capabilities for last-mile transportation. The proposed system dynamically assigns vehicles to passengers based on real-time demand, optimizing efficiency while reducing operational costs. For transit operators, this approach balances fare revenue with penalties for passenger delays, ensuring profitability even during fluctuating demand. Cities adopting shared autonomous vehicle services can benefit from reduced congestion and lower carbon emissions, aligning with sustainability goals. Our experiments validate that the method improves service reliability and fleet utilization across demand scenarios. Implementing this solution requires integrating demand prediction tools with a centralized autonomous vehicle dispatch platform, compatible with existing mobility apps. Future extensions could adapt the system to mixed fleets (combining autonomous vehicles with traditional vehicles) or expand to multi-city networks. By addressing both economic and environmental needs, this framework offers a scalable path toward smarter, greener urban transportation.}, archive = {J_TASE}, author = {Yongwu Liu and Binglei Xie and Yuying Long and Jiawei Chen and Gangyan Xu}, doi = {10.1109/TASE.2025.3611676}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21195--21211}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based heterogeneous autonomous vehicles scheduling for on-demand last-mile transportation}, volume = {22}, year = {2025}, }
• (2025). Adaptive excavation automation in complex soil environments using reinforcement learning. TASE, 22, 21181--21194. (www)

  @article{ , abstract = {This paper presents an advanced excavation automation framework that improves trajectory efficiency and improves operational stability in diverse and complex soil conditions. The framework integrates the fundamental equation of earthmoving (FEE) with the proximal policy optimization (PPO) algorithm to enable adaptive trajectory planning during excavation. Using these estimates, the PPO algorithm iteratively optimizes excavation strategies to minimize applied force and improve trajectory efficiency, contributing to potential energy savings. The validation of the proposed framework is performed through simulations conducted under varying soil conditions and initial height configurations. The results show an 18.1 % reduction in total excavation resistance compared to baseline models, achieving a mean absolute error (MAE) of 0.065 m. These findings confirm the effectiveness of the framework in reducing excavation resistance and improving task precision in simulated environments. Note to Practitioners—This paper is inspired by the lack of applicable excavation mechanisms due to the variability of soil conditions and proposes an excavation framework that combines soil resistance modeling and artificial intelligence (AI) to improve operational efficiency in various soil environments. The system models key soil parameters and dynamically adjusts excavation trajectories to minimize resistance and improve leveling accuracy. The focus is on improving robustness to flexibly adapt to diverse soil conditions. Industry professionals in construction, mining, and infrastructure development can utilize this framework to reduce energy consumption and improve operational precision. Experimental results demonstrate that the proposed solution effectively reduces excavation resistance and improves task accuracy, demonstrating its applicability to real-world automated environments. However, this approach assumes static soil properties during single cycles, which may pose limitations in rapidly changing environments. In addition, obtaining precise soil property data in real-world scenarios remains a challenge. Future research will address the design of excavation algorithms based on real-time sensor data integration for soil property estimation.}, archive = {J_TASE}, author = {Mingyu Shin and Junhyung Cho and Joongheon Kim and Soyi Jung}, doi = {10.1109/TASE.2025.3611032}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21181--21194}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive excavation automation in complex soil environments using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). An end-to-end large model framework of wearable augmented vision device for the visually impaired. TASE, 22, 21167--21180. (www)

  @article{ , abstract = {Visual impairments significantly affect individuals’ ability to perform essential tasks such as communication, object search, and navigation. Traditional wearable augmented vision devices rely on modular designs that separate functions like perception and path planning, leading to cumulative errors and inefficiencies in real-world applications. To address these challenges, we propose an end-to-end multimodal large model framework, UniANS, specifically designed for wearable augmented vision devices. UniANS integrates visual perception, speech interaction, and path planning into a unified framework. Such integration improves task coordination, reduces error propagation, and enhances overall performance. We also propose a prompt design strategy with a mixture of cluster-conditional low-rank adaptation experts architectures and dual-branch encoders, combined with advanced preprocessing techniques for visual and speech modules. The framework has been validated through ablation studies, showing superior performance in accuracy and task effectiveness compared to existing methods. We further showcase its capabilities in addressing challenges related to communication, object search, and indoor navigation tasks. The design of UniANS enhances mobility and quality of life for visually impaired individuals. Note to Practitioners—UniANS is an end-to-end multimodal framework for wearable vision devices that assist visually impaired users. Unlike traditional modular designs, which suffer from error accumulation due to separated perception, interaction, and planning, UniANS integrates these functions into a unified system; such integration improves coordination and robustness. It employs a progressive training strategy with cluster-conditional low-rank adaptation experts and dual-branch encoders, along with advanced preprocessing, to enhance performance across dynamic environments. In practice, UniANS achieves strong results in communication, object search, and indoor navigation without pre-built maps, making it well-suited for unfamiliar settings. Experimental results show it outperforms existing methods in accuracy and effectiveness, offering practitioners a reliable tool to enhance user mobility. We advocate UniANS as a foundation for next-generation, user-centered assistive technologies with potential for further expansion into complex tasks and improved user experience.}, archive = {J_TASE}, author = {Yang Song and Zhijun Li and Yu Kang and Guoxin Li and Haisheng Xia}, doi = {10.1109/TASE.2025.3602402}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21167--21180}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An end-to-end large model framework of wearable augmented vision device for the visually impaired}, volume = {22}, year = {2025}, }
• (2025). Adaptive gradient neural networks for solving the time-varying sylvester equation. TASE, 22, 21151--21166. (www)

  @article{ , abstract = {This paper develops several new dynamical designs, based on the gradient neural network (GNN), from the perspective of control theory to solve the time-varying Sylvester equation (TVSE). We start with an adaptive gradient neural network called AGNN-S. Then, based on Lyapunov theory, we propose three improved models: ACGNN, AIGNN, and ABGNN. Among them, the ABGNN model stands out for its fastest convergence speed and strongest robustness to noise. Theoretical analysis confirms that all proposed models solve the TVSE effectively, with ACGNN, AIGNN, and ABGNN converging faster than AGNN-S. Theoretical analysis shows that using certain nonlinear activation functions can further boost convergence speed. A robustness analysis indicates that the AGNN-S, ACGNN, and ABGNN models maintain stable convergence even in the presence of differentiation or model-implementation errors. Comparisons with the classical zeroing neural network (ZNN) and other six state-of-the-art GNN- and ZNN-type models demonstrate the superior accuracy and efficiency of our approaches, especially the ABGNN model. Numerical experiments notably highlight ABGNN’s advantages over other models under certain parameter setting. Finally, we showcase applications in time-varying quadratic programming and robotic arm trajectory tracking to verify the practical value of the models. Note to Practitioners—This paper targets practitioners seeking efficient and robust neural network-based solvers for time-varying matrix equations, particularly the time-varying Sylvester equation, which commonly arises in adaptive control, signal processing, and dynamic system modeling. Classical methods, such as the GNN, may suffer from limited solution accuracy. The proposed models, referred to as AGNN-S, ACGNN, AIGNN, and ABGNN, are designed for rapid convergence and strong noise resistance under time-varying conditions. Among them, ABGNN exhibits the fastest convergence and strongest robustness, making it particularly suitable for deployment in real-world engineering systems with uncertain or fluctuating dynamics for the goal of good performance. These models are relevant for hardware implementation where stability and speed are critical. Additionally, the adaptability of these neural networks extends to broader applications, such as solving constrained optimization problems and robot mobile arm tracking.}, archive = {J_TASE}, author = {Changxin Mo and Hongyan Dai and Predrag S. Stanimirović and Yimin Wei}, doi = {10.1109/TASE.2025.3610407}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21151--21166}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive gradient neural networks for solving the time-varying sylvester equation}, volume = {22}, year = {2025}, }
• (2025). Real-time stochastic reference tracking nonlinear predictive functional control: A case study on wind turbines. TASE, 22, 21139--21150. (www)

  @article{ , abstract = {In recent years, the development of Nonlinear Model Predictive Control (NMPC) methods has shifted focus from set-point stabilization to tracking reference trajectories. However, uncertainties in reference signals and the need for dynamic forecasting have underscored the demand for robust NMPC frameworks. To simplify analysis and ease implementation, these approaches typically assume constant forecasting of the stochastic reference, which often leads to performance degradation. In this paper, we move beyond this simplifying assumption and propose a novel nonlinear Predictive Functional Control (PFC) framework for tracking stochastic references, leveraging fundamental PFC concepts to significantly reduce the computational burden. Feasibility and convergence to a robust invariant set are guaranteed under a cost function constraint, and the implementation is carried out using a parallel Newton method. The proposed controller is designed and evaluated for a wind turbine, with stochastic wind speed as the reference. Results show superior performance over existing methods, especially in reducing computation time and improving tracking robustness. Note to Practitioners—This paper proposes a NMPC method for tracking stochastic references, applicable to industrial systems with uncertain inputs, such as wind turbines. Many NMPC methods assume that the stochastic reference remains constant, which compromises control accuracy. This research addresses this limitation by proposing an advanced nonlinear version of PFC, offering an approach that enhances tracking accuracy and reduces computational time. PFC is utilized in various industries due to its simplicity and efficiency. The proposed method, by incorporating dynamic reference forecasting and parallel Newton iterations, enables real-time implementation. Experiments on a wind turbine demonstrate that this approach increases aerodynamic power extraction and reduces mechanical stresses. Furthermore, it can be implemented with minimal modifications in other industrial systems requiring fast and accurate predictive control.}, archive = {J_TASE}, author = {Mohammad Soleymani}, doi = {10.1109/TASE.2025.3608971}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21139--21150}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time stochastic reference tracking nonlinear predictive functional control: A case study on wind turbines}, volume = {22}, year = {2025}, }
• (2025). Tracailer: An efficient trajectory planner for tractor-trailer robots in unstructured environments. TASE, 22, 21118--21138. (www)

  @article{ , abstract = {The tractor-trailer robot consists of a drivable tractor and one or more non-drivable trailers connected via hitches. Compared to typical car-like robots, the addition of trailers provides greater transportation capability. However, this also complicates motion planning due to the robot’s complex kinematics, high-dimensional state space, and deformable structure. To efficiently plan safe, time-optimal trajectories that adhere to the kinematic constraints of the robot and address the challenges posed by its unique features, this paper introduces a lightweight, compact, and high-order smooth trajectory representation for tractor-trailer robots. Based on it, we design an efficiently solvable spatial-temporal trajectory optimization problem. To deal with deformable structures, which leads to difficulties in collision avoidance, we fully leverage the collision-free regions of the environment, directly applying deformations to trajectories in continuous space. This approach not requires constructing safe regions from the environment using convex approximations through collision-free seed points before each optimization, avoiding the loss of the solution space, thus reducing the dependency of the optimization on initial values. Moreover, a multi-terminal fast path search algorithm is proposed to generate the initial values for optimization. Extensive simulation experiments demonstrate that our approach achieves several-fold improvements in efficiency compared to existing algorithms, while also ensuring lower curvature and trajectory duration. Real-world experiments involving the transportation, loading and unloading of goods in both indoor and outdoor scenarios further validate the effectiveness of our method. The source code is accessible at https://github.com/Tracailer/Tracailer Note to Practitioners—This paper addresses the challenges of motion planning for tractor-trailer robots, which are crucial in industries like logistics and agriculture. Our approach introduces a lightweight trajectory planning method that efficiently generates safe and time-optimal paths while considering the unique kinematic constraints of these vehicles. By utilizing the environment’s collision-free areas, we enhance safety and reduce reliance on initial conditions. While our method shows significant efficiency improvements in simulations and real-world tests, its effectiveness may vary in extremely dynamic environments. Practitioners can easily adopt our open-source code to integrate this planning technique into their operations, improving the efficiency and safety of tractor-trailer systems and paving the way for future advancements in autonomous transportation.}, archive = {J_TASE}, author = {Long Xu and Kaixin Chai and Boyuan An and Shuhang Ji and Zhenyu Hou and Jiaxiang Gan and Qianhao Wang and Yuan Zhou and Xiaoying Li and Junxiao Lin and Zhichao Han and Chao Xu and Yanjun Cao and Fei Gao}, doi = {10.1109/TASE.2025.3608827}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21118--21138}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tracailer: An efficient trajectory planner for tractor-trailer robots in unstructured environments}, volume = {22}, year = {2025}, }
• (2025). Comparison between behavior trees and finite state machines. TASE, 22, 21098--21117. (www)

  @article{ , abstract = {Behavior Trees (BTs) were first conceived in the computer games industry as a tool to model agent behavior, but they received interest also in the robotics community as an alternative policy design to Finite State Machines (FSMs). The advantages of BTs over FSMs had been highlighted in many works, but there is no thorough practical comparison of the two designs. Such a comparison is particularly relevant in the robotic industry, where FSMs have been the state-of-the-art policy representation for robot control for many years. In this work we shed light on this matter by comparing how BTs and FSMs behave when controlling a robot in a mobile manipulation task. The comparison is made in terms of reactivity, modularity, readability, and design. We propose metrics for each of these properties, being aware that while some are tangible and objective, others are more subjective and implementation dependent. The practical comparison is performed in a simulation environment with validation on a real robot. We find that although the robot’s behavior during task solving is independent on the policy representation, maintaining a BT rather than an FSM becomes easier as the task increases in complexity. Note to Practitioners—Behavior Trees (BTs) have gained significant traction in robotics, with over one hundred research papers published annually in the past five years. BTs have demonstrated their advantages in various applications, such as mobile manipulation, underwater and aerial robotics, and humanoids. Choosing the right policy representation is crucial for successful task completion. While Finite State Machines (FSMs) have been popular for their simplicity, they can become cumbersome as tasks grow more complex. We show this by implementing BTs and FSMs in a mobile manipulation task, tested in simulation and on a real robot. As task difficulty increases, FSMs become harder to maintain due to the rising number of transitions. Our findings provide practical guidance for engineers, helping them choose the best policy representation for their specific robotic applications.}, archive = {J_TASE}, author = {Matteo Iovino and Julian Förster and Pietro Falco and Jen Jen Chung and Roland Siegwart and Christian Smith}, doi = {10.1109/TASE.2025.3610090}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21098--21117}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Comparison between behavior trees and finite state machines}, volume = {22}, year = {2025}, }
• (2025). Coordination of fleets of autonomous vehicles for logistics operations in industrial environments: A grid-based receding horizon control approach. TASE, 22, 21083--21097. (www)

  @article{ , abstract = {Fleets of autonomous vehicles offer an innovative solution to enhance logistic operations in manufacturing systems by increasing efficiency and safety. However, navigating in unstructured environments with moving obstacles presents significant challenges. This paper addresses this problem by proposing a novel control architecture that combines grid-based mapping with receding horizon control techniques. Specifically, unicycle models are used to describe the dynamics of the robots, and sequences of feasible and safe way-points are computed by leveraging the real-time grid map of the entire operating scenario. A distributed set-theoretic model predictive control scheme leverages this information to ensure safe navigation. The effectiveness of the proposed approach is demonstrated through experimental results in a real-world context, where a team of three autonomous robots assists human workers along a steel production line manufacturing angle pipes, block-stages, and caps. Note to Practitioners—This research addresses the challenge of coordinating AMR fleets for logistics in dynamic, unstructured industrial environments, such as manufacturing facilities with moving obstacles. The proposed control architecture combines grid-based path planning with set-theoretic receding horizon control techniques to ensure safe and efficient navigation. This approach adopts distributed control and real-time mapping techniques to enable each AMR to make decentralized, adaptive decisions while fulfilling collision avoidance and logistical requirements. From a practical perspective, this work provides a powerful and scalable framework that enhances operational safety and efficiency by automating transportation tasks within an industrial setting. Although this approach has been validated in a steel production environment with only three unicycle-type AMRs, its scalability paves the way for possible applications with larger fleets in more complex scenarios. Furthermore it can be easily extended to account for other vehicles types and/or more advanced path planning algorithm.}, archive = {J_TASE}, author = {Antonello Venturino and Luigino Filice and Giovanni Mezzatesta and Francesco Tedesco and Giuseppe Franzè}, doi = {10.1109/TASE.2025.3609637}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21083--21097}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Coordination of fleets of autonomous vehicles for logistics operations in industrial environments: A grid-based receding horizon control approach}, volume = {22}, year = {2025}, }
• (2025). Input-output data-driven safe control of unknown discrete-time linear systems. TASE, 22, 21070--21082. (www)

  @article{ , abstract = {This paper presents a data-driven quadratically constrained quadratic program for designing safe and stable controllers using only input-output measurements. Input-output-based control Lyapunov functions (CLFs) and control barrier functions (CBFs) are integrated to derive safe and stable control inputs for discrete-time linear systems with unknown system dynamics. An augmented system is formed using historical input-output measurements and is leveraged to formalize learning of data-driven CLFs and CBFs. In sharp contrast to existing data-driven results in the literature, the presented approach does not require the restrictive assumption of availability of the full system’s state for measurement, which broadens its applicability. Conditions under which the resulting CBF-CLF optimization is convex are outlined, and an interior-point method is presented to handle non-convexities arising from some choices of CBFs. In convex scenarios, closed-form expressions for control inputs are derived. Two special cases that lead to a convex optimization, i.e., systems with a scalar control input and systems with a compact safe set, are thoroughly analyzed. The computational complexity of the proposed interior-point method is discussed and its efficacy is validated through simulations and real-world experiments with mobile robots. Two scenarios are explored: a nonholonomic differential-drive mobile robot and a holonomic mobile robot equipped with Mecanum wheels. Simulations demonstrate the framework’s capability to handle nonlinear constraints, while experiments on the ROSbot XL with Mecanum wheels, highlight its computational efficiency, practical applicability, and effectiveness in ensuring safety and stability in real-world environments. Note to Practitioners—Ensuring safety is paramount in safety-critical systems such as robotics, autonomous vehicles, and industrial automation, where operations in uncertain environments can lead to costly delays, equipment damage, or even harm to human operators. Although designing controllers for such systems typically requires comprehensive knowledge of system dynamics, acquiring an exact model is often challenging due to inherent uncertainties and disturbances. Data-driven control offers a solution by enabling the design of controllers directly from data, circumventing the need for complete dynamic models. In many real-world applications, collecting comprehensive state data is impractical; we often have access only to a subset of the system’s states. To address this limitation, we propose designing safe and stable control inputs using only input-output measurements. Leveraging a quadratic control barrier function, our method accommodates a wide array of safety constraints, including both linear and quadratic forms. This reliance on readily available input-output data simplifies control design, allowing engineers to develop effective controllers without the laborious process of constructing detailed models. This approach not only facilitates faster deployment and easier adaptation to changes but also simplifies the implementation process. Future work will focus on extending the method to handle complex, nonlinear stochastic systems.}, archive = {J_TASE}, author = {Niyousha Ghiasi and Bahare Kiumarsi}, doi = {10.1109/TASE.2025.3609211}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21070--21082}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Input-output data-driven safe control of unknown discrete-time linear systems}, volume = {22}, year = {2025}, }
• (2025). Hierarchical adaptive control framework for autonomous bicycles: Integrating residual decisions and dynamic optimization. TASE, 22, 21057--21069. (www)

  @article{ , abstract = {This paper introduces the Adaptive Residual Decision-Control Synthesis (ARDCS) framework, a hierarchical control architecture that synergizes a model-based Linear Quadratic Regulator (LQR) with an adaptive Dynamic Proximal Policy Optimization (DPPO) agent for autonomous bicycle control. ARDCS is designed to take control of the bicycle’s complex nonlinear dynamics and adapt to environmental uncertainties by leveraging the stability of traditional control with the flexibility of reinforcement learning. A key innovation is a momentum-enhanced dynamic entropy adjustment mechanism within DPPO, which optimizes the exploration-exploitation trade-off for more stable and efficient learning. Comprehensive experiments on balancing and multi-target navigation tasks demonstrate that ARDCS consistently and significantly outperforms both conventional control methods and pure reinforcement learning strategies across varying levels of difficulty. The framework achieves enhanced adaptability and robust stability without relying on auxiliary mechanical stabilizers, offering a potent and generalizable solution for the control of under-actuated systems. Note to Practitioners—This research addresses the practical challenge of controlling inherently unstable two-wheeled vehicles in dynamic environments, particularly relevant for developing low-cost autonomous delivery robots and personal mobility devices. Our control framework enables bicycles to maintain balance and navigate complex obstacles without requiring expensive stabilization mechanisms, significantly reducing hardware complexity and energy consumption. The key innovation lies in combining traditional control theory with machine learning: a basic controller ensures fundamental stability while an AI layer handles complex decision-making. This approach has been successfully tested in simulations mimicking real-world conditions like uneven terrain and sudden wind gusts. Engineers can implement this solution using standard sensors and actuators, with potential applications extending to warehouse robots, drone stabilization, and adaptive industrial manipulators. Current limitations include computational demands for real-time decision-making, which may require optimization for low-power embedded systems. Future work will focus on simplifying the algorithm for edge computing devices while maintaining its adaptive capabilities. This technology could revolutionize last-mile delivery services and enable new categories of agile service robots within 3-5 years.}, archive = {J_TASE}, author = {Shiyu Sha and Yanhong Liu and Benyan Huo and Zhang Chen and Xingang Zhao}, doi = {10.1109/TASE.2025.3608870}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21057--21069}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical adaptive control framework for autonomous bicycles: Integrating residual decisions and dynamic optimization}, volume = {22}, year = {2025}, }
• (2025). A particle swarm optimization event-triggered approach to adaptive sliding mode control of markov jump networked systems. TASE, 22, 21047--21056. (www)

  @article{ , abstract = {This work focuses on adaptive sliding mode control for networked Markov jump systems using a dynamic event-triggered strategy. Unlike previous studies, this mechanism employs a particle swarm optimization algorithm to compute the optimal triggering threshold based on the system’s continuous output values. First, this study designs an event-triggered state observer based on the system conditions, and the error dynamics can be obtained according to the definition of the error. Second, a hyperbolic-type cost function is designed for the event-triggered mechanism, and a dynamic event-triggered scheme is constructed using the particle swarm optimization algorithm. Third, an integral sliding surface is established to obtain the sliding mode dynamics. Following this, a sliding mode controller incorporating adaptive laws is designed, and the reachability is formally demonstrated, and the closed-loop system’s stochastic stability is examined through stochastic Lyapunov function method. Finally, the effectiveness and superiority of the proposed method are verified through RLC circuit. Note to Practitioners—This paper was motivated by the effective control problem in systems such as robotic manipulators and microgrids, which may exhibit stochastic behaviors over a network. As the signal data is transmitted through the network, an important issue arises: how to conserve network bandwidth to prevent congestion when the bandwidth is limited, and how to save system resources when the system reaches a steady state, rendering further updates of control signals unnecessary. Recent studies have proposed static or dynamic event-triggered approaches, which have improved efficiency and system resource usage to some extent. This paper introduces an innovative particle swarm optimization-based adaptive sliding mode approach, which not only enhances the efficiency of existing event-triggered methods but also ensures the robustness of the plant in the presence of unknown matched nonlinear disturbances. The advantages of the proposed approach are verified through a simulation study conducted on a circuit.}, archive = {J_TASE}, author = {Haocheng Lou and Baoping Jiang and Zhen Liu}, doi = {10.1109/TASE.2025.3608783}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21047--21056}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A particle swarm optimization event-triggered approach to adaptive sliding mode control of markov jump networked systems}, volume = {22}, year = {2025}, }
• (2025). A mixed-reality-augmented deep reinforcement learning approach for multi-robot safe motion generation in Human–Robot collaborative manufacturing cells. TASE, 22, 21033--21046. (www)

  @article{ , abstract = {Augmenting capabilities of human operators with multi-robot cells offers substantial advantages for increasing productivity in manufacturing applications. This synergy effectively combines the strengths of both robots and humans, maximizing operational efficiency and leveraging human capabilities. However, achieving these benefits requires real-time, reactive coordination of multi-robot motion generation in response to human motion. Current approaches face significant challenges, particularly in dealing with uncertainties in human motions. To address these issues, this paper introduces the Deep Reinforcement Learning (DRL) approach for end-to-end safe motion generation in human multi-robot collaborative workspaces. First, the DRL approach is augmented by adopting mixed-reality (MR) features to facilitate efficient state perception and representation of tasks, humans, robots, and scenes for enabling effective learning motion generation policy. Moreover, to better promote high-dimensional action generation of the multi-robot systems involving human, an advanced DRL approach is developed. The approach leverages memory-enhanced representation learning, intrinsic reward-guided exploration, and action space pruning to better address the motion generation challenges. Empirical testing demonstrates the effectiveness of the proposed system, with experiments showing high success rates across tasks with varying team sizes and difficulty levels, thereby demonstrating applicability in human-robot collaborative manufacturing tasks. Note to Practitioners—The work reported in this paper is expected to be useful in applications requiring complex multi-robot motion generation within human-robot collaborative manufacturing cells, particularly in scenarios where the number and layout of robots may change, for example, low-volume high-mix manufacturing cells. Challenges in HRC robot motion generation are often compounded by the uncertainties associated with human worker movements and the interactions among robots in the cells. The proposed DRL-based motion policy leverages MR-assisted features to enhance state extraction capabilities and incorporates algorithmic modules to improve representation learning and exploration capabilities, thereby enhancing reactive motion generation for safe collaboration. The system demonstrates high success rates across tasks of varying complexities and team sizes. Owing to the flexible policy generation settings and the generalization capabilities of the perception approach, this approach is applicable to a variety of robotic cells and other close-proximity human-robot/machine collaborative motion generation scenarios with minimal tuning.}, archive = {J_TASE}, author = {Chengxi Li and Yue Yin and Hantao Ye and Pai Zheng and Satyandra K. Gupta}, doi = {10.1109/TASE.2025.3605990}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21033--21046}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A mixed-reality-augmented deep reinforcement learning approach for multi-robot safe motion generation in Human–Robot collaborative manufacturing cells}, volume = {22}, year = {2025}, }
• (2025). Robustness of panoptic segmentation for degraded automotive cameras data. TASE, 22, 21020--21032. (www)

  @article{ , abstract = {Precise situational awareness is essential for the safe deployment of artificial intelligence in real-world applications, particularly in assisted and automated driving (AAD) systems. Among perception techniques, panoptic segmentation is a promising technique to identify and categorise objects, impending hazards, and drivable space at a pixel level. While panoptic quality might be affected by automotive camera data quality, a comprehensive understanding and modelling of their relationship remains underexplored. Motivated by such a need, this work proposes a unifying pipeline to evaluate the robustness of panoptic segmentation models for automotive cameras, correlating it with 8 traditional image quality metrics (IQA). The proposed pipeline begins by generating a novel degraded dataset, D-Cityscapes+, featuring 19 realistic automotive degradation types at varying severity levels, including novel models for darkness and snowfall conditions with veiling effect. Evaluations on 14 state-of-the-art segmentation model backbones yielded key insights: 1) large-particle degradations (e.g., lens droplets, heavy snow) severely degrade segmentation performance, increasing uncertainty and edge-concentrated segmentation errors; 2) Transformer-based models outperform CNN models under adverse conditions; however, longer processing time, a higher number of parameters, and computational cost are limiting their real-world deployment; 3) frequency-based IQA metrics, such as CW-SSIM, strongly correlate with segmentation performance, serving as reliable predictive tools. 4) visual enhancements via restoration do not coherently benefit downstream segmentation tasks, underscoring the need for perception-specific restoration techniques. The benchmark and code:https://github.com/Warwick-Jocelyn/BRPS Note to Practitioners—Reliable situational awareness is vital for autonomous systems operating under adverse conditions and potentially degraded sensor data. This study shows that while Transformer-based architectures outperform traditional CNNs in robustness, their long processing time (often <2 FPS) and high computational costs (often >500 GFLOps) limit real-world deployment. We also find that simply increasing the quantity of training data without considering ‘noise coverage’ offers very limited improvement in model robustness under extreme weather conditions, due to the long-tail effect. These findings underscore the practical value of degraded datasets such as D-Cityscapes+, which simulate 19 realistic conditions at multiple severity levels. Moreover, frequency-based image quality metrics such as CW-SSIM are shown to correlate strongly with perception performance, offering a practical diagnostic tool for benchmarking segmentation robustness. The presented insights provide practitioners with a valuable benchmark, predictive metrics, and clear guidelines for developing robust, real-time-capable segmentation models suitable for challenging automotive environments. Future research could also explore multimodal sensing integration to enhance overall system robustness.}, archive = {J_TASE}, author = {Yiting Wang and Haonan Zhao and Daniel Gummadi and Mehrdad Dianati and Kurt Debattista and Valentina Donzella}, doi = {10.1109/TASE.2025.3605753}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21020--21032}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robustness of panoptic segmentation for degraded automotive cameras data}, volume = {22}, year = {2025}, }
• (2025). Integrated adaptive repetitive learning control of linear motor servo systems with periodic tasks. TASE, 22, 21010--21019. (www)

  @article{ , abstract = {Periodic repetitive tasks are common in industrial production, especially in intelligent equipment for batch processing. This paper studies the disturbance suppression problem of linear motor servo system (LMSS) in periodic trajectory tracking, and proposes an integrated adaptive repetitive learning control (IARLC) algorithm. Switching dynamics involving linear and nonlinear exponentials are constructed to achieve finite-time convergence or boundedness of the tracking error. For periodic uncertainty, a feedforward compensation term that does not depend on model prior information is designed, and a fully saturated learning law is used to limit the estimation error to a preset interval. Specifically, parameter estimation based on co-resonance reset and a projection-type adaptive algorithm are used to achieve compensation of physical model parameters. Non-periodic uncertainties and unknown disturbances are optimally attenuated by a robust feedback control law. The stability of the closed-loop system is analyzed based on the Lyapunov theory, and comparative experimental results verify the advantages of the proposed method in the presence of parametric and non-parametric uncertainties. Note to Practitioners—Modern industrial production usually uses linear motors as power components to benefit from their high precision and conversion efficiency. In addition, batch assembly and processing require manufacturing equipment to perform repetitive tasks over and over again. Based on this background, this paper studies the periodic trajectory tracking problem of LMSS and proposes a new finite time IARLC method. The controller is designed based on model compensation and robust stabilization, and the periodic dynamic characteristics are effectively processed using historical information, which reduces the complexity of system design. The legacy non-periodic disturbances and parameter time-varying effects are weakened by the designed feedback control law and parameter estimation algorithm. In particular, the linear and nonlinear error switching surface is constructed to accelerate the stability of the system. In the presence of parameter and non-parameter uncertainties, the stability and precise tracking of the closed-loop system are achieved, which has practical significance for practical engineering applications.}, archive = {J_TASE}, author = {Pengwei Shi and Yuzhe Zhao and Jiahu Qin and Xinghu Yu and Weichao Sun}, doi = {10.1109/TASE.2025.3609458}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {21010--21019}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated adaptive repetitive learning control of linear motor servo systems with periodic tasks}, volume = {22}, year = {2025}, }
• (2025). Non-force-sensing variable admittance control of lower limb rehabilitation exoskeleton robots using class κ∞ function-based adaptive sliding mode. TASE, 22, 20995--21009. (www)

  @article{ , abstract = {In this paper, a non-force-sensing variable admittance control approach is proposed for lower limb rehabilitation exoskeleton robots. This approach aims to provide satisfactory assistance and rehabilitation-training performance for users of such exoskeletons. Our method initiates with a novel fixed-time sliding mode observer that estimates human-machine interaction torques according to generalized momentum. This observer-based torque estimation strategy can estimate the interaction torque precisely, thereby enabling a non-force-sensing effect. Next, a variable admittance control framework is formulated for lower limb exoskeleton robots to ensure superior compliance. This framework comprises two essential components. First, a newly designed adaptive fixed-time sliding mode controller based on a class $\kappa _{\infty } $ function for the inner loop, which operates without requiring prior knowledge of the upper bound of lumped perturbations and guarantees precise gait trajectory-tracking performance. Second, a variable-parameter admittance model for the outer loop, which utilizes an exponential function to dynamically adjust the admittance parameters, thereby achieving a balance between the exoskeleton’s compliance and gait-correction efficacy. Finally, both simulation and experimental results are presented and analyzed to validate the effectiveness and superiority of the proposed non-force-sensing variable admittance control approach. Specifically, simulation results demonstrate that the root-mean-square (RMS) values of the inner-loop tracking errors for the proposed method are reduced by 26.1% and 20.4% at the hip and knee joints, respectively, compared with the top-performing benchmark algorithm. Meanwhile, the precision of the outer-loop observation is improved by 21% and 18% at these joints. Experimental validation further shows reductions of 14.2% and 20.1% in the RMS inner-loop tracking errors at the hip and knee joints, respectively, versus this benchmark algorithm. Note to Practitioners—Motivated by the problem of how to realize effective control of lower limb exoskeleton robots to provide the wearers with appropriate comfort and gait-correction effect, this paper formulates an adaptive variable admittance control framework. In the inner loop of this framework, a novel gait trajectory-tracking controller using adaptive sliding mode is designed to ensure high tracking accuracy. In the outer loop, a new adaptive observer is designed to estimate the human-robot interaction torque, and an adaptive admittance model is formulated to provide appropriate compliance for the robot. The proposed strategies are experimentally validated and can be utilized in real-word applications. The work of this paper has reference significance for the development of lower limb exoskeleton technologies.}, archive = {J_TASE}, author = {Zhe Sun and Tianyu Chai and Bo Chen and Hai Wang and Jinchuan Zheng and Zhihong Man}, doi = {10.1109/TASE.2025.3608042}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20995--21009}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Non-force-sensing variable admittance control of lower limb rehabilitation exoskeleton robots using class κ∞ function-based adaptive sliding mode}, volume = {22}, year = {2025}, }
• (2025). Stiffness-guided adaptive path planning for cable-driven dual-arm manipulators. TASE, 22, 20982--20994. (www)

  @article{ , abstract = {Owing to their lightweight design and structural flexibility, cable-driven dual-arm manipulators offer significant advantages in collaborative tasks. However, practical applications pose several challenges. The existence of environmental obstacles and the closed kinematic chain considerably increase the complexity of path planning. Furthermore, the inherently low stiffness of cable-driven systems can result in end-effector deformation, thereby reducing precision during path-following tasks. This paper proposes an adaptive, sampling-based path planning algorithm that addresses obstacle avoidance and closed kinematic chain constraint, facilitating efficient path planning in diverse operational environments. Additionally, an evaluation metric for the Cartesian stiffness of cable-driven dual-arm manipulators is introduced and incorporated into the proposed algorithm for optimization. This approach enhances both Cartesian stiffness and end-effector precision, thereby improving the reliability and accuracy of cable-driven dual-arm manipulators in practical applications. Note to Practitioners—This paper addresses the path planning problems of cable-driven dual-arm manipulators in practical applications, with a focus on obstacle avoidance, closed kinematic chain constraint, and stiffness optimization. Existing methods rely heavily on task-specific parameter tuning and struggle to find a valid path in complex environments. We propose a novel path planning algorithm with adaptive parameters capable of generating paths efficiently without violating the constraint. Despite variations in planning time across scenarios of different complexity, the proposed algorithm maintains superior efficiency compared to baseline methods in all experiments. An optimizing version of this algorithm is also designed to enhance the manipulator stiffness along the planned paths and reduce the end-effector deformation. While designed specifically for cable-driven dual-arm manipulator with closed kinematic chain and stiffness-optimization objectives, the proposed algorithm can be extended to general path planning problems with arbitrary constraints or optimization goals. Currently, the algorithm faces its limitations in highly obstructed environments, and the stiffness evaluation model does not account for complex characteristics of the cables. Future work could focus on improving planning efficiency through task decomposition strategies and developing more accurate stiffness models to better quantify end-effector deformation.}, archive = {J_TASE}, author = {Shuqing Dai and Bin Zhang and Shunxiang Pang and Weiwei Shang}, doi = {10.1109/TASE.2025.3607403}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20982--20994}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stiffness-guided adaptive path planning for cable-driven dual-arm manipulators}, volume = {22}, year = {2025}, }
• (2025). Trajectory and flow optimization for multi-part, multi-location pick-and-place tasks using nonlinear model predictive control. TASE, 22, 20966--20981. (www)

  @article{ , abstract = {This paper presents a comprehensive trajectory and flow optimization framework in multi-part, multi-location pick-and-place operations using Nonlinear Model Predictive Control (NLMPC). The proposed system enables a single robotic manipulator to execute multiple sequential pick-and-place actions across spatially distributed locations within a single operational cycle, significantly improving throughput and flexibility in industrial automation tasks. A central contribution of this work is the introduction of terminal cost penalization in the NLMPC formulation, targeting joint velocity and acceleration at the end of the trajectory to enable smooth and precise motion termination. Additionally, Euclidean distance constraints are incorporated to enhance the final pose accuracy of the end effector. The system is validated through extensive simulation experiments using a KINOVA Gen3 robotic arm in both obstacle-free and obstacle-present environments, where object and obstacle positions are predefined. Results show that penalizing the cost function improves end-effector precision, reducing Euclidean distance error by 35.9% in the obstacle-free case and 10.6% in the obstacle-present scenario. The NLMPC framework also maintains real-time feasibility, with an average computation time of 0.045 seconds per control update, well below the 0.55-second control loop interval. These findings confirm the practical viability of the proposed approach for high-performance, constraint-aware robotic control. Supplementary materials, including simulation videos and open-source MATLAB code, are provided to support reproducibility and future research. Note to Practitioners—Modern industrial automation systems demand flexible and precise robotic solutions capable of handling diverse objects in dynamic workspaces. This paper presents an NLMPC-based trajectory planning strategy tailored for multi-part, multi-location pick-and-place operations. The proposed approach supports flexible, precise, and collision-free robotic motion, making it well-suited for advanced material handling applications. Incorporating penalties on terminal velocity and acceleration leads to smoother stops, increased placement precision, and reduced mechanical wear, while applying Euclidean distance constraints ensures accurate final positioning of the end-effector. The method is computationally efficient and runs in real time on a standard CPU. Although the simulations assume known object locations, the framework is extensible to real environments with integrated perception. Open-source code and videos are provided to support adoption and replication.}, archive = {J_TASE}, author = {Amanuel Abrdo Tereda and Sun Yi and Jagannathan Sankar and Yimesker Yihun and Richard Holdbrook}, doi = {10.1109/TASE.2025.3607109}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20966--20981}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory and flow optimization for multi-part, multi-location pick-and-place tasks using nonlinear model predictive control}, volume = {22}, year = {2025}, }
• (2025). SWIFT: A distributed one-stage planner for efficient multi-quadrotor trajectory optimization. TASE, 22, 20951--20965. (www)

  @article{ , abstract = {This paper presents SWIFT (Swarm-Wise Inference for Fast Trajectory Planning), a distributed one-stage planner designed for efficient multi-quadrotor trajectory optimization in cluttered environments. SWIFT unifies depth-based perception, interaction-aware modeling, and trajectory prediction into a single lightweight network, enabling decentralized real-time planning without reliance on global maps. Each quadrotor processes its local depth observation and asynchronously received neighbor intentions to predict multiple candidate trajectories, leveraging a confidence-guided cross-attention mechanism to selectively model critical inter-agent interactions. The proposed framework achieves sub-millisecond planning latency while ensuring collision avoidance, dynamic feasibility, and cooperative behavior under limited communication. Furthermore, a visual-inertial odometry drift correction strategy based on inter-agent visual detections enhances positioning accuracy during real-world flights. Extensive simulations and forest experiments validate that SWIFT outperforms existing baselines in trajectory smoothness, success rate, scalability, and robustness across diverse scenarios. Note to Practitioners—This work presents a fully decentralized system that enables a team of quadrotors to navigate cluttered environments without GPS, global maps, or centralized control. Each quadrotor plans its trajectory in real time using onboard depth sensing and lightweight asynchronous communication with nearby teammates. The proposed SWIFT planner quickly evaluates multiple motion options through a single neural network pass, enabling safe, smooth, and dynamically feasible trajectories with sub-millisecond latency. A collaborative visual-inertial odometry correction module further improves positioning accuracy by using visual detections between agents. This system is well suited for tasks such as search and rescue, warehouse logistics, and environmental monitoring, where fast and reliable navigation in complex GPS-denied settings is critical. Current limitations include reduced adaptability to unknown moving obstacles and challenges under extremely high-speed flight. However, the SWIFT architecture is designed to be extensible, allowing future integration of dynamic obstacle handling, predictive interaction modeling, and swarm-level objectives such as formation control or cooperative tracking.}, archive = {J_TASE}, author = {Haowen Wang and Shuting Zhang and Yuyang Sun and Zhaohui Wang and Junren Sun and Bocheng Zhu}, doi = {10.1109/TASE.2025.3606896}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20951--20965}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SWIFT: A distributed one-stage planner for efficient multi-quadrotor trajectory optimization}, volume = {22}, year = {2025}, }
• (2025). Designing pinning control synchronization scheme for inertial memristive neural networks with dynamic granger causality. TASE, 22, 20934--20950. (www)

  @article{ , abstract = {Understanding the communication and information processing mechanisms in pinning control is highly influenced by the connectivity structure of the network, particularly in higher-order networks. However, existing studies predominantly focus on static network structures and utilize complete error information, often neglecting underlying inter-node dependencies, thereby limiting the effectiveness of low-energy control in higher-order dynamic systems. To address these limitations, this paper proposes a novel pinning control scheme for the synchronization of inertial memristive neural networks (IMNNs) based on dynamic Granger causality analysis (DGCA). First, an interpre table IMNN model is constructed to explicitly characterize higher-order interactions without decomposing the system into first-order subsystems. Then, unlike traditional pinned-node selection strategies relying on static interaction rules, a causality-aware selection algorithm is developed using DGCA to dynamically identify influential nodes via time-series analysis, enhancing control efficiency under dynamic network conditions. Furthermore, a local information-based pinning controller is designed by leveraging local causal relationships and control influence regions extracted from DGCA, ensuring the stability of the synchronization error system. Theoretical guarantees are provided by deriving sufficient conditions for controller design based on Lyapunov stability theory. Finally, three numerical examples are presented to demonstrate the effectiveness and practicality of the proposed scheme. Note to Practitioners—In practical applications, the dynamic behaviors of memristive neural networks (MNNs) are often influenced by both overdamped and underdamped responses, which play a critical role in generating complex or even chaotic dynamics. These characteristics are particularly evident in inertial MNNs (IMNNs). Therefore, this study addresses the synchronization issue of interpretable IMNNs directly, without resorting to variable substitution. Additionally, efficiently utilizing error information for pinning control is crucial for reducing energy consumption and enhancing control performance. However, strong coupling and correlation between network nodes are inherent in IMNNs and cannot be neglected. Therefore, how can the dynamic correlations among nodes be utilized to improve the selection of pinned nodes in a scalable manner? How can synchronization error information be selectively integrated to reduce unnecessary communication and energy overhead? These questions puzzled the authors and prompted this study.}, archive = {J_TASE}, author = {Manman Yuan and Jiapei Li and Yunzhou Li and Wenjun Xiong and Guanrong Chen}, doi = {10.1109/TASE.2025.3603010}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20934--20950}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Designing pinning control synchronization scheme for inertial memristive neural networks with dynamic granger causality}, volume = {22}, year = {2025}, }
• (2025). Hierarchical control for robust standing stability and fall recovery of task-performing humanoid robots. TASE, 22, 20920--20933. (www)

  @article{ , abstract = {Maintaining robust standing stability is critical for humanoid robots performing precision manipulation tasks, where traditional balance strategies (e.g., stepping) can disrupt task execution. Designing a controller that ensures both stability and effective disturbance rejection is challenging, as these goals impose conflicting requirements on body sway and stepping behavior. To address this, we propose a hierarchical control framework that decouples these objectives into three parts: 1) a Task-Oriented Standing Policy trained via reinforcement learning with strict penalties on stepping and body sway, specialized for non-stepping disturbance rejection; 2) Ankle Strategy Modulation embedded within the standing policy, which dynamically adjusts the Zero-Moment Point (ZMP) to eliminate Divergent Component of Motion (DCM) errors, thereby accelerating recovery from disturbances; and 3) a fall prediction and recovery mechanism that triggers a transition to a robust stepping policy when disturbances exceed the stability threshold. Extensive experiments on a Q-series humanoid robot, in both simulation and hardware, validate the controller’s effectiveness. We achieved a 100% task success rate over 100 trials in practical scenarios including robot archery and part grasping/placement. Note to Practitioners—This study addresses a critical challenge in deploying humanoid robots for manipulation tasks: maintaining operational continuity during frequent upper-body movements, where traditional balance strategies—such as stepping—can disrupt workflow. We propose a hierarchical control framework that resolves the inherent conflict between non-stepping stability and fall recovery through three integrated components: a reinforcement learning-based standing policy enforcing minimal body sway and zero foot movement, ankle-based Zero-Moment Point (ZMP) modulation for real-time Dynamic Capture Model (DCM) error correction, and a fall prediction mechanism that triggers stepping recovery when necessary. This architecture enables separate training of non-stepping and stepping strategies, improving both stability and disturbance resilience. Experimental validation demonstrates 100% task success over 100 trials across realistic tasks, such as archery and part manipulation. Limitations include simplified sensory, and disturbance models and the decoupled control of upper and lower body motions, which constrain whole-body coordination. Future work will focus on improving multimodal perception and implementing whole-body cooperative control for more complex tasks.}, archive = {J_TASE}, author = {Chiyu Zhang and Jihe Bai and Ziyu Chen and Jie Gao and Hong Qiao}, doi = {10.1109/TASE.2025.3608186}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20920--20933}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical control for robust standing stability and fall recovery of task-performing humanoid robots}, volume = {22}, year = {2025}, }
• (2025). Lyapunov stability-driven control algorithm for heterogeneous multi-robot coordination. TASE, 22, 20906--20919. (www)

  @article{ , abstract = {Recent advancements in autonomous swarm systems have made a pivotal point in robotic science. Utilising a large-scale swarm of simple robots to accomplish complex tasks offers efficient, robust, and reliable solutions inspired by natural phenomena. Although bio-inspired methodologies have presented competent algorithms, those approaches inspired by the physical interactions in viscoelastic materials demonstrate more structured methods to prove the stability and robust performance of the algorithms mathematically. This paper proposes a new viscoelastic swarm algorithm which applies to heterogeneous swarm systems. In this paper, the algorithm development utilises the Lyapunov method to determine stability criteria and corresponding conditions. Therefore, the resulting approach does not rely on complex optimisation to obtain the parameters that guarantee stable performance. In addition to the theoretical framework, a series of Monte Carlo simulations have been conducted to assess the algorithm’s performance and its sensitivity to the key variables. Furthermore, the algorithm’s performance has been evaluated by a series of experiments with real robots to examine the effect of different variables, such as neighbourhood conditions and the stiffness coefficient, on the algorithm’s output. The results obtained from the simulations and experiments demonstrate the stable and bounded performance of the algorithm and how the key variables, such as stiffness coefficient and number of neighbours for each robot, affect the swarm performance. The comparison results, obtained from real-world experiments with a state-of-the-art algorithm, show that the proposed framework significantly reduces the control effort for the robots while improving the swarm behaviour. Note to Practitioners–This paper presents a viscoelastic swarm control algorithm for heterogeneous multi-robot coordination, emphasising stability, robustness, and reduced computational effort. Swarm robotics addresses complex tasks using simple autonomous agents, offering adaptability and resilience. However, practical deployment faces challenges in ensuring system stability and optimising performance under resource constraints. The proposed algorithm applies the Lyapunov method for a theoretical stability proof. Its robustness was validated through Monte Carlo simulations and real-world experiments using the Mona robot platform. Key findings demonstrate that increasing stiffness coefficients enhances swarm cohesiveness but may induce fluctuations in alignment, while a larger number of neighbours and scouts improves alignment stability and robustness. Practitioners in automation, robotics, and related fields can benefit from this work by applying the algorithm to distributed systems requiring efficient coordination under uncertain conditions, such as intelligent transportation and warehouse automation. The approach reduces control effort, a critical factor for battery-operated robots, and offers a systematic framework for parameter tuning based on interaction graph properties. Future work will extend this framework to dynamic interaction networks, extending applicability to real-world scenarios with time-varying constraints.}, archive = {J_TASE}, author = {Fatemeh Rekabi-Bana and Mazen Bahaidarah and Ognjen Marjanovic and Farshad Arvin}, doi = {10.1109/TASE.2025.3606553}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20906--20919}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lyapunov stability-driven control algorithm for heterogeneous multi-robot coordination}, volume = {22}, year = {2025}, }
• (2025). Automated top stitching via vision-based macro-mini approach: Retrofitting legacy machines for enhanced precision in garment manufacturing. TASE, 22, 20891--20905. (www)

  @article{ , abstract = {The apparel industry is increasingly adopting automation technologies and investing in research to address global reshoring efforts and rising labor costs. One essential process for automation in garment factories is top stitch, which involves sewing two fabric sheets together along a seam line while maintaining a consistent distance from it. This paper presents an automated top stitch system that retrofits a legacy machine using a vision-based macro-mini approach. The system mainly comprises a two-degree-of-freedom automatic sewing machine as the legacy machine, a vision sensor module, a mini-actuator module, and an infrared sensor module. The vision sensor module detects the seam line on the fabric, even in the presence of positional uncertainties, such as, human errors. The infrared sensor module monitors the sewing sequence, while the mini-actuator module operates simultaneously with the sewing machine to compensate for any misalignment. The system demonstrates improvements, with the average and the maximum errors reduced by 79% and 67%, respectively, and the standard deviation improved by 73%, compared to those from the operation without our system. To the best of our knowledge, this is the first development of an autonomous top stitch system using a retrofitted legacy machine, presenting a novel architecture that integrates sensors and actuators through a macro-mini approach. Note to Practitioners—This paper implements key technologies essential for automating top stitch. Notably, since conventional vision recognition methods struggle to distinguish a seam line, an asymmetric lighting environment and a specialized algorithm were developed to address the challenge. Additionally, the system architecture was designed with consideration for upgrading existing legacy machines, enabling seamless integration without requiring full equipment replacement. While this paper focuses on the practical challenge of top stitch automation, the proposed architecture can be extended beyond sewing applications. It can be applied to other precision-driven processes that rely on vision-based object differentiation, such as welding or dispensing. This approach is particularly beneficial for enterprises hesitant to adopt new equipment without prior validation, offering a cost-effective alternative to full system replacements. Given the expenses associated with dismantling, disposing of, and replacing existing machinery, introducing modular upgrades provides a practical means of verify process improvements while minimizing expenses. Although the design and performance of additional modules may vary depending on the legacy machine being retrofitted, substantial improvements can still be achieved. Experimental results demonstrate that this approach reduces the average error by 79% and enhances the standard deviation by 73%, validating its feasibility from both engineering and economic perspectives.}, archive = {J_TASE}, author = {Taehwan Kim and HyunWoong Choi and Jaejin Kim and Byung-Hyun Song and Ho-Young Kim and Yong-Lae Park}, doi = {10.1109/TASE.2025.3605235}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20891--20905}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated top stitching via vision-based macro-mini approach: Retrofitting legacy machines for enhanced precision in garment manufacturing}, volume = {22}, year = {2025}, }
• (2025). Obstacle-avoiding path planning for robotic manipulators based on recursive segmentation point migration optimization and progressive inverse kinematics. TASE, 22, 20876--20890. (www)

  @article{ , abstract = {This paper introduces a novel robotic path planning method that integrates Recursive Segmentation Point Migration Optimization (RSPMO) with Progressive Inverse Kinematics (PIK). RSPMO continuously refines the path by detecting collisions on a direct trajectory, adjusting segmentation points to avoid obstacles, and simplifying the path by removing redundant points, thereby rapidly generating an optimal and efficient trajectory for the manipulator’s end-effector. PIK utilizes second derivative information to iteratively compute joint angles, employing dynamic interpolation and damping factor adjustments near anomalous configurations to enhance system stability. Extensive experimental validation confirms the efficacy of both methods. In a multi-obstacle 2D scenario, RSPMO can plan paths of 789.4516 units in merely 0.0170 seconds, significantly outperforming comparative methods. Additionally, PIK achieves robust performance, with average iteration times of 0.0041 seconds near singular points over 8.45 iterations. Real-world tests on a 6-DOF robotic arm further confirm that the integration of RSPMO and PIK provides an efficient solution for path planning in complex scenarios. Note to Practitioners— This study addresses the practical challenge of path planning for robotic manipulators in complex and confined environments where collision-free navigation is essential. Traditional methods often face issues such as high computational complexity, suboptimal path smoothness, and limited real-time applicability, which can hinder production efficiency. Our proposed solution, combining Recursive Segmentation Point Migration Optimization (RSPMO) and Progressive Inverse Kinematics (PIK), enhances obstacle avoidance and path planning efficiency in complex scenarios. This approach not only minimizes path length but also dynamically adjusts joint angles near anomalies, ensuring smoother and more stable trajectories. This method is highly relevant for industrial applications, as it reduces planning time, improves system stability, and lowers mechanical wear and maintenance costs. However, practical challenges remain, such as the need for fine-tuning PIK near anomalies, especially for nonstandard manipulators where the current damping adjustment may not fully adapt. Future work will focus on developing smarter obstacle detection and dynamic path adjustment mechanisms to maintain efficient path planning in densely cluttered environments.}, archive = {J_TASE}, author = {Na Li and Zhenguo Gao and Yun Ouyang and Yule Zeng}, doi = {10.1109/TASE.2025.3604604}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20876--20890}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Obstacle-avoiding path planning for robotic manipulators based on recursive segmentation point migration optimization and progressive inverse kinematics}, volume = {22}, year = {2025}, }
• (2025). An adaptive-gain reinforcement iterative learning control scheme for nonlinear systems to track morphologically similar targets. TASE, 22, 20863--20875. (www)

  @article{ , abstract = {This paper investigates the problem of tracking morphologically similar targets for nonlinear systems and proposes an adaptive-gain reinforcement iterative learning control (AG-RILC) scheme. Unlike existing control approaches, where gain schemes are predefined before the control process, this study incorporates a reinforcement learning (RL) mechanism into a novel iterative learning control (ILC) framework. Specifically, this novel ILC scheme can address the tracking problem for morphologically similar targets using the condition-performance matching algorithm and time-scale transformation. Further, an adaptive-gain scheme is designed to adapt to the control performance metrics by introducing a weighting factor that integrates the decreasing gain sequences of ILC with RL’s powerful exploratory capabilities. In addition, it is proved that under the proposed scheme, the control input error converges to zero with the number of iterations approaches infinity. Finally, illustrative simulations demonstrate that the AG-RILC scheme achieves a balance between high tracking accuracy and fast convergence. Experimental outcomes validate the performance benefits of the AG-RILC algorithm. Note to Practitioners—Many engineering systems, such as rail transportation systems, robotic systems, and chemical process systems, require tracking repetitive tasks with morphological similarity over finite time intervals. Iterative learning control (ILC) is a well-established approach for such repetitive control tasks. However, many existing ILC methods for nonlinear systems rely on constant or predefined decreasing learning gains before the learning process, which prevents real-time adaptation to control performance. As a result, these methods struggle to effectively balance fast convergence with high tracking accuracy. In this paper, we overcome the difficulty of learning parameter tuning in classical ILC algorithms, and draw inspiration from reinforcement learning (RL) algorithm to propose an adaptive-gain reinforcement iterative learning control (AG-RILC) scheme for nonlinear systems. The scheme strikes an optimal balance between fast convergence and accurate tracking performance, resulting in superior overall control performance.}, archive = {J_TASE}, author = {Chunjun Chen and Lu Yang and Kexin Yin and Yaowen Zhang}, doi = {10.1109/TASE.2025.3604450}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20863--20875}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptive-gain reinforcement iterative learning control scheme for nonlinear systems to track morphologically similar targets}, volume = {22}, year = {2025}, }
• (2025). Energy-efficient four-mode electromagnetic mini valve with single-coil drive: Modular design and application in pneumatic soft actuators. TASE, 22, 20854--20862. (www)

  @article{ , abstract = {In minimally invasive surgery, pneumatic soft actuators have gained significant attention due to their promising applications. Valves, which are critical components in pneumatic driving systems, typically operate in only two modes in current research. In multi-actuator systems, this limitation results in a large number of valves and a bulky driving system, which restricts the system’s portability. In this study, we report a four-mode, multi-stable, energy-efficient electromagnetic mini valve. A coil serves as the driving source. The valve has three outlets, enabling it to control up to three actuators independently. The moving components include a cylindrical magnet and a ring magnet. They are concentrically arranged and can only move along the axial direction. These magnets can realize four distinct positions based on the electromagnetic force and their mutual magnetic force. Moreover, the multi-stable design reduces energy consumption. The two cores make the magnets stable when no current is applied. The valve exhibits a shortest response time of 4.75 ms and an energy cost of 0.007 J. It can handle a maximum flow rate of 9.7 L/min at a pressure of 200 kPa, with a maximum back pressure of 100 kPa. Its energy-efficient and four-operating-mode characteristics prove its potential in pneumatic soft actuator systems. A modular design by connecting multiple mini valves reduces the number of pumps and valves, thereby lowering the overall size and weight of pneumatic driving systems. Note to Practitioners—Pneumatic soft actuators are widely used in automation systems because they are flexible and safe. However, controlling multiple actuators often needs many valves, making the system large and heavy. This work introduces a small valve with four operating modes. It can control up to three actuators by itself. The valve holds its position without continuous power, which saves energy. It also responds quickly and can deliver a high airflow. By connecting several of these valves, users can build simpler and lighter pneumatic systems. This valve is useful for portable automation devices, medical robots, and wearable systems. In the future, the valve design could be connected to smart controllers for more advanced automatic control.}, archive = {J_TASE}, author = {Jingjia Zhu and Pengfei Han and Hubiao Fang and Jie Hu and Guofang Gong and Huayong Yang and Fuzhou Niu and Dong Han}, doi = {10.1109/TASE.2025.3603647}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20854--20862}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy-efficient four-mode electromagnetic mini valve with single-coil drive: Modular design and application in pneumatic soft actuators}, volume = {22}, year = {2025}, }
• (2025). Personalized federated learning-based distributed model predictive control with predictive error compensation for nonlinear networked systems. TASE, 22, 20837--20853. (www)

  @article{ , abstract = {This paper develops a personalized federated learning-based distributed model predictive control (PFL-DMPC) method with predictive error compensation (PEC) for nonlinear networked systems with multiple subsystems to preserve data privacy and achieve the desired control performance despite the presence of modeling error of federated learning (FL) models. First, a personalized FL (PFL) algorithm is presented to obtain a global model using dynamic alternating iteration mode to update and aggregate personalized models, which preserves data privacy and addresses the heterogeneity issue among subsystems. Then, the convergence of the PFL algorithm with strongly convex property is proved, showing a faster convergence rate and higher computational efficiency than the existing FL algorithms. Next, since PFL models are integrated into a distributed model predictive control (DMPC) scheme, a DMPC scheme with PEC is developed to reduce the steady-state offset by incorporating the estimated prediction error into optimization problems. Finally, the simulation results for a nonlinear chemical process network demonstrate the convergence of the PFL algorithm and the superiority of the PFL-DMPC method with PEC compared to the PFL-DMPC method without accounting for model-plant mismatch. Note to Practitioners—This work addresses challenges in controlling nonlinear networked systems with multiple interconnected subsystems, particularly in ensuring privacy, managing heterogeneity, and compensating for prediction errors. These challenges often arise in industrial automation settings such as chemical processing, where subsystems exhibit diverse dynamics and data privacy is critical. We propose a PFL-DMPC framework with PEC. Unlike traditional centralized approaches, this PFL approach preserves data privacy by training models locally in each subsystems and sending them to a server to aggregate into a global model, thereby addressing subsystem-specific heterogeneity. The PEC mechanism corrects steady-state offsets caused by modeling inaccuracies, which ensures the system’s output aligns closely with desired trajectories. Practitioners can benefit from the faster convergence and improved adaptability of this framework, particularly in scenarios with limited communication bandwidth or data privacy concerns. However, implementing PFL-DMPC requires sufficient computational resources for iterative optimization and may demand fine-tuning of hyperparameters. Future work may explore extending this approach to real-time applications with stringent timing constraints. Potential applications include distributed control of multi-vehicle systems, smart grids, and robotic networks, where privacy-preserving and learning-based control is vital.}, archive = {J_TASE}, author = {Zeyuan Xu and Zhe Wu}, doi = {10.1109/TASE.2025.3608014}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20837--20853}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Personalized federated learning-based distributed model predictive control with predictive error compensation for nonlinear networked systems}, volume = {22}, year = {2025}, }
• (2025). 3D flexible locomotion of magnetic soft flapping milliswimmers based on adaptive MPC. TASE, 22, 20825--20836. (www)

  @article{ , abstract = {Magnetic soft flapping milliswimmers hold significant potential for biomedical applications due to their flexible mobility and multifunctionality. However, advanced motion control tasks for such forms remain underexplored, hindering their flexible mobility and various biomedical applications. The multi-physics coupling inherent in the flapping swimming process presents substantial challenges for dynamics modeling and control. In this study, we fabricated a magnetic soft flapping millirobot and conducted dynamics analysis using rigid-body modeling and quasi-static wing-flapping analysis method. The resulting state-space model is employed to build a navigation control framework based on adaptive Model Predictive Control (MPC) method, realizing autonomous navigation and obstacle avoidance performances. To validate the effectiveness of the proposed control framework, the complex path-following task in 3D space is performed, demonstrating superior performance compared to geometry-based following methods. Furthermore, the robot’s navigation ability in complex environments is evaluated in both physical and simulated setups featuring static and dynamics obstacles. The milliswimmer’s 3D flexible mobility and its capability to operate effectively in challenging and intricate environments may reveal great potential for real-world biomedical and other advanced applications. Note to Practitioners—This paper is motivated by the problem of precise motion control task of the magnetic soft flapping milliswimmer but the method also applies to other magnetic torque driven swimmers. Existing literature has not fully investigated the modeling and control strategy for the magnetic driven flapping swimming motion. This paper suggests a modeling method to describe the magnetic field driven propulsion from body-environment interactions and proposed a control framework which is suitable for the mechanism. We then demonstrate the efficacy of the method with a series of experiments on advanced motion control tasks. In future research, we will explore how instantaneous magnetic actuation dynamics shape the robot’s motion profile, with the goal of improving both the generalization capability and precision of the proposed approach.}, archive = {J_TASE}, author = {Ruokun He and Yunlang Zhou and Zhiang Zhang and Bing Han and Hesheng Wang and Zhuo-Chen Ma}, doi = {10.1109/TASE.2025.3600421}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20825--20836}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {3D flexible locomotion of magnetic soft flapping milliswimmers based on adaptive MPC}, volume = {22}, year = {2025}, }
• (2025). Joint design for hybrid bumps suppression and disturbance attenuation of switched fuzzy systems under bump-dependent dynamic event-triggering. TASE, 22, 20814--20824. (www)

  @article{ , abstract = {In this paper, the anti-disturbance bumpless transfer control problem is addressed for switched fuzzy systems without stable subsystems under event-triggering. First, we establish an original characterization of the anti-disturbance hybrid bumpless transfer performance to simultaneously attenuate disturbances and abrupt large hybrid bumps caused by both switching and triggering. Second, a novel dynamic event-triggered scheme that incorporates triggering-induced bumps and switched turning parameters is designed to improve bumpless transfer performance further. Then, solvability conditions are developed for the considered anti-disturbance hybrid bumpless transfer control problem. Moreover, the bump-dependent dynamic event-triggered scheme, a state-dependent switching law, and disturbance-observer-based event-triggered controllers are designed simultaneously. Finally, a tunnel diode circuit model and a numerical example are used to verify the effectiveness of our method. Note to Practitioners—The undesired abrupt large control bumps and disturbances commonly exist in practical systems with switching dynamics, such as microgrids, aeroengines, and circuit systems, degrading both transient performance and steady-state performance. Accordingly, this paper focuses on simultaneous suppression of hybrid control bumps and disturbances in event-triggered switched fuzzy systems. By proposing anti-disturbance hybrid bumpless transfer performance, we establish a new framework for bumpless transfer control to cope with the challenges posed by hybrid control bumps as well as disturbances. Furthermore, unlike traditional event-triggered strategies, we design a novel bump-dependent dynamic event-triggered mechanism to further suppress the control bumps and improve bumpless transfer performance. Finally, a co-design scheme integrating the event-triggered mechanism, switching law, and hybrid bumpless transfer controllers is presented.}, archive = {J_TASE}, author = {Siyuan Zhang and Jun Zhao}, doi = {10.1109/TASE.2025.3607490}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20814--20824}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Joint design for hybrid bumps suppression and disturbance attenuation of switched fuzzy systems under bump-dependent dynamic event-triggering}, volume = {22}, year = {2025}, }
• (2025). Modular soft exoskeleton design and control for assisting movements in multiple lower limb joint configurations. TASE, 22, 20799--20813. (www)

  @article{ , abstract = {Soft lower limb exoskeletons have shown potential to enhance mobility in patients with motor disorders and improve their quality of life. This study introduces a novel modular soft lower-limb exoskeleton that provides assistance to a single hip joint or/and a single ankle joint, configurable on either the left or right side according to clinical needs. The control system employs adaptive gait phase recognition and a hierarchical strategy to accurately identify user intentions and track desired assistance profiles. An assistance profile optimization method is also designed based on the users’ muscle surface electromyography signals, aiming to further enhance the adaptability of robotic assistance to different users and tasks. Multi-subject experiments were conducted with 10 participants. The experimental results demonstrated that the exoskeleton exhibited good modular performance and effectively assisted walking under different configurations. Specifically, compared to natural walking, the exoskeleton assistance reduced the net metabolic rate by up to 15.29% while the hip and ankle joints are assisted simultaneously. Furthermore, the optimized assistance profiles achieved lower metabolic rates and reduced muscle activation compared to the generic assistance. Note to Practitioners—Lower-limb exoskeletons are becoming an essential tool for improving mobility in individuals with motor impairments, yet many existing systems lack the flexibility to adapt to varying user needs and movement patterns. This work addresses this challenge by introducing a modular wearable suit and targeted optimized assistance profile to soft lower-limb exoskeletons that allow for personalized assistance at the hip and ankle joints. The system is designed to be highly flexible, enabling different combinations of joint assistance based on the user’s specific requirements, making it suitable for a wide range of rehabilitation and assistive applications. Stroke survivors with unilateral challenges or patients requiring partial joint support can benefit from this approach. To achieve this, the system employs an adaptive gait phase recognition method and hierarchical control strategy to ensure accurate tracking of user intentions and the use of corresponding assistance profiles. Additionally, the profiles are optimized using surface electromyography signals, enhancing the system’s ability to adjust to individual users and tasks. Experiments demonstrated that the use of the exoskeleton reduced the respiratory metabolic rate and muscle activation of the subjects while walking, and it also had good modular performance and could provide diverse assistance. While promising, current testing is limited to controlled environments. Further research is needed to validate the system in real-world scenarios, assess long-term durability, and explore multi-patient experiments. Future work will focus on enhancing multi-joint coordination, long-term stability, and integration into broader rehabilitation settings.}, archive = {J_TASE}, author = {Bi Zhang and Weizhong Jiang and Xiaowei Tan and Juhua Su and Yiwen Zhao and Xingang Zhao}, doi = {10.1109/TASE.2025.3607351}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20799--20813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modular soft exoskeleton design and control for assisting movements in multiple lower limb joint configurations}, volume = {22}, year = {2025}, }
• (2025). Distributed output feedback consensus control for nonlinear multiagent systems under output event-triggered communication. TASE, 22, 20788--20798. (www)

  @article{ , abstract = {This paper focuses on the leader-following consensus control problem for nonlinear multiagent systems (MASs) under output event-triggered communication. A novel distributed discontinuous backstepping control strategy is presented, which utilizes information exclusively from intermittent output instants. First, a distributed adaptive event-triggered mechanism (ETM), along with a continuous-discrete time compensator and observer are designed jointly to compensate for consensus errors and reconstruct the system state, where all parameters designed are flexibly chosen. The triggering sampling instants are asynchronous and aperiodic, eliminating the necessity for continuous neighbor information monitoring. Second, we introduce a dynamic variable into the coordinate transformation to overcome the challenge of intermittent output in backstepping design, which is the key to address input errors caused by triggering mechanisms. Compared with the widely used backstepping method using the first-order filter to address intermittent signals, the proposed scheme does not require additional triggering design for the filter and achieves a full-state consensus in the global sense. Theoretical analysis shows that the system is asymptotically stable, all consensus errors converge to zero, and Zeno behavior is excluded. Finally, simulation examples are presented to demonstrate the effectiveness of the theoretical result. Note to Practitioners—This paper addresses the challenge of reducing communication costs in multi-agent systems (e.g., autonomous vehicles, sensor networks) where frequent data exchange strains network resources. Traditional continuous or periodic communication methods waste bandwidth and energy. Our solution employs output event-triggered communication, enabling agents to transmit only when necessary. By integrating adaptive triggering thresholds with a simplified control design, we eliminate complex filters or continuous monitoring while ensuring full-state consensus (e.g., synchronized positions/speeds). This approach enhances scalability for large networks and resource-limited systems like drone swarms or smart grids. Future extensions could address unknown disturbances as well as consider time-varying and switching topological networks. The method is ideal for applications prioritizing energy efficiency, such as autonomous logistics, environmental monitoring, or industrial automation.}, archive = {J_TASE}, author = {Hao Li and Changchun Hua and Kuo Li}, doi = {10.1109/TASE.2025.3605716}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20788--20798}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed output feedback consensus control for nonlinear multiagent systems under output event-triggered communication}, volume = {22}, year = {2025}, }
• (2025). HIGHSTAR: High-speed and efficient online autonomous UAV exploration. TASE, 22, 20770--20787. (www)

  @article{ , abstract = {Unmanned aerial vehicles (UAVs) are widely used in autonomous exploration, but their motion speed is underutilized due to inaccurate motion time cost evaluation and high computational cost. Existing methods either fail to consider UAV’s motion tendency and environment simultaneously or can’t ensure real-time planning in large 3-D environments. This paper presents a consistent, high-speed, and efficient online autonomous UAV exploration method. First, a motion primitive activated graph search method is proposed to fully take advantage of the UAV’s current velocity and acceleration. It improves motion time cost evaluation by simulating short-term motion tendencies with motion primitives and reduces the computational cost by searching on a voxel graph with a dynamic upper bound. Then, a minimum time trajectory to the optimal viewpoint with a non-zero terminal velocity constraint in a convex hull is optimized. Finally, an SE(3) coverage trajectory for unknown space around the exploration path is further optimized. Simulations in various environments with different speed settings show that the proposed method’s average UAV velocity is $18.6\%-116.1\%$ faster and its exploration efficiency is $13.8\%-49.5\%$ higher than state-of-the-art methods. Real-world tests verify its effectiveness. The source code of our method will be released at: https://github.com/NKU-MobFly-Robotics/HighStar Note to Practitioners—This paper proposes a real-time UAV planning method for high-speed exploration of large unknown 3-D environments. The key challenge the proposed method addresses is how to exploit the high-speed potential of UAVs in exploration. The proposed method uses a motion primitive activated graph search method to take advantage of the current UAV speed and acceleration in exploration target planning and plans an informative coverage trajectory. The proposed method achieves real-time and high-speed exploration planning in simulations. The experiments conducted in structured underground parking and large wild outdoor environments show the effectiveness of the proposed method.}, archive = {J_TASE}, author = {Qianli Dong and Xuebo Zhang and Shiyong Zhang and Ziyu Wang and Zhe Ma and Tianyi Li and Haobo Xi}, doi = {10.1109/TASE.2025.3605375}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20770--20787}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {HIGHSTAR: High-speed and efficient online autonomous UAV exploration}, volume = {22}, year = {2025}, }
• (2025). Design and control of a dual-drug-chamber drug delivery capsule robot for multi-target drug delivery. TASE, 22, 20759--20769. (www)

  @article{ , abstract = {With the advancement of capsule endoscopy technology, the treatment of gastrointestinal (GI) tract diseases has entered a new era. Capsule robots enable the diagnosis and treatment of lesions in a painless and minimally invasive manner. Research on highly controllable Drug Delivery Systems (DDSs) using capsule endoscopy is significant for treating GI tract diseases. Despite the variety and complexity of DDS designs, most systems lack the capability for multi-target drug delivery and simultaneous carriage of multiple drugs. This paper proposes a dual-drug-chamber drug delivery capsule robot (DDCR) that utilizes a single Internal Permanent Magnet (IPM) for both drug delivery and propulsion. The design of the dual-drug-chamber primarily aims to carry one or two drugs simultaneously for treating one or multiple target sites. The proposed DDCR uses a balloon mechanism for drug containment. An external magnetic drive system activates a needle attached to a piston, which punctures the balloon and, thereby, releases the medication. This mechanism ensures rapid and effective coverage of lesions by the drugs. Based on both theoretical and experimental results regarding balloon drug loading and the control distance of the external permanent magnet (EPM) control distance, it was established that a drug load of 0.3 ml per chamber was most suitable for the design, and the feasibility of the dosing method was demonstrated. The size (29mm in length and ~13mm in diameter) of the DDCR was also found to be suitable for biological experiments, including multi-target drug delivery. By continuously adjusting the distance between the DDCR and the EPM in porcine small intestine samples, the optimal driving distance and drug delivery distance were found to be 80mm and 140mm. Note to Practitioners—The proposed dual-drug-chamber DDCR can potentially address the limitations of current systems in multi-drug and multi-target GI treatments. The use of a single internal permanent magnet (IPM) for both propulsion and drug release enables precise, minimally invasive delivery of one or two drugs to different target sites within the GI tract. Its promising design with a compact structure and balloon-based drug release mechanism provides practical value for treating conditions like Crohn’s disease and gastric ulcers and offers greater flexibility and control compared to traditional methods. The external magnetic control system ensures accurate positioning and drug release, making the DDCR a promising tool for targeted therapy. Despite its potential, there are practical limitations, including the need for enhanced real-time control and adaptation to various patient anatomies. The current study demonstrates feasibility, but further research is necessary to refine the system’s accuracy in dynamic GI environments and ensure its effectiveness in clinical applications. These developments could help advance capsule endoscopy and drug delivery technology in the near to mid-term, providing more efficient and patient-friendly treatment options.}, archive = {J_TASE}, author = {Bo Ye and Zhi Shu and Shufang Wang and Bo Wang and Guangzheng Gao and Shoujun Dai and Sixian Liu and Yihan Xu and Princy Randhawa and Lalit Garg and Amit Krishna Dwivedi and Sheng Liu}, doi = {10.1109/TASE.2025.3604374}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20759--20769}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and control of a dual-drug-chamber drug delivery capsule robot for multi-target drug delivery}, volume = {22}, year = {2025}, }
• (2025). CODN-GS: Coupled optimization of depth and normal in 3D gaussian splatting for scene reconstruction. TASE, 22, 20746--20758. (www)

  @article{ , abstract = {Scene reconstruction has attracted widespread attention due to its extensive applications in intelligent devices. Recently, 3D Gaussian Splatting (3DGS) has gained recognition as a prominent technique owing to its impressive pixel-level rendering quality and high reconstruction speed. However, 3DGS scenes constructed with only photometric supervision are highly prone to RGB overfitting. This problem neglects the depth of objects and the surface normal in the reconstructed scene, which leads to the lack of geometric consistency. To address these issues, this study introduces a coupled optimization of depth and normal within the 3D Gaussian Splatting (CODN-GS) framework. Firstly, a normal–depth–normal transformation is applied to ensure accurate capture of geometric features in the reconstructed scenes. Secondly, a robust monocular depth supervision model generates depth maps that are refined via global and local adjustments, which serve as guidance for the model to accurately learn scene geometry. Thirdly, a normal supervision model is incorporated as a complement to depth supervision, jointly optimizing the overall scene geometry. Finally, comprehensive experiments on the Replica, MipNerf360, and ScanNet datasets demonstrate that CODN-GS reduces RMSE-D and RMSE-N by at least 9% and 13%. These results confirm that the proposed method outperforms state-of-the-art methods in both depth and normal accuracy. Note to Practitioners—While 3D Gaussian Splatting (3DGS) has shown promise for scene reconstruction, achieving accurate geometric consistency with purely image-based methods remains challenging. Existing approaches primarily optimize 3DGS using RGB information, with some incorporating depth and normal constraints. However, these methods struggle in complex regions, such as textureless and reflective surfaces, which often lead to geometric distortion. To address these issues, we introduce CODN-GS (Coupled Optimization of Depth and Normal within 3D Gaussian Splatting), a novel reconstruction framework designed to improve geometric accuracy without the need for additional sensors. CODN-GS first applies a surface refinement transformation to improve the model’s ability to learn depth and normal information. It then derives depth and normal maps for each input RGB image using transfer learning and fine-tuning, providing robust geometric supervision for reconstruction. We evaluated CODN-GS on three public datasets and found that it significantly improves geometric consistency while maintaining high photometric reconstruction accuracy.}, archive = {J_TASE}, author = {Wentao Hu and Ke Feng and Xin Ye and Huafeng Ding and Long Wen}, doi = {10.1109/TASE.2025.3607491}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20746--20758}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {CODN-GS: Coupled optimization of depth and normal in 3D gaussian splatting for scene reconstruction}, volume = {22}, year = {2025}, }
• (2025). Event-triggered-based fault-tolerant output synchronization of heterogeneous multi-agent systems under denial-of-service attacks. TASE, 22, 20737--20745. (www)

  @article{ , abstract = {This paper studies the problem of dynamic event-triggered output synchronization for heterogeneous multi-agent systems (MASs) under asynchronous Denial-of-Service (DoS) attacks and actuator hybrid faults. First, an innovative dynamic event-triggered estimator abandoning the information of the attacked channels is proposed to estimate the leader’s state under asynchronous attacks, with stability of the estimation error system ensured based on a novel lemma combined with Lyapunov stability theory. Furthermore, a decentralized fault-tolerant controller is developed based on the estimator’s state, which integrates adaptive parameters to compensate for actuator hybrid faults. It is manifested that the output synchronization can be achieved with the designed fault-tolerant controller. Compared with the existing studies assuming simultaneous attacks on all channels, this paper considers each channel’s attack to be independent. Finally, a simulation example not only shows that the dynamic event-triggered mechanism can effectively lengthen the minimum interevent interval due to the internal dynamical variable but also substantiates the effectiveness of the proposed control scheme. Note to Practitioners—In this paper, the output synchronization problem is addressed for heterogeneous MASs, which play significant roles in smart power grids, uncrewed aerial vehicles (UAVs) cluster combat, and so on. When a channel is subjected to DoS attacks, the information of that channel is unavailable. Thus, an estimator that discards the information of the attacked channels is proposed. In addition to potential attacks at the cyber layer, the physical layer may undergo actuator faults due to factors such as equipment aging. Consequently, a fault-tolerant controller is developed to compensate for actuator hybrid faults. Moreover, the dynamic event-triggered mechanism is utilized to save communication resources, and the Zeno behavior is excluded.}, archive = {J_TASE}, author = {Kaiyue Dong and Guang-Hong Yang}, doi = {10.1109/TASE.2025.3600659}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20737--20745}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered-based fault-tolerant output synchronization of heterogeneous multi-agent systems under denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Less conservative h∞ loop-shaping control design and implementation for low-damping piezo-positioning stages under varying loads. TASE, 22, 20726--20736. (www)

  @article{ , abstract = {Due to hysteresis nonlinearity and load variation-induced low-damping resonance drift, the piezo-positioning stages struggle to achieve high-precision and high-speed tracking control. To address this issue, this paper proposes a less conservative $H_{{\infty }}$ loop-shaping controller design method based on the $\nu $ -gap metric. Resonance drift caused by load variations is modeled as coprime factor uncertainty, and the uncertainty boundaries are quantified by computing the $\nu $ -gap between the nominal model and frequency responses under varying load conditions. A parameter selection range is established based on the relationship between the robust stability margin and the $\nu $ -gap to guide the design of the weighting function, enhancing control performance while ensuring robust stability under varying load conditions. The $H_{{\infty }}$ LS controller is implemented using normalized coprime factor decomposition, which eliminates resonance peaks and reduces step response overshoot while preserving disturbance rejection capability. Experimental results under both no-load and 300 g load conditions show that, compared to the traditional $H_{{\infty }}$ LS, the proposed method reduces overshoot by 9.4% and 14.6%, and shortens settling time by 46.4% and 25.4%. Compared to the $\mu $ -synthesis method, the proposed method increases bandwidth by 68.4 Hz and 144.1 Hz, and decreases the maximum relative tracking error by 40.6% and 42.5%. These results validate the superiority of the proposed method. Note to Practitioners—Piezo-positioning stages have been widely employed in atomic force microscopes due to their high resolution and rapid response characteristics. However, load variations can induce low-damping resonance drift, which degrades tracking accuracy and system stability. The proposed method specifically addresses low-damping resonance drift caused by load variations. The design of the weighting function is guided by the relationship between the model $\nu $ -gap and the robust stability margin, which significantly reduces the conservatism and trial-and-error tuning in traditional $H_{{\infty }}$ LS methods. Notably, the proposed method imposes no strict constraints on model structure or order, and can be implemented using only basic knowledge of frequency-domain shaping, making it highly practical for engineering applications. It is particularly suitable for piezo-positioning systems and other high-precision motion control platforms with load disturbances, resonance drift, and stringent performance requirements.}, archive = {J_TASE}, author = {Haotian Liu and Shiyu Duan and Guilin Zhang}, doi = {10.1109/TASE.2025.3606029}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20726--20736}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Less conservative h∞ loop-shaping control design and implementation for low-damping piezo-positioning stages under varying loads}, volume = {22}, year = {2025}, }
• (2025). OPEN: Lightweight map-based semantic navigation for GPS-free last-mile delivery. TASE, 22, 20713--20725. (www)

  @article{ , abstract = {The growing demand for efficient last-mile delivery highlights the need for autonomous robots to improve operational efficiency and reduce costs. Traditional navigation methods rely on high-precision maps, which are expensive to produce and maintain, while learning-based approaches often struggle to adapt to diverse real-world environments. To address these challenges, this paper presents OpenStreetMap-enhanced oPen-air sEmantic Navigation (OPEN), a system that combines foundation models with classic navigation algorithms to enable scalable outdoor navigation. By leveraging off-the-shelf OpenStreetMap (OSM), OPEN eliminates the need for extensive pre-mapping and provides a lightweight, readily available map representation. The system uses Large Language Models (LLMs) to interpret delivery instructions and Vision Language Models (VLMs) for global localization without relying on GPS, real-time map updates, and entrance recognition, ensuring robust navigation in complex environments. To further enhance adaptability, OPEN incorporates a local replanning method that dynamically adjusts waypoints in response to environmental changes and OSM inaccuracies. Since existing benchmarks do not adequately reflect the challenges of last-mile delivery, this work introduces a new benchmark designed for residential navigation. Experiments conducted in both simulated and real-world settings demonstrate that OPEN improves navigation accuracy, efficiency, and reliability, outperforming existing semantic navigation methods. To facilitate further research, the code and benchmark are publicly available at https://github.com/EI-Nav/light-map-navigation.git Note to Practitioners—Autonomous robot navigation for last-mile delivery faces significant challenges, including high mapping costs, environmental variability, and unreliable or unavailable GPS signals in urban areas, all of which hinder large-scale deployment. Traditional high-precision maps are expensive to create and maintain, while learning-based approaches often struggle to generalize to new and dynamic environments. Additionally, traditional navigation systems lack the ability to comprehend and execute tasks based on natural language instructions, limiting their flexibility and interactivity. To address these challenges, this work introduces an innovative framework that eliminates the need for costly pre-mapping by leveraging OSM data and foundation models for efficient navigation. The system utilizes LLMs to interpret delivery tasks expressed in natural language, offering a more intuitive and interactive user experience, while VLMs enable GPS-free localization, real-time map updates, and entrance recognition, enhancing adaptability in complex environments. Additionally, a local replanning mechanism dynamically adjusts waypoints when OSM data proves inaccurate, ensuring greater safety and reliability. In practical applications, OPEN provides a low-cost, highly adaptable solution that allows robots to operate in diverse environments without extensive pre-mapping efforts, while its ability to process natural language instructions significantly enhances usability. However, its effectiveness depends on the completeness and accuracy of OSM data, and integrating higher-quality maps from providers like Google or Baidu could further improve navigation performance. Beyond last-mile delivery, this framework holds promising potential for applications in drone logistics, warehouse automation, and intelligent transportation systems, paving the way for scalable and cost-effective autonomous mobility solutions.}, archive = {J_TASE}, author = {Junhui Wang and Dongjie Huo and Yongliang Shi and Chao Gao and Yan Qiao and Guyue Zhou}, doi = {10.1109/TASE.2025.3605356}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20713--20725}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {OPEN: Lightweight map-based semantic navigation for GPS-free last-mile delivery}, volume = {22}, year = {2025}, }
• (2025). Nav-SCOPE: Swarm robot cooperative perception and coordinated navigation. TASE, 22, 20702--20712. (www)

  @article{ , abstract = {This paper proposes a lightweight decentralized solution for multi-robot coordinated navigation with cooperative perception. First, we introduce a rapid way to process sensory data, thus obtaining safe directions and key environmental features. Then, an information flow is created to facilitate real-time perception sharing over wireless ad-hoc networks. Consequently, the environmental uncertainties of each robot are reduced by interaction fields that deliver complementary information. Finally, path optimization is achieved in a probabilistic way, enabling self-organized coordination with effective convergence, divergence, and collision avoidance. Our method is fully interpretable and ready for deployment without gaps. Comprehensive simulations and real-world experiments demonstrate reduced path redundancy, robust performance across various tasks, and minimal demands on computation and communication. Note to Practitioners—Local perception is the information source for robot navigation in unknown environments. The extended perception at the swarm level provides complementary information for each robot. This can optimize robot paths and achieve coordinated navigation in a distributed way. However, this purpose is currently unachievable due to the high demands of communication and computation on board. This study reduces these two demands to the extreme by Fast Fourier Transform (FFT), digital filtering, probabilistic fusion, and swarm interaction forces. At the same time, our fully interpretable method avoids sim-to-real gaps and can be deployed directly on ground mobile robots. Simulation and real-world experiments outperform a state-of-the-art approach. Future work aims to extend our method to aerial robots as drones.}, archive = {J_TASE}, author = {Chenxi Li and Weining Lu and Qingquan Lin and Litong Meng and Haolu Li and Bin Liang}, doi = {10.1109/TASE.2025.3604178}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20702--20712}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nav-SCOPE: Swarm robot cooperative perception and coordinated navigation}, volume = {22}, year = {2025}, }
• (2025). Attack-resilient distributed control of multi-agent systems under output feedback-driven multi-channel DoS attacks. TASE, 22, 20691--20701. (www)

  @article{ , abstract = {This paper investigates the distributed resilient output-tracking control problem for heterogeneous multi-agent systems (MASs) subject to multi-channel denial-of-service (DoS) attacks. First, we propose a simplified framework to address the distributed output tracking control for systems with uncertain dynamics. Then, we develop a distributed output feedback-driven resilient control protocol under asynchronous multi-channel DoS attacks, considering both inter-follower communication channels and leader-follower links. To enhance resilience against attacks, a set of channel-dependent decay rates is introduced and designed to ensure self-healing output tracking capabilities under active attacks. Unlike conventional distributed output control methods that heavily depend on distributed observers and assume the leader’s state is globally accessible or transmitted via distributed communication, our resilient control algorithm eliminates the need for distributed observers while reducing communication overhead through compact output information encoding. Finally, the effectiveness of the proposed algorithm is demonstrated through a practical example involving a multi-vehicle system. Note to Practitioners—For practical systems such as ground vehicles, aerial vehicles, and robots, the dynamics often exhibit heterogeneity and uncertainty due to real-world operational conditions. To ensure robust tracking performance under such challenges, this paper investigates resilient secure tracking control for heterogeneous and uncertain multi-agent systems (MASs) under distributed denial-of-service (DoS) attacks. Unlike existing methods that rely on distributed observers requiring the leader’s full state, we propose a novel output feedback-driven resilient control framework for heterogeneous and uncertain systems to address challenges arising from model uncertainty and limited information availability in practice. The proposed distributed resilient control algorithm is applicable to formation control of autonomous ground vehicles, aerial vehicles, and intelligent robots. Its advantages lie in avoiding the transmission of observed variables and enabling lightweight communication. Additionally, an attack immunity mechanism is integrated to counter multi-channel DoS attacks. A set of equivalent decay rates is defined to provide sufficient conditions for secure tracking. Finally, the effectiveness of the proposed algorithm is demonstrated through a practical example involving a multi-vehicle system.}, archive = {J_TASE}, author = {Meng-Ying Wan and Yong Xu and Zheng-Guang Wu}, doi = {10.1109/TASE.2025.3604414}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20691--20701}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attack-resilient distributed control of multi-agent systems under output feedback-driven multi-channel DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Semi-supervised privacy-preserving EEG-based motor imagery classification via self and adversarial training. TASE, 22, 20679--20690. (www)

  @article{ , abstract = {Electroencephalogram (EEG)-based motor imagery (MI) signals are frequently used in brain-computer interfaces (BCIs) due to their wide applications in the rehabilitation field. However, cross-subject variations often result in a model trained on one participant failing when applied to another. Additionally, privacy concerns regarding sensitive health and mental information in EEG-based MI signals further complicate the situation. Source-free domain adaptation aims to address these cross-subject variations by transferring knowledge from a source domain (i.e., a previous participant) to a target domain (i.e., a new participant) without accessing sensitive source data. However, source-free unsupervised domain adaptation models often face issues with incorrect pseudo-labels, which can lead to unstable and ineffective adaptation. To address this, we propose a source-free semi-supervised domain adaptation algorithm for EEG-based MI signal classification. This algorithm tackles noise accumulation caused by incorrect pseudo-labels while effectively handling data distribution variations and privacy concerns, similar to source-free unsupervised domain adaptation models. Specifically, we train the classifier head using only a limited amount of labeled target data to prevent noise accumulation, and generate pseudo-labels for the unlabeled target data. Furthermore, we introduce an independent self-training head that learns better representations using the generated pseudo-labels, mitigating overfitting caused by the limited labeled target data. Additionally, we design an adversarial head that plays a minimax game to extract more discriminative feature representations from the unlabeled target data. Extensive experiments on three benchmark datasets, compared with eighteen state-of-the-art SFDA methods, demonstrate the superiority of our approach. Note to Practitioners—The practical problem addressed in this work is the challenge of classifying electroencephalogram (EEG)-based motor imagery (MI) signals in brain-computer interfaces (BCIs) across different subjects, which is crucial for applications such as stroke and paralysis recovery. Our solution introduces a source-free semi-supervised domain adaptation algorithm that leverages a small amount of labeled target data to prevent noise accumulation caused by incorrect pseudo-labels, a common issue in unsupervised domain adaptation. By incorporating a self-training head and an adversarial head, we improve feature representation and reduce overfitting, leading to more accurate and reliable classification. This approach enhances the efficiency and reliability of BCI systems by reducing the need for extensive labeled data, ensuring privacy, and improving generalization capabilities across different subjects. The key insights include proposing the first semi-supervised source-free domain adaptation method for MI signal classification, developing a self-training head to mitigate noise accumulation, and designing an adversarial head to optimize feature extraction adversarially. Future work could explore further reducing computational cost, extending the method to other BCI signals, and investigating more advanced neural network architectures. The results can be extended to real-time BCI systems, personalized medical diagnostics, and enhancing AI-driven health monitoring systems.}, archive = {J_TASE}, author = {Jian Zhu and Ganxi Xu and Zhizhe Lin and Jinyi Long and Teng Zhou and Bin Sheng and Xiaokang Yang}, doi = {10.1109/TASE.2025.3604283}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20679--20690}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semi-supervised privacy-preserving EEG-based motor imagery classification via self and adversarial training}, volume = {22}, year = {2025}, }
• (2025). A nonlinear trajectory tracking approach of double-pendulum aerial transportation systems: Transient performance improvement and saturated inputs. TASE, 22, 20667--20678. (www)

  @article{ , abstract = {This paper considers aerial cargo transportation tasks in the presence of double-pendulum swing effects. Due to the higher degree of freedom and the more intricate dynamic coupling of the system, an effective nonlinear trajectory tracking control method is proposed to effectively track a given trajectory and simultaneously suppress the swing motion during tracking. The controller is designed to achieve improved transient tracking performance and generate amplitude-saturated inputs. Specifically, a tracking performance improvement term is elaborately designed through a dynamic generator, achieving more accurate trajectory tracking with bounded overshoots. Also, saturated functions are induced to provide bounded inputs, being able to deal with input saturation problems. Using the Lyapunov stability analysis and LaSalle’s invariance theorem, the stability of the system is rigorously proven. The author’s search of relevant literature shows that the proposed method is the first method of a nonlinear anti-swing trajectory tracking controller for double-pendulum aerial transportation systems in three-dimensional space. The tracking performance, swing suppression, and robustness against external disturbances are verified through comparative simulation. Note to Practitioners—The utilization of quadrotors for aerial transportation has become a prevalent method. A cable securely fastens the hook to the quadrotor fuselage, suspending the cargo. This configuration establishes quadrotor air transportation systems with the advantageous of flexible, light weight, and large load. Due to the absence of direct forces acting on the hook and load, coupled with a high degree of freedoms of the system, a controller is essential. This controller is elaborately designed to mitigate the swing of the hook and cargo during flight while ensuring satisfactory trajectory tracking performance. This paper addresses these challenges by proposing a nonlinear trajectory tracking method. The system is engineered to follow a predefined trajectory, suppresses hook and cargo swing during tracking, and dynamically adjusts parameters to maintain input within a specified range, aligning with real-world operational conditions.}, archive = {J_TASE}, author = {Yiming Wu and Yuchun Li and Jiuxiang Dong}, doi = {10.1109/TASE.2025.3604026}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20667--20678}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A nonlinear trajectory tracking approach of double-pendulum aerial transportation systems: Transient performance improvement and saturated inputs}, volume = {22}, year = {2025}, }
• (2025). Distributed potential game optimization to 3-path vertex cover of networks. TASE, 22, 20649--20666. (www)

  @article{ , abstract = {3-path vertex cover of networks is a typical optimization problem in network science, which has a wide range of applications. Toward a 3-path vertex cover of networks from distributed optimization, we first established a potential game to describe the 3-path vertex cover problem. Next, we analyze the inherent relationship between potential game and 3-path vertex cover, that is, only the solution to minimum value of potential function are minimum 3-path vertex covered solutions, and strict Nash equilibriums are intermediate solutions between minimum 3-path vertex covered solutions and 3-path vertex covered solutions. Then, we propose a bounded best response and memory-based distributed algorithm, and prove that our proposed algorithm can guarantee any initial solution converge to a strict Nash equilibrium, and further analyze the complexity of this algorithm. Finally, numerical simulations verify the effectiveness and superiority of our proposed algorithm on some representative networks and benchmark by comparing with existing representative algorithms.This work paves an effective way for distributed optimization that could be modeled as distributed potential game. Note to Practitioners—The 3-path vertex cover problem has a wide range of practical applications. Many optimization algorithms for the vertex cover problem are available in the existing literature. However, those algorithms are essentially centralized algorithms. Thus, one of the main challenges is to develop a distributed algorithm for solving the 3-path vertex cover problem. For this, this work establishes a potential game for the 3-path vertex cover problem, and propose a bounded best response and memory-based distributed algorithm, and prove that it can converge to a strict Nash equilibrium, and further analyze the complexity of this algorithm. In addition, our proposed distributed algorithm can favor better strict Nash equilibrium through the adjustment of bounded strength $\epsilon $ and memory length m. This work can serve as a supplement to the existing works for solving the 3-path vertex cover problem.}, archive = {J_TASE}, author = {Jie Chen and Jie Wu and Rongpei Zhou and Weihua Gui}, doi = {10.1109/TASE.2025.3602818}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20649--20666}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed potential game optimization to 3-path vertex cover of networks}, volume = {22}, year = {2025}, }
• (2025). Anomaly monitoring of dynamic wastewater regeneration process based on recursive broad learning system. TASE, 22, 20636--20648. (www)

  @article{ , abstract = {In the wastewater regeneration process, a condition monitoring system must have dynamic adaptability to new operating conditions while maintaining a lasting adaptability to existing conditions to ensure efficient and stable operation. This paper proposes a Slow Feature Recursive Broad Learning System (SF-RBLS) to achieve incremental learning of new knowledge and sustained adaptation to condition characteristics. The model takes advantage of the incremental learning mechanism of the broad learning system, enabling online learning of unknown features as new operating conditions emerge. An innovative coupling mechanism between the slow feature space and the recursive structure is constructed. Through the slow feature analysis mechanism, temporal constraints are imposed on the dynamic evolution process of the recursive window, enabling the feature extraction process to simultaneously generate long-term steady state representations and short-term dynamic responses, achieving enduring adaptability to various conditions. Experimental results demonstrate that SF-RBLS exhibits significant advantages over state-of-the-art methods in both accuracy and stability for real-time monitoring and anomaly detection, confirming its effectiveness and potential in intelligent wastewater reclamation monitoring. Note to Practitioners—Monitoring wastewater regeneration processes is challenging due to dynamic environmental conditions, sudden pollution events, and changing operational states. Traditional models struggle to adapt to these variations, leading to reduced accuracy in fault detection. To address this, we propose a Slow Feature Recursive Broad Learning System (SF-RBLS) that combines slow feature analysis with a recursive mechanism to capture both long-term trends and short-term fluctuations. In addition, an incremental learning strategy allows the model to update in real time without retraining from scratch. The experimental results show that SF-RBLS outperforms existing methods in detecting anomalies such as toxicity shocks, habitat destruction, and sludge bulking. This approach offers a more adaptive and efficient solution for intelligent wastewater monitoring, with potential applications in other industrial process monitoring tasks.}, archive = {J_TASE}, author = {Chang Peng and Lu Yan}, doi = {10.1109/TASE.2025.3601377}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20636--20648}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Anomaly monitoring of dynamic wastewater regeneration process based on recursive broad learning system}, volume = {22}, year = {2025}, }
• (2025). Surgeon supervised autonomous surgical system for oral and maxillofacial surgery. TASE, 22, 20624--20635. (www)

  @article{ , abstract = {Oral and maxillofacial surgery (OMS) imposes an increasing workload on even the most experienced surgeons due to long operation time, high skill requirements, limited observation field, constrained workspace, and fast-growing patient population. Robot-assisted OMS is particularly challenging, requiring technological advancements to replicate complex surgical workflows executed by human surgeons and novel working concepts to properly address human-machine relationships. We introduced a Surgeon Supervised Autonomous Surgical System (SSASS) aiming to solve emerging bottlenecks in OMS. SSASS custom develops a deep-learning-assisted virtual planning module, a teeth-based monocular camera navigation module, and a six-degree-of-freedom compact robot module to function as surgeons’ auxiliary brain, eye, and hand, respectively. These three modules are further seamlessly integrated to autonomously complete most labor-intensive procedures, while prioritizing surgeons to supervise and be responsible for the overall procedure. Le Fort I experiments on five human head models demonstrated that the surgical results of SSASS closely matched the preoperative plan, with high drilling accuracy and acceptable cutting accuracy under a fundamentally new and significantly simplified surgical workflow. Compared to its existing OMS counterparts, SSASS integrates the latest technologies such as deep learning, medical 3D printing, markerless navigation, virtual reality, and collaborative robotics, providing a comprehensive surgical solution for encompassing the entire OMS loop. Note to Practitioners—This study was motivated by the increasing needs of automating the surgical procedure of oral and maxillofacial surgery (OMS) to reduce the mismatch between the inadequate number of professional surgeons and the fast-growing patient population. Although more and more surgical hardware and software were introduced into the operating room to assist the surgeon, their workloads are only partially alleviated as surgeons still have to perform many vital procedures manually. This study proposed a Surgeon Supervised Autonomous Surgical System (SSASS) with a significantly simplified workflow for achieving predictable, controllable, and repeatable OMS outcomes. SSASS could avoid many previously necessary analyzing procedures and surgical tools, freeing surgeons from the need to conduct time-consuming manual cephalometric analysis, prefabricate physical surgical models, primary robot programming knowledge or repeatedly program robot trajectory in every surgery, and use physical tracking markers and frequently measure the separated skeleton parts. SSASS introduced a systematical surgical solution aiming to finish most high-workload, tedious, repetitive work while giving the highest priority to surgeons for surveillance. We disclosed key software and hardware information of this system in the manuscript and supplementary materials so that the practitioners could rebuild or improve our current prototype without major difficulties. The experiment results show the high application potential of this system in OMS.}, archive = {J_TASE}, author = {Qingchuan Ma and Etsuko Kobayashi and Kazuaki Hara and Junchen Wang and Ken Masamune and Hideyuki Suenaga and Yubo Fan}, doi = {10.1109/TASE.2025.3604891}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20624--20635}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Surgeon supervised autonomous surgical system for oral and maxillofacial surgery}, volume = {22}, year = {2025}, }
• (2025). RASP: Robot active scene perception with joint viewpoint planning and depth completion in cluttered environments. TASE, 22, 20611--20623. (www)

  @article{ , abstract = {Capturing dense visual perception in cluttered environments using depth sensors is crucial for downstream robotics tasks. However, occlusions between objects and unreliable depth data make it very challenging for robots to perceive comprehensive and accurate scene information. To address these issues, we propose a novel robot active scene perception method called RASP, which is composed of two parts. First, we introduce a temporal attention-based view planning algorithm, which actively plans the minimum feasible viewpoint sequence based on latent dependencies in all previous observations to maximize the information perception of the cluttered environments. Subsequently, for the unreliable depth data obtained by the depth sensor, especially caused by transparent and specular objects in the scene, we design a geometry-guided depth completion network that fully utilizes the 3D scene information during the progressive perception process. Specifically, multi-level scene geometric features are extracted and projected into the image space, combining with the image features to guide the depth completion step. These two parts are learned jointly to achieve consistent and accurate results. Extensive experiments demonstrate that our method outperforms state-of-the-art methods. Furthermore, we show that our method significantly improves the performance of downstream grasping tasks. Note to Practitioners—With the development of depth sensing technology, consumer-level depth cameras have been widely used in many robotics and automation tasks. However, in cluttered environments, due to the occlusions between objects, it is difficult for depth cameras to capture comprehensive scene information from a single perspective. In addition, depth sensors often generate unreliable depth data when facing transparent and specular objects. To address these challenges, this article proposes a robot active scene perception method that combines view planning and depth completion. Our view planning strategy actively explores the minimum feasible viewpoint sequence to maximize the acquisition of scene information, while the depth completion network accurately restores unreliable depth information. Experimental results show that our method performs well in cluttered environments containing transparent and specular objects. The proposed method provides high-quality perception input for robots to perform intelligent tasks in practice.}, archive = {J_TASE}, author = {Yizhe Liu and Tong Jia and Haiyu Zhang and Guowei Yang and Hao Wang and Dongyue Chen}, doi = {10.1109/TASE.2025.3604455}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20611--20623}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RASP: Robot active scene perception with joint viewpoint planning and depth completion in cluttered environments}, volume = {22}, year = {2025}, }
• (2025). Edge delayed deep deterministic policy gradient: Efficient continuous control for edge scenarios. TASE, 22, 20596--20610. (www)

  @article{ , abstract = {Deep Reinforcement Learning (DRL) has emerged as a powerful paradigm for learning complex policies directly from high-dimensional input spaces, enabling advances across a variety of domains. Modern DRL algorithms often rely on dual-network Q-learning architectures to approximate optimal policies to overcome overestimation bias. Recent research has introduced approaches leveraging multiple Q-functions to further mitigate overestimation effects and enhance policy reliability. However, there is a growing emphasis on deploying DRL in edge scenarios, where privacy concerns and stringent hardware constraints necessitate highly efficient algorithms. In such environments, the computational and memory efficiency of learning methods is of critical importance. In this context, we propose Edge Delayed Deep Deterministic Policy Gradient (EdgeD3), a novel reinforcement learning algorithm specifically designed for edge computing settings. EdgeD3 offers significant reductions in GPU time (by 25%) and computational and memory usage (by 30%), while consistently achieving or surpassing the performance of state-of-the-art algorithms across multiple benchmarks and in real-world tasks. Note to Practitioners— Driven by the growing need for efficient computational solutions in automation, this research introduces the Edge Delayed Deep Deterministic Policy Gradient (EdgeD3), a novel Deep Reinforcement Learning algorithm designed for applications on edge devices with a limited computational budget. EdgeD3 is developed to enable on-device execution of policy learning, when computing resources are at a premium, such as in smart manufacturing and autonomous vehicle systems. EdgeD3 can learn highly effective policies, on par of state-of-the-art algorithms, while utilizing significantly fewer resources. This would empower autonomous devices to operate independently of cloud-based systems, fostering faster operational speeds and enhancing data privacy.}, archive = {J_TASE}, author = {Alberto Sinigaglia and Niccoló Turcato and Ruggero Carli and Gian Antonio Susto}, doi = {10.1109/TASE.2025.3604290}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20596--20610}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Edge delayed deep deterministic policy gradient: Efficient continuous control for edge scenarios}, volume = {22}, year = {2025}, }
• (2025). Risk-aware informative path planning for information gathering of a 3D surface. TASE, 22, 20584--20595. (www)

  @article{ , abstract = {Surface information acquisition by robots faces challenges such as sensor uncertainty, limited resources, and dynamic environment, all of which often lead to reduced collection efficiency and accuracy. To address these issues, this paper proposes a Risk-aware Informative Path Planning (RIPP) framework. The framework is capable of adaptively selecting the target region according to the mutual information between the expected detection viewpoints. The uncertainty risk caused by noisy sensing can be effectively managed using the Conditional Value at Risk (CVaR)-based method. Therefore, a CVaR-based Greedy Algorithm (CGA) is proposed to select the optimal set of inspection viewpoints. To further enhance information acquisition efficiency, the drone’s path is optimized using a novel Adaptive Fractional Particle Swarm Optimization (AFPSO) algorithm. This approach enables the drone to autonomously select trajectories rich in high-value information. This framework is evaluated in the context of 3D surface temperature inspection of large storage tanks. Simulation and experimental results show that RIPP framework significantly reduces information reconstruction errors in tank surface inspections by demonstrating clear advantages over existing methods. The effectiveness and feasibility of RIPP framework in surface information acquisition task are verified, which provides a new solution for efficient monitoring in complex environment. Note to Practitioners—This study is to propose a risk-aware online information path planning method to enhance the efficiency of information collection by autonomous aerial vehicles (AAVs) under uncertainty-induced risks. Conventional information path planning methods typically overlook the correlation between observation viewpoints and fail to address the risks posed by uncertainties, focusing merely on uncertainty estimation. In contrast, the proposed method effectively manages risks associated with uncertainties and optimally plans AAV information collection paths. Performance evaluations demonstrate that the proposed approach outperforms several existing methods.}, archive = {J_TASE}, author = {Mengfei Xu and Yang Chen and Mian Hu and Yanhua Yang}, doi = {10.1109/TASE.2025.3604174}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20584--20595}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Risk-aware informative path planning for information gathering of a 3D surface}, volume = {22}, year = {2025}, }
• (2025). Depth enhancement mask mapping network with multi-teacher distillation for RGB-D scene parsing. TASE, 22, 20572--20583. (www)

  @article{ , abstract = {The proposed method in the study, called DEMMNet-KD, aims to address the challenges of scene parsing in RGB-D indoor scenes. The study recognizes the importance of lightweight models for pixel-intensive tasks like scene parsing, and proposes a depth enhancement mask mapping network (DEMMNet) with multi-teacher knowledge distillation approach to achieve both reduced model size and high accuracy. DEMMNet introduces a new segmentation paradigm called class-position separate segmentation to reduce the difficulty of delineating and classifying the segmentation region. It also includes a depth spatial enhancement module based on the attention mechanism to effectively utilize depth information in a single-stream network design. To improve performance without increasing complexity, DEMMNet-KD utilizes Mix Transformer encoder (MiT) as a backbone and employs multi-teacher knowledge distillation. This allows semantic knowledge to be transferred from large teacher networks with different performances to the student network. The experiments conducted on NYUDv2 and SUN RGB-D datasets demonstrate that DEMMNet-KD achieves state-of-the-art performance with fewer additional learnable parameters and minimal computational effort. Utilizing MiT-B2 as the backbone network, the DEMMNet-KD method attained mean Intersection over Union (mIoU) scores of 54.71% and 51.03% on the two datasets, respectively, thereby achieving an optimal balance between accuracy and efficiency. Furthermore, the scalability of this method has been demonstrated across various domains and modalities. Overall, the proposed DEMMNet-KD approach provides an efficient solution for scene parsing in RGB-D indoor scenes, enhancing the capabilities of bionic binocular robots in perceiving their environments and serving human society effectively. The code is available at https://github.com/SHARKALAKALA/DEMMNet Note to Practitioners—This study presents an efficient network known as the Depth Enhancement Mask Mapping Network (DEMMNet) and a multi-teacher knowledge distillation (KD) approach that utilizes several teacher networks to transfer knowledge to the student network. Its encoding component differs from those employed in most other scene parsing methods, as it uses only the RGB modality to extract multi-level semantic features through a backbone network. The depth modality is directly integrated with the multi-level semantic features extracted from the RGB modality using the DSE module, bypassing any processing by the backbone network. The CPSS module introduces a lightweight segmentation framework that partitions the segmentation task into two components: region partitioning and region classification.}, archive = {J_TASE}, author = {Yuxiang Xiao and Jiajun Meng and Fangfang Qiang and Xiena Dong and Wujie Zhou}, doi = {10.1109/TASE.2025.3603265}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20572--20583}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Depth enhancement mask mapping network with multi-teacher distillation for RGB-D scene parsing}, volume = {22}, year = {2025}, }
• (2025). Integrated nesting and scheduling for SLA 3D printing: A pixel-based evolutionary algorithm with convolutional acceleration. TASE, 22, 20559--20571. (www)

  @article{ , abstract = {Additive manufacturing (AM) constructs complex products through the layer-by-layer deposition of materials. AM enables complex geometry fabrication but faces spatiotemporal optimization challenges: maximizing space utilization and minimizing printing time, requiring intelligent nesting of products and scheduling of resources. The integration of nesting and scheduling further expands the solution space, but existing methods frequently overlook essential geometric intricacies in irregular products. This paper proposes a novel pixel-based grey wolf optimizer algorithm (PGWO) to improve packing density and reduce time cost in stereolithography (SLA) printing. In the proposed PGWO, point-cloud pixelization is employed to simplify 3D irregular parts. Each part independently performs autonomous orientation to minimize local space occupancy at a low cost. Based on the time-frequency domain conversion, a convolutionally accelerated localization strategy (Cals) is proposed to improve the speed of nesting. For efficient scheduling, a bottleneck-balanced local search phase is designed with two operators. By targeting critical bottlenecks in printing, two operators can effectively reduce the frequency of layer changes, further optimizing local optimal solutions. PGWO is evaluated on 70 instances with diverse scales, and shows significant superiority over other state-of-the-art methods in over 88% of instances. The results demonstrate its superior performance in reducing printing time and enhancing packing density. Note to Practitioners—This work is motivated by the challenge of optimizing additive manufacturing (AM) for irregularly shaped parts, particularly in industries requiring high customization like aerospace and medical devices. While AM enables efficient production of complex geometries, existing methods struggle to simultaneously optimize spatial arrangement (nesting) and temporal scheduling, leading to inefficient material usage and prolonged printing time, which are critical barriers in automation science. Current research often simplifies irregular geometries or decouple nesting from scheduling, sacrificing either precision or efficiency. This paper proposes an evolutionary algorithm to achieve speed-fidelity balancing abstraction of 3D irregular parts and high-quality generation of scheduling solutions. The experimental results demonstrate superior performance in packing density and time reduction compared to conventional methods. For practitioners, this approach enables automated, resource-efficient additive manufacturing that reduce material waste and shorten production cycles, which are critical factors for scaling customized manufacturing. Current limitations include computational overhead for ultra-high-resolution parts, which could be mitigated through GPU parallelization in future implementations. In the future, this framework could be extended to advanced AM technologies such as multi-material printing or multi-machine collaboration, where efficient spatial-temporal optimization is essential.}, archive = {J_TASE}, author = {Zipeng Yang and Xinyu Li and Qihao Liu and Haoran Li and Liang Gao}, doi = {10.1109/TASE.2025.3603257}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20559--20571}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated nesting and scheduling for SLA 3D printing: A pixel-based evolutionary algorithm with convolutional acceleration}, volume = {22}, year = {2025}, }
• (2025). An efficient geometric points-guided generalized exponential splatting. TASE, 22, 20546--20558. (www)

  @article{ , abstract = {This paper presents a novel Generalized Exponential Splatting framework for scene reconstruction. Unlike neural radiance fields (NeRF), 3D Gaussian Splatting (3DGS) offers rapid rendering and high fidelity. However, existing 3DGS frameworks integrating RGB-D SLAM face challenges in system efficiency and aliasing artifacts. Inspired by geometric features, we first categorize the space into three regions—edge, planar, and non-planar—based on their information richness. The edge extraction is realized through an energy function, while planar and non-planar regions are segmented via a fast segmentation algorithm proposed in this paper. Secondly, a sparseness function is established to screen key points within the these regions to build a geometric points-guided Generalized Exponential Splatting model (GP-GES), strategically placing primitives to reduce redundancy. Finally, by integrating a 3D low-pass filter, a smooth GP-GES model is further built, ensuring that the frequency of the primitives adheres to the Nyquist-Shannon theorem, to mitigate high-frequency artifacts. Extensive experiments show that our method reduces the map size by $3.7\times $ to $6\times $ without sacrificing rendering quality, thereby accelerating the computational efficiency of the system. In addition, the segmentation algorithm has a fast detection speed, with a frame rate exceeding 90Hz. Note to Practitioners—3DGS has paved the way for practical applications in scene reconstruction, providing technical support across many fields. However, a review of current research and open-source projects reveals two main challenges: the 3DGS-based SLAM systems are commonly associated with inefficiency, and the reconstruction results are prone to artifacts. These drawbacks have hindered the broader adoption of the technology. In this paper, we propose another scene reconstruction framework. Through on-site deployment and experimental validation, our method has demonstrated superior performance in terms of cost-effectiveness and precision, achieving high-quality reconstruction results with fewer primitives. In future research, we will consider the effects of motion blur and depth measurement errors to develop a more robust reconstruction framework.}, archive = {J_TASE}, author = {Donglin Zhu and Zhongli Wang and Xin Jiang}, doi = {10.1109/TASE.2025.3602403}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20546--20558}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An efficient geometric points-guided generalized exponential splatting}, volume = {22}, year = {2025}, }
• (2025). Performance-enhanced intelligent fault-tolerant control for unknown nonlinear systems with multiple faults. TASE, 22, 20533--20545. (www)

  @article{ , abstract = {This article investigates the practical problem of the fault-tolerant tracking control for nonlinear systems subjected to dynamic uncertainties, unknown disturbances, and simultaneous actuator and sensor faults. A novel performance-enhanced intelligent fault-tolerant control (IFTC) framework is developed by integrating the predefined-time stability theory with the optimal control principle. A 1-dimensional convolutional neural network is proposed as a restorer to efficiently detect and isolate sensor faults, upon whose outputs a single radial basis function neural network is constructed to approximate two distinct components of the unknown system dynamics in real time. The proposed fault-tolerant controller ensures the boundedness of all signals within the closed-loop system and guarantees that the tracking errors converge within a user-specified time. Moreover, by incorporating a cost function related to the control effort, the developed controller avoids the excessive control input, achieving an enhanced control performance alongside the predefined-time convergence. Comprehensive simulation results validate the effectiveness and practicality of the proposed performance-enhanced IFTC scheme, highlighting its potential for real-world applications. Note to Practitioners—General input-affine nonlinear systems are capable of modeling a wide range of physical systems, such as motors, robotic arms, and aircraft. Due to the inherent complexity of these systems, which involve numerous parameters and are often subject to harsh operating conditions, multiple uncertainties like faults and disturbances frequently arise and may coexist. To address these challenges, this paper develops a performance-enhanced intelligent fault-tolerant controller for unknown input-affine nonlinear systems. The proposed IFTC framework utilizes offline system data, collected under both normal and faulty conditions, to detect and isolate sensor faults. Additionally, online adaptive control laws are designed to compensate for dynamic uncertainties, actuator faults, and disturbances, ensuring the system convergence within a user-specified time with reduced control effort. These improvements in the control performance have significant practical implications, demonstrating substantial potential for real-world engineering applications.}, archive = {J_TASE}, author = {Ran Chen and Donghua Zhou and Li Sheng}, doi = {10.1109/TASE.2025.3601727}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20533--20545}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Performance-enhanced intelligent fault-tolerant control for unknown nonlinear systems with multiple faults}, volume = {22}, year = {2025}, }
• (2025). Active fuzzy-based fault-tolerant control for uncertain electro-hydraulic actuators subject to simultaneous actuator and sensor faults. TASE, 22, 20519--20532. (www)

  @article{ , abstract = {This paper introduces an active fault-tolerant control (FTC) for servo-valve-driven electro-hydraulic actuators (EHAs) subject to unstructured dynamics, uncertainties, and actuator and sensor faults. The proposed methodology is established based on the backstepping control scheme integrated with observers and a fuzzy logic engine. The fuzzy logic-based approximation is utilized to compensate for the unstructured dynamics, while the observer is hybridized to suppress the effects of disturbance and lumped uncertainties. Moreover, under the adverse impact of the sensor malfunction, a disturbance-state observer (DSO) is developed to decouple the unknown residual signal. Notably, unlike other techniques using a bank-of-observers (BOb) to detect and isolate faults with a linear matrix inequality (LMI) tool to determine observer gains, the proposed observer views the residual as an extended state to identify it directly. Subsequently, the backstepping-based active FTC is developed to fulfill specific control requirements. The stability of the closed-loop system is theoretically achieved through the Lyapunov theorem, while the effectiveness of the proposed methodology is validated through comparative simulations under different faulty scenarios. Note to Practitioners—The main focus of this paper is to clarify the unsolved difficulties in addressing different faults, unknown uncertainties and unstructured dynamics in uncertain electro-hydraulic actuators simultaneously. The feasibility of the proposed methodology is verified through different scenarios of faults for intuitive evaluations. As the challenge in detecting, isolating, and estimating sensor faults for the FTC reconstruction, this paper brings a new methodology that can help instantly estimate and decouple residual signal from a measured one to ease the control development and is potentially expanded to other industrial applications.}, archive = {J_TASE}, author = {Hoai Vu Anh Truong and Wan Kyun Chung}, doi = {10.1109/TASE.2025.3599177}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20519--20532}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active fuzzy-based fault-tolerant control for uncertain electro-hydraulic actuators subject to simultaneous actuator and sensor faults}, volume = {22}, year = {2025}, }
• (2025). Adaptive fixed-time control for an uncertain robot with input quantization: A broad learning system approach. TASE, 22, 20505--20518. (www)

  @article{ , abstract = {In this paper, an adaptive fixed-time control approach is designed for a robot with dynamic uncertainty in the presence of input quantization by using the Broad learning system (BLS). The proposed BLS-based control algorithm is constructed by fusing the BLS with the radial basis function neural network, which is improved in terms of node selection rule with a self-adjusting Gaussian function center and enhancement layer. A hysteresis quantizer is applied to the requirement of a low transmission rate. For the nonlinearity occurring in the quantized input, a novel adaptive fixed-time method is developed such that 1) the adverse effect of quantization nonlinearity is removed in a finite interval; 2) the BLS-based approximation technique can improve the approximation accuracy, which enhances the robustness of the closed-loop system; and 3) via the Lyapunov stability method, the fixed-time convergence of the closed-loop system is proved. Finally, numerical simulations and experiments validate the effectiveness of the proposed control scheme. Note to Practitioners—This paper presents a novel approach to achieving fixed-time quantization for broad learning neural network-based controllers in robotic systems. The primary application of this research is in the field of automation, specifically for improving the performance of robotic joint tracking. The proposed method addresses the practical problem of ensuring that robotic systems can achieve precise tracking in fixed time, regardless of the initial conditions. This is particularly useful in industrial automation where consistent and reliable performance is crucial. Our approach leverages a fixed-time quantization technique to enhance the efficiency and reliability of neural network controllers. This method ensures that the tracking errors are confined to a small neighborhood of zero within a predetermined fixed time, thus significantly improving the productivity and reliability of the system. The results demonstrate that the proposed control scheme can handle various initial states and maintain robust performance.}, archive = {J_TASE}, author = {Donghao Zhang and Wenke Sun and Linghuan Kong and Xinbo Yu and Yifan Wu and Wei He}, doi = {10.1109/TASE.2025.3594597}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20505--20518}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fixed-time control for an uncertain robot with input quantization: A broad learning system approach}, volume = {22}, year = {2025}, }
• (2025). A structure-modification-based petri net modeling and reachability analysis method for automated manufacturing systems. TASE, 22, 20493--20504. (www)

  @article{ , abstract = {Due to their graphical representation and capability for property analysis, Petri nets (PNs) have been widely used in developing automated manufacturing systems (AMS). When doing so, it is important to perform their state reachability analysis and verify their functionality. The most common approach is to traverse a reachability tree of their PN models. However, it has the problem of state space explosion. A way to determine a state’s reachability is to find a firing sequence (FS) that corresponds to nonnegative integer solutions (NISs) of a state equation. Our prior work has given an algorithm to decide the existence of FS corresponding to an NIS in polynomial time. Yet it is impossible to decide the reachability of a marking given an infinite number of NISs of a state equation. This work studies the relationship between the PN model and NIS count of any state equation. An innovative method is proposed to modify a PN structure such that, for any initial state and destination one, its state equation has no more than one NIS. Given a PN model of an AMS, by analyzing the relationship between its properties and the modified one’s, we have that the proposed method can maintain the functionality of the modeled AMS. As a result, by using the modified PN as the AMS’s final model, the reachability of any marking can be determined in polynomial time, which can be viewed as a breakthrough result in the field of PN analysis. The proposed method is illustrated via case studies. Note to Practitioners—Petri nets (PNs) are established formal tools with extensive application in the modeling and analysis of automated manufacturing systems (AMS). Reachability is to determine whether an AMS can reach a state from an initial one, which is crucial for ensuring system robustness. However, for PN-based AMS, existing methods have difficulty in determining the reachability of an arbitrary state due to excessive computational complexity. This research introduces a novel reachability-decidable PN modeling method, where the structure of the PN is modified while preserving the functionality of the modeled system. Consequently, the reachability of any marking can be determined in polynomial time by employing a state-equation-based method. The proposed approach avoids exhaustive state enumeration, resulting in significant time and computational resource savings. It is important in promoting PN’s industrial applications.}, archive = {J_TASE}, author = {Yue Su and MengChu Zhou and Liang Qi and Aiiad Albeshri and Abdullah Abusorrah}, doi = {10.1109/TASE.2025.3588429}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20493--20504}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A structure-modification-based petri net modeling and reachability analysis method for automated manufacturing systems}, volume = {22}, year = {2025}, }
• (2025). Robust adaptive optimal fault-tolerant control scheme for multi-machine power systems via the stabilizing two-stage policy iteration framework. TASE, 22, 20481--20492. (www)

  @article{ , abstract = {For a class of multi-machine power systems (MMPSs) with unknown dynamics and actuator faults, two model-free optimal fault-tolerant (FT) control frameworks based on policy iteration (PI) are proposed. First, the actuator failure is modeled within the interconnected MMPSs. Building on this foundation, a stable FT-PI framework is developed that eliminates the conventional dependence on initial admissible controllers through a proposed two-stage implementation. Additionally, an enhanced Q-learning equivalent framework is introduced, which further relaxes the conditions on the initial stabilizing policy typically required by traditional integral Q-learning methods. Both of these data-driven algorithms dynamically solve the FT continuous algebraic Riccati equation (FT-CARE) without requiring knowledge of the system dynamics and actuator fault information. Meanwhile, the convergence and stability of the algorithms are rigorously proved. Finally, the effectiveness of the proposed algorithms is verified through a three-machine power system. Note to Practitioners—MMPSs are crucial components of modern power grids. However, actuator failures and uncertain dynamics pose threats to system stability. Conventional FT control, when aiming for optimal performance, often demands accurate system modeling and fault information. This becomes impractical or costly when dealing with MMPSs having unknown dynamics, as achieving optimal control under such uncertainties is challenging. This work focuses on addressing these practical challenges. We propose PI-based optimal FT control frameworks for adaptive optimal FT control in MMPSs. The two proposed frameworks simplify the implementation of optimal control—mitigating reliance on practitioners’ prior experience, especially in choosing the initial stabilization gain. From a practical application perspective, these frameworks offer engineering-oriented solutions. They provide a strong theoretical foundation to ensure system stability and feasibility during operation, enabling engineers to implement optimal FT control in MMPSs more easily in real-world scenarios, even without complete knowledge of system dynamics and fault details.}, archive = {J_TASE}, author = {Hai Wang and Junmin Li and Chao He and Chaochen Song}, doi = {10.1109/TASE.2025.3603734}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20481--20492}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust adaptive optimal fault-tolerant control scheme for multi-machine power systems via the stabilizing two-stage policy iteration framework}, volume = {22}, year = {2025}, }
• (2025). Actuator fault estimation for a class of euler-lagrange systems using super-twisting observers. TASE, 22, 20469--20480. (www)

  @article{ , abstract = {This paper develops an adaptive actuator fault estimation scheme for a class of Euler-Lagrange systems with relative degree two, which allows the loss of actuator effectiveness and generalised velocities to be estimated. In particular, a super-twisting observer is created for Euler-Lagrange systems with multiple degrees of freedom and the ‘equivalent output error injection’ signals, associated with the super-twisting sliding motion, are exploited to reconstruct the loss of effectiveness for all the actuators in an adaptive sense. The proposed scheme is then applied to a six-link underactuated snake robot model and the simulation results demonstrate the efficacy of the scheme. Note to Practitioners—This paper is motivated by the problem of actuator condition monitoring for snake robots. The presented results also apply to a wide range of underactuated systems formulated in a Euler-Lagrange manner, especially those that satisfy virtual holonomic constraints. Since snake-like robots are often underactuated, their control performance is more likely to degrade in the presence of actuator faults. In both the academic community and in engineering practice, very few examples have attempted to address actuator fault estimation problems for snake-like robots due to the inherent structural limitations. Furthermore, the mathematical model of the snake robot in this paper is well-known and has been successfully employed for the development of advanced control laws applied to real-world snake robots. Consequently, the work presented in this paper is unique and possesses a degree of engineering value.}, archive = {J_TASE}, author = {Lejun Chen and Sarah Spurgeon and Bin Jiang}, doi = {10.1109/TASE.2025.3604035}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20469--20480}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Actuator fault estimation for a class of euler-lagrange systems using super-twisting observers}, volume = {22}, year = {2025}, }
• (2025). Learning-based tube MPC for multi-area interconnected power systems with wind power and HESS: A set identification strategy. TASE, 22, 20458--20468. (www)

  @article{ , abstract = {With the development of intelligent automation technology and advancement of modernization, the degree of interconnection between power systems is increasing. With the main purpose of involving hybrid energy storage systems (HESS) in optimizing system frequency, this work proposes a learning-based tube model predictive control (MPC) for the multi-area interconnected power systems with wind power and HESS. The suggested method has strong adaptability due to the introduction of a new robust constraint handled by a learning mechanism. By identifying the uncertainty set of coupling strength of online data in the learning stage, the optimal MPC problem is calculated in the adaptive stage, which effectively reduces the adverse effects of disturbances and noises in multi-area interconnected power systems. Moreover, an input to state stability criterion is provided to ensure the robust stability of the system with uncertain disturbances and noises. With simulations on a four-area interconnected power system with wind power and HESS, the effectiveness of proposed method is discussed on an improved IEEE 39-bus system. Note to Practitioners—To achieve a balance between load demand and power generation in multi-area interconnected power systems, various load frequency controls have been widely designed. The disturbances caused by high interconnectivity and the noise generated by electrical equipment pose a threat to the reliable operation of the power system. There is an urgent need to develop appropriate strategies to resist these interferences. So far, HESS have been widely applied in power systems with new energy to improve key system responses, such as power system frequency. This prompts defenders to design different types of external controllers to optimize the frequency deviation of the power system based on HESS. To tackle disturbances and noises which impose serious threat to system stability, we introduce a learning mechanism on the property of invariant sets and propose a learning-based tube MPC strategy that identifies disturbance invariant sets during the learning stage and calculates the optimal output during the adaptation stage. Moreover, simulations are presented to demonstrate that the proposed tube MPC strategy can provide satisfactory stability performance for interconnected power systems under disturbances and noises.}, archive = {J_TASE}, author = {Zhuoer An and Xinghua Liu and Gaoxi Xiao and Meng Zhang and Zhongmei Pan and Yu Kang and Nicholas Jenkins}, doi = {10.1109/TASE.2025.3603607}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20458--20468}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based tube MPC for multi-area interconnected power systems with wind power and HESS: A set identification strategy}, volume = {22}, year = {2025}, }
• (2025). A differential friction driving piezoelectric stepping robot for 3-DOF planar micromanipulation. TASE, 22, 20448--20457. (www)

  @article{ , abstract = {Miniature piezoelectric robots exhibit significant potential in high-precision actuation applications; however, their inherent structural complexity often limits their widespread adoption in the miniature robotics field. To address this limitation and develop a miniature robot characterized by high precision, 3-DOF, structural simplicity, and low driving requirements, this paper proposes a piezoelectric stepping robot based on a novel differential friction principle. This differential friction piezoelectric stepping robot effectively leverages the friction differential principle to achieve high-precision actuation while maintaining structural simplicity. Its 3-leg driving configuration provides excellent motion stability and enables 3-DOF planar motion (translational and rotational). Experimental results demonstrate that the robot (dimensions: 44.5 mm x 50.3 mm x 16mm) achieved a translational resolution of 6.97 nm and a maximum speed of 339.35 $\mu $ m/s, along with a rotational resolution of 2.82 $\mu$ rad and a maximum speed of 4.31 mrad/s. The robot also demonstrated an excellent maximum payload-to-weight ratio of 64.52. Furthermore, the robot successfully performed precision alignment of a 200 $\mu $ m diameter optical fiber. Its advantages such as simple structure and few driving units enhance its potential for practical applications and provide a promising approach for developing high-performance micro-robotic systems. Note to Practitioners—This paper is driven by the practical need for a miniature robotic positioner that is simple in design, easy to control, yet capable of delivering high-precision, multi-degree-of-freedom motion with significant load-carrying capacity. Such robots are essential for applications ranging from semiconductor probing and optical device alignment to in-situ sample manipulation inside scientific instruments. We present a novel 3-DOF planar stepping robot that achieves these goals using a unique “differential friction” principle, requiring only three simple piezoelectric actuators. This design significantly reduces the structural complexity and the number of driving channels compared to traditional multi-DOF micro-robots. The experimental results validate its practical utility: the robot demonstrates nanometer-level positioning resolution (6.97 nm linear, 2.82 $\mu $ rad rotational) and a remarkable payload-to-weight ratio of over 64. We have successfully demonstrated its capability in a challenging fiber optic alignment task, showcasing its ability to perform fine translational and rotational adjustments. For practitioners looking to build or integrate a compact and robust micro-positioning system, this work offers a complete blueprint—from the operating principle and design model to demonstrated performance. The inherent simplicity of the robot makes it a promising platform for developing next-generation automated micromanipulation systems for both research laboratories and industrial settings.}, archive = {J_TASE}, author = {Chutian Dai and Shupeng Wang and Xinbao Wang and Wencheng Wu}, doi = {10.1109/TASE.2025.3602855}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20448--20457}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A differential friction driving piezoelectric stepping robot for 3-DOF planar micromanipulation}, volume = {22}, year = {2025}, }
• (2025). Visual localization using 3D gaussian splatting representation for mobile robots with geometric feature correspondences synthesis. TASE, 22, 20435--20447. (www)

  @article{ , abstract = {Achieving visual localization with excellent interactive performance is challenging for mobile robots. Based on real-time photo-realistic view synthesis, 3D Gaussian splatting (3DGS) representation has demonstrated vast potential for robots engaging with the physical world. In this work, we propose a novel coarse-to-fine visual localization method named L3DGS based on the 3DGS radiance field representation. Particularly, during the coarse stage, we exploit novel views synthesized by the pretrained 3DGS map to create geometric feature correspondences to perform geometric alignment. Then, we integrate both geometric and photometric alignment to refine the camera pose. Unlike previous radiance field-based approaches, we leverage geometric feature correspondences and the innovative 3DGS map to improve the localization accuracy. In our experiments, we evaluate the proposed method across two real-world indoor and outdoor datasets. Consequently, compared to the baselines, the proposed method achieves competitive or superior experimental results. Note to Practitioners—This paper is motivated by the problem of map-based visual localization for autonomous mobile robots but is also applied to other robotic systems that represent a 3D scene using 3D Gaussian splatting (3DGS) representation. In practical applications, the accuracy of existing radiance field-based localization methods is susceptible to image retrieval. In this paper, we propose a visual localization method applied to both indoor and outdoor scenarios for mobile robots. Our technical innovation lies in using 3DGS as the map and utilizing its rendered images to enhance the robustness of the visual localization algorithm. Compared with structure-based methods, the proposed method requires less time for 3D map reconstruction, while enhancing both the accuracy and robustness of visual localization. We conduct extensive experimental studies in real-world indoor and outdoor scenarios to validate the effectiveness of the proposed method. In the future, we will focus on improving the efficiency of the proposed method and carry out applications in visual navigation and path planning of mobile robots.}, archive = {J_TASE}, author = {Zhiyu Zhou and Feng Hui and Xing Li and Yu Liu}, doi = {10.1109/TASE.2025.3601848}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20435--20447}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Visual localization using 3D gaussian splatting representation for mobile robots with geometric feature correspondences synthesis}, volume = {22}, year = {2025}, }
• (2025). RINO: Accurate, robust radar-inertial odometry with non-iterative estimation. TASE, 22, 20420--20434. (www)

  @article{ , abstract = {Odometry in adverse weather conditions, such as fog, rain, and snow, presents significant challenges, as traditional vision- and LiDAR-based methods often suffer from degraded performance. Radar-Inertial Odometry (RIO) has emerged as a promising solution due to its resilience in such environments. In this paper, we present RINO, a non-iterative RIO framework implemented in an adaptively loosely coupled manner. Building upon ORORA as the baseline for radar odometry, RINO introduces several key advancements, including improvements in keypoint extraction, motion distortion compensation, and pose estimation via an adaptive voting mechanism. This voting strategy facilitates efficient polynomial-time optimization while simultaneously quantifying the uncertainty in the radar module’s pose estimation. The estimated uncertainty is subsequently integrated into the maximum a posteriori (MAP) estimation within a Kalman filter framework. Unlike prior loosely coupled odometry systems, RINO not only retains the global and robust registration capabilities of the radar component but also dynamically accounts for the real-time operational state of each sensor during fusion. Experimental results conducted on publicly available datasets demonstrate that RINO reduces translation and rotation errors by 1.06% and 0.09°/100m, respectively, when compared to the baseline method, thus significantly enhancing its accuracy. Furthermore, RINO achieves performance comparable to state-of-the-art methods. Our code is available at https://github.com/yangsc4063/rino Note to Practitioners—Odometry for autonomous vehicles faces significant challenges under adverse conditions such as fog, rain, and snow. Commonly used sensors, including cameras and LiDAR, perform poorly in such environments, leading to failures in traditional odometry methods. Scanning millimeter-wave radar, with its strong penetration capabilities, offers a promising alternative by providing dense, wide-ranging, and robust perception under these conditions. However, radar data is often noisy and prone to ghosting artifacts, making processing challenging and hindering its widespread research and practical application. In this paper, we propose a Radar-Inertial Odometry (RINO) system that integrates pose estimation from scanning radar and IMU while incorporating uncertainty estimation within a non-iterative point cloud registration framework. Experimental validation on both datasets and real-world vehicle tests shows that RINO outperforms existing methods in terms of accuracy, efficiency, and robustness, providing a reliable solution for autonomous navigation in adverse weather conditions.}, archive = {J_TASE}, author = {Shuocheng Yang and Yueming Cao and Shengbo Eben Li and Jianqiang Wang and Shaobing Xu}, doi = {10.1109/TASE.2025.3601838}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20420--20434}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RINO: Accurate, robust radar-inertial odometry with non-iterative estimation}, volume = {22}, year = {2025}, }
• (2025). Network-side carbon emission reduction via dispatching power electronic devices in AC–DC hybrid distribution systems. TASE, 22, 20406--20419. (www)

  @article{ , abstract = {Heightened environmental concerns have arisen the urge for low-carbon power systems. Existing literatures on low-carbon dispatching have fully studied how to reduce carbon emissions on generation side or demand side. But, carbon emissions related to network power losses (i.e., network-side carbon emissions), especially in distribution system (DS), significantly affects the whole carbon distributions as well. However, the objective of conventional low-carbon economic dispatching fails to precisely reduce this part of carbon emissions. To address this issue, we propose a network-side carbon emission reduction method. After determining active power flows by the day-ahead economic dispatching, this method innovatively explores utilizing reactive power compensation capability of some emerging power electronic devices in AC-DC hybrid DS for reducing network-side carbon emissions. First, the traditional Var compensation devices and the emerging voltage source converters (VSCs) with specific control strategies are modelled. Second, models of inverter-based PV units and the aggregated feasible power region (AFPR) of PV aggregators are also proposed. Finally, the low-carbon dispatching model for network-side carbon emission reduction is established, and an ALSO-X+ method is used to solve joint chance constraints (JCC) on the AFPR caused by the uncertainties of PV generation. Case studies on AC-DC test systems with different scales verify that dispatching reactive power can also influences carbon intensities among the whole distribution network. Moreover, our proposed method could effectively reduce network-side carbon emissions and demand-side carbon emissions. Note to Practitioners—Optimizing carbon emissions on network side via dispatching reactive power significantly affect carbon distributions as well, which shows great potential on achieving the low-carbon goal in DS. And in practice, it is essential to unlock the reactive power compensation capability of power electronic devices, especially in AC-DC hybrid DS. This paper proposes a network carbon emission reduction method considering dispatching various reactive power devices. This method takes various control modes of VSC into account, and utilizes JCC to deal with uncertainties during the PV aggregation process. Moreover, the novel ALSO-X+ algorithm is adopted to efficiently decompose the JCC constraints. The proposed method not only provides new sights and extends current researches for achieving the low-carbon goal in DS, but also demonstrates the adaptability for AC-DC hybrid DS operation and exhibits significant scalability, which offers the considerable practical values for industry practitioners.}, archive = {J_TASE}, author = {Qianhao Sun and Yao Zhang and Hanting Zhao and Wei Huo and Jian Liao and Jianxue Wang}, doi = {10.1109/TASE.2025.3601738}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20406--20419}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Network-side carbon emission reduction via dispatching power electronic devices in AC–DC hybrid distribution systems}, volume = {22}, year = {2025}, }
• (2025). Collisions-free and connectivity-preserving distributed cooperative output regulation of nonlinear multiagent systems. TASE, 22, 20394--20405. (www)

  @article{ , abstract = {The existing distributed cooperative output regulation schemes have designed safe command signals for collisions/obstacle avoidance. However, in large-scale multiagent system (MAS) formations, due to transient tracking errors or formation changes for collision/obstacle avoidance, the distance between an agent and its neighboring agents exceeds their communication range, leading to disconnection that changes the command signal, ultimately disrupting cooperation. To address this, we design a novel nonlinear command signal, featuring two artificial-delayed trajectories and a delay-free observer-governor state, whose coefficients are adaptively adjusted by three dynamic barrier functions to avoid collisions and preserve connectivity. It is demonstrated that the MAS can adaptively reconfigure its formation in a distributed manner to avoid collisions and preserve connectivity when entering the region covered by the barrier functions, and rapidly recover the desired formation along with the command signal after moving away from the coverage range of obstacle-avoidance barrier function. Finally, the results obtained from comparative simulations confirm the efficacy of the proposed method. Note to Practitioners—This work addresses a critical problem in automated systems such as drone swarms or factory robots: sudden formation changes for collision/obstacle avoidance or transient tracking errors often cause agents to lose communication with teammates, leading to mission failures. For example, in a robotic assembly line, a robot dodging a fallen object might disconnect from others, halting production. However, existing results only design safe command signals for collision/obstacle avoidance but neglect connectivity preservation during dynamic maneuvers. This paper proposes a novel command signal that integrates two artificial-delayed trajectories and a delay-free observer-governor state to enable agents to autonomously reconfigure formations around obstacles while avoiding collision/obstacle avoidance and maintaining connectivity. This approach requires only low-cost sensors (e.g., IoT devices) and integrates with existing industrial systems, avoiding costly infrastructure upgrades. However, it currently assumes undirected communication networks and requires further validation in real-world production settings. Future work will investigate novel collision/obstacle avoidance strategies under unreliable networks and extend the framework to applications such as warehouse logistics (robots navigating narrow aisles) or disaster rescue (drones coordinating in collapsed buildings).}, archive = {J_TASE}, author = {Longbin Fu and Liwei An}, doi = {10.1109/TASE.2025.3601609}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20394--20405}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Collisions-free and connectivity-preserving distributed cooperative output regulation of nonlinear multiagent systems}, volume = {22}, year = {2025}, }
• (2025). A gravity compensation system for touchdown process verification of large spacecraft with quasi-zero stiffness mechanism and passivity-based controller. TASE, 22, 20380--20393. (www)

  @article{ , abstract = {To ensure the reliability of an extraterrestrial lander during its touchdown process, it is crucial to replicate the landing dynamics within a simulated low-gravity environment. In this study, a servo-driven system(SDS) is proposed and integrated with a quasi-zero stiffness(QZS) system to address the limitations of traditional methods, including restricted accuracy and limited movement range. An analytical method for describing the dynamic characteristics of the low-gravity simulation system is presented based on passivity theory, enhancing the compensation accuracy. Subsequently, a fuzzy controller for the QZS-SDS composite system is designed to autonomously adjust the equilibrium point and convergence speed, ensuring rapid error reduction in tension control. Furthermore, the global asymptotic stability conditions of the closed-loop system are established and analyzed, and the quantitative parameter design guidelines for practical implementation are provided. Finally, a series of physical experiments demonstrate the efficiency and robustness of the proposed system. Compared to conventional methods using only QZS, the proposed approach reduces the maximum force compensation error from 13.9% to 4.0%, and extends the supported simulation range from 800mm to over 12,000mm. Note to Practitioners—This study addresses the challenges of providing a reliable low-gravity environment for the touchdown phase of extraterrestrial landers. The primary motivation is to verify the robustness of the engine shutdown algorithm and evaluate the strength/stiffness design of the landing legs under realistic impact conditions. While some low-gravity compensation systems have been widely used in orbiting and descent phases, their effectiveness during the touchdown process, where strong impacts occur, remains insufficiently explored. At higher landing velocities(approximately 1.1 m/s), most low-gravity compensation systems struggle to maintain accuracy due to bandwidth limitations, except for quasi-zero stiffness(QZS) systems. However, QZS-based approaches suffer from mechanical deviations such as stiffness inconsistencies in energy storage components and friction, leading to reduced compensation accuracy. Additionally, the movement range of QZS is constrained by mechanical structures, restricting their application in spacecraft landing missions. To overcome these limitations, we propose an actively controlled servo-drive system with a passivity-based control law for precise force compensation. The introduction of servo motors not only improves compensation accuray but also significantly extends the system’s operational range, enabling it to meet the requirements of various landing tasks. This work has direct implications for practitioners developing ground-based low-gravity simulation systems for extraterrestrial exploration. In the future, we aim to extend our low-gravity compensation technology to support extraterrestrial launch missions, broadening its application from single-degree-of-freedom landing tasks to multi-degree-of-freedom non-vertical take-off scenarios.}, archive = {J_TASE}, author = {Rongqiao Zhang and Zhengqing Li and Haining Sun and Xiaoqaing Tang}, doi = {10.1109/TASE.2025.3600626}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20380--20393}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A gravity compensation system for touchdown process verification of large spacecraft with quasi-zero stiffness mechanism and passivity-based controller}, volume = {22}, year = {2025}, }
• (2025). Content caching for IoT devices by using self-feedback adversarial semi-bandits learning. TASE, 22, 20366--20379. (www)

  @article{ , abstract = {As massive data is generated by Internet of Things (IoT) devices, user-end devices are required to implement computation-intensive functionalities, including multi-sensory data processing and analysis, sophisticated system control schemes, and artificial intelligence. Mobile Edge Computing (MEC) is a significant technology that has the potential to extend the computation and storage capacities of user-end devices by the decentralization of required resources and contents near users and at the edge. A crucial challenge in this direction is the development of a smart mechanism to effectively cache contents upon Edge Servers (ESs) near users for high effectiveness and low latency of content delivery with the constraints on computational and storage capacities of ESs. This study employs a queuing model for analyzing total request delay and interprets the content caching problem as an adversarial semi-bandits problem. We propose an Online Self-feedback Adversarial semi-bandits Learning (OSAL) algorithm that incorporates a dual-layer learning architecture for dynamically generating caching strategies and maximizes the long-term reward. Experimental results demonstrate that the proposed method significantly outperforms the state-of-the-art methods across various performance metrics in a real-world multi mobile-user content caching case. Note to Practitioners—This work addresses the important trade-off issue between delays and cache storage cost for an edge-cloud computing system that operates with limited cache storage budgets but must meet real-time requirements. Our proposed Online Self-feedback Adversarial semi-bandits Learning (OSAL) algorithm can be used to improve the operational efficiency of the systems with edge computing setting in the area of Internet of Things (IoT). These systems employ edge servers to cache the frequently reusable contents with the aim to minimize accumulative delays of all tasks while maintaining the energy cost at the acceptable level. Theoretical analysis and simulation results jointly show that the proposed solution outperforms several recently proposed ones, thus making it readily applicable to industrial scenarios.}, archive = {J_TASE}, author = {Jinpeng Li and Peng Chen and Yunni Xia and MengChu Zhou and Yong Ma and Yumin Dong and Hui Liu and Qinglan Peng and Xifeng Xu}, doi = {10.1109/TASE.2025.3598689}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20366--20379}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Content caching for IoT devices by using self-feedback adversarial semi-bandits learning}, volume = {22}, year = {2025}, }
• (2025). OM-koop: Online memorable koopman operator learning for marine robots steering dynamics. TASE, 22, 20354--20365. (www)

  @article{ , abstract = {The steering dynamics of marine robots play a pivotal role in achieving precise maneuvering. However, complex and unpredictable ocean disturbances pose challenges for rapid online learning of accurate dynamics. This paper presents the Online Memory Koopman Learning (OM-Koop) framework, a hybrid model that combines physical priors, online data and stability preserving mechanisms to solve the nonlinear dynamic capture challenge and dynamically adapt to the marine environment. Firstly, we construct the Koopman operator-based uncertainty model online using the state error of the steering model and sliding window methods. The model can effectively capture the nonlinear features that are not represented in the predefined steering model. To ensure stability, the eigenvalues of the Koopman operator are constrained during online learning, guaranteeing Lyapunov stability. Secondly, in order to improve the efficiency of online learning, the Long Short-Term Memory (LSTM) neural network is involved in the construction process of the Koopman operator, which enhances the model’s memory capability. Finally, through field experiments using Autonomous Surface Vehicles (ASVs) and Autonomous Underwater Vehicles (AUVs) in field environments, comparative analyses with other learning strategies show that OM-Koop has excellent adaptability and robustness while guaranteeing Lyapunov stability. Note to Practitioners—The motivation of this article is that the steering dynamic behavior of marine robots is highly affected by unpredictable ocean environments, which poses great challenges in achieving precise manipulation in practical applications. In this paper, we propose the OM-Koop framework to address these challenges by integrating physical priors, online data learning, and stability preserving mechanisms. Theoretical analyses show that the proposed framework ensures Lyapunov stability while dynamically adapting to nonlinear disturbing forces imposed by the environment. Field experiments on AUV and ASV validate the robustness and adaptability of the framework, demonstrating its potential for practical deployment in dynamic marine environments. And the proposed framework has prospects for practical applications in other robotic systems.}, archive = {J_TASE}, author = {Hongde Qin and Siju Yuan and Yifan Xue and Hongkun He and Zehao Wu and Jing Zhao and Qingsong Xu}, doi = {10.1109/TASE.2025.3597887}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20354--20365}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {OM-koop: Online memorable koopman operator learning for marine robots steering dynamics}, volume = {22}, year = {2025}, }
• (2025). Finite-time prescribed performance h∞ control of multi-agent systems based on sample-data event-triggered mechanism. TASE, 22, 20341--20353. (www)

  @article{ , abstract = {This study considers the finite-time prescribed performance $H_{\infty } $ tracking consensus control problem in nonlinear multi-agent systems with time-varying disturbance and designs a distributed event-triggered control strategy. First, a class of time-dependent error performance constraint functions is introduced to effectively address the negative impact of disturbances on the transient and steady-state performance of a system. Second, a dynamic event-triggered protocol defined using periodic sampling data is proposed to reduce the information transmission between agents and decrease the controller’s update frequency. Third, the designed dynamic event-triggered mechanism is combined with the finite-time $H_{\infty } $ control theory, and sufficient conditions for the system to achieve tracking consensus in a finite time are obtained by employing the Lyapunov stability theory and linear matrix inequality. Finally, the feasibility of the proposed distributed control strategy is verified through numerical simulations and practical examples. Note to Practitioners—This paper aims to achieve the finite-time prescribed performance consensus for nonlinear multi-agent systems, which is of great practical significance in multi-mechanical arm cooperative control, UAV formation control, etc. A distributed dynamic event-triggered control protocol is proposed to address the challenge of time-varying disturbances within nonlinear multi-agent systems. Utilizing periodic sampling data, this protocol can potentially resolve two particularly salient problems in practice: the shortage of control resources and the limited storage space that characterizes many systems. Furthermore, most extant sample-data event-triggered consensus controls exclusively consider the system’s steady-state performance or asymptotic consensus control, which also restricts their application in practical engineering. Therefore, this paper combines the prescribed performance control with sample-data event-triggered mechanism and finite-time control theory to ensure the convergence speed of the system while taking into account the transient and steady-state performance of the system.}, archive = {J_TASE}, author = {Yuejie Yao and Yiping Luo and Jinde Cao and Anping Li and Dong Yue}, doi = {10.1109/TASE.2025.3603105}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20341--20353}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time prescribed performance h∞ control of multi-agent systems based on sample-data event-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Motion coordination of swarm robots for mobile target search. TASE, 22, 20323--20340. (www)

  @article{ , abstract = {With the rapid advancement of robotics technologies, a group of robots are able to communicate with one another by wireless transmissions and form a robot swarm. Robot swarm has many applications and a typical one is target search in which swarm robots are sent to places that might be dangerous for human workers, and they coordinate with one another to search for targets such as survivors in a disaster. However, motion coordination of swarm robots for target search has received little attention especially when the targets are mobile. In this work, we develop a motion coordination algorithm for swarm robots to search for targets in an unknown area. Our basic idea is to divide the search area into grids and build a gray-scale map in which each grid is associated with a gray scale indicating the efficiency of searching targets in this grid. The Voronoi diagram is adopted to coordinate swarm robots to search different portions of the search area for maximizing search efficiency. By theoretical analysis, our motion coordination algorithm is validated to ensure that all static targets are guaranteed to be found. We derive an upper-bound on the total time for robots to traverse all the grids in the search area. Extensive simulations are conducted and the results show that the proposed motion coordination algorithm outperforms the state-of-the-art and achieves a success rate of over 90% in finding all mobile targets with low search latency. Note to Practitioners—This paper was motivated by the problem of target search in an unknown area (e.g., the search for Malaysia Airlines MH370). With advanced robotics technologies, swarm robots could be sent to the area to perform a search mission. This work suggests a motion coordination algorithm for swarm robots to search for both static and mobile targets in an unknown area with possible obstacles. The proposed algorithm divides the search area into grids and each grid is associated with a gray scale indicating the efficiency of searching targets in this grid. This gray-scale map guides the robots in target search. In real applications, robots are normally powered by batteries and are prone to fail, requiring efficient coordination of the robots to maximize search efficiency. A Voronoi mask is further introduced to coordinate swarm robots to search different portions of the area. This mask divides the space based on the robots’ current positions, assigning each robot a specific region that is closer to it than to others, leading to efficient and coordinated operation. We mathematically derive an upper-bound on the total time for robots to traverse all the grids in the search area. We then show by simulations that the proposed motion coordination algorithm outperforms the state-of-the-art and coordination of robots is critical to improving success rates and minimizing search latency. The proposed algorithm has not yet been tested in a robot testbed which is our future work.}, archive = {J_TASE}, author = {Hai Liu and Shujin Ye and Chris Y. T. Ma and Yue Wang and Tse-Tin Chan}, doi = {10.1109/TASE.2025.3600954}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20323--20340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Motion coordination of swarm robots for mobile target search}, volume = {22}, year = {2025}, }
• (2025). SGCF: An MPC-inspired control framework using sequence generation networks for flotation process. TASE, 22, 20311--20322. (www)

  @article{ , abstract = {In order to integrate the advantages of Model Predictive Control (MPC) and artificial neural networks, this paper proposes a control framework based on sequence generation (SGCF). The main idea of SGCF is to first generate an ideal subsequent state trajectory conditioned on the current working condition, and then infer feasible control sequences accordingly. To achieve implicit optimization of the performance metric during sequence generation, different learning weights are assigned to labeled samples. A confidence score based on model interaction is further introduced to evaluate the reliability of each inference and assist training. Theoretical analysis is provided to explain how the sequence generation network learns under performance guidance, promoting better trajectories. The effectiveness of SGCF is validated through experiments on inverted pendulum control and lead-zinc rougher flotation reagent control. Multiple baseline methods are considered, including state feedback control (SFC) models and data-driven MPC approaches. Results demonstrate that SGCF achieves advantages in both control performance and cost, while maintaining favorable inference efficiency. Moreover, SGCF exhibits potential for interpretable reasoning. Note to Practitioners–Model predictive control (MPC) is widely used for balancing control performance and cost, but its application to flotation processes faces challenges due to system modeling difficulties and online optimization complexity. On the other hand, conventional neural network control may suffer from overfitting and poor interpretability, making its deployment in industrial settings risky. This paper proposes a control framework that combines MPC principles with sequence generation networks. The approach first generates target adjustment strategies based on current flotation conditions—such as desired froth size or brightness—and then infers the required reagent flow rates to achieve these targets. Beyond providing recommended control actions, the framework also offers users additional feedback on predicted process changes and the confidence of each decision, improving transparency and user trust. Experimental results on both inverted pendulum and flotation reagent control tasks demonstrate that this method can improve both product quality and reagent usage efficiency, while keeping computational requirements practical for real-time deployment. Further research is still needed to explore which types of sequence generation models are most suitable for various scenarios, as well as to develop more user-friendly explanations based on predicted system behavior.}, archive = {J_TASE}, author = {Zhiqiang Qian and Yongfang Xie and Shiwen Xie and Zhaohui Tang}, doi = {10.1109/TASE.2025.3599660}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20311--20322}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SGCF: An MPC-inspired control framework using sequence generation networks for flotation process}, volume = {22}, year = {2025}, }
• (2025). Force-compliance MPC and robot-user CBFs for interactive navigation and user-robot safety in hexapod guide robots. TASE, 22, 20296--20310. (www)

  @article{ , abstract = {Guiding the visually impaired in complex environments requires real-time two-way interaction and safety assurance. We propose a Force-Compliance Model Predictive Control (FC-MPC) and Robot-User Control Barrier Functions (CBFs) for force-compliant navigation and obstacle avoidance in Hexapod guide robots. FC-MPC enables two-way interaction by estimating user-applied forces and moments using the robot’s dynamic model and the recursive least squares (RLS) method, and then adjusting the robot’s movements accordingly, while Robot-User CBFs ensure the safety of both the user and the robot by handling static and dynamic obstacles, and employ weighted slack variables to overcome feasibility issues in complex dynamic environments. We also adopt an Eight-Way Connected DBSCAN method for obstacle clustering, reducing computational complexity from O( $n^{2}$ ) to approximately O(n), enabling real-time local perception on resource-limited on-board robot computers. Obstacles are modeled using Minimum Bounding Ellipses (MBEs), and their trajectories are predicted through Kalman filtering. Implemented on the HexGuide robot, the system seamlessly integrates force compliance, autonomous navigation, and obstacle avoidance. Experimental results demonstrate the system’s ability to adapt to user force commands while guaranteeing user and robot safety simultaneously during navigation in complex environments. Note to Practitioners—Guiding visually impaired individuals in complex environments remains a significant challenge, as traditional methods like guide dogs are limited by high costs and scalability issues. To address this, we propose a Hexapod guide robot framework that integrates Force-Compliance Model Predictive Control (FC-MPC) and Robot-User Control Barrier Functions (CBFs) to ensure real-time two-way interaction and dual safety for both the user and the robot. The framework supports diverse interaction modes by allowing parameter adjustments to combine functionalities such as force compliance, global autonomous navigation, and obstacle avoidance. An efficient clustering algorithm is adopted in the local environment perception process, which enables real-time deployment on resource-constrained devices of guide robot. While effective in experiments, future work will focus on map-less navigation to improve performance in previous unseen environments.}, archive = {J_TASE}, author = {Zehua Fan and Feng Gao and Zhijun Chen and Yunpeng Yin and Limin Yang and Qingxing Xi and En Yang and Xuefeng Luo}, doi = {10.1109/TASE.2025.3596630}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20296--20310}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Force-compliance MPC and robot-user CBFs for interactive navigation and user-robot safety in hexapod guide robots}, volume = {22}, year = {2025}, }
• (2025). Autonomous PID tuning: Two-phase reinforcement learning through adversarial imitation learning under imperfect demonstrations. TASE, 22, 20280--20295. (www)

  @article{ , abstract = {This paper presents stable reinforcement learning through adversarial imitation learning (SRAIL), a two-phase learning framework for automatic proportional integral derivative (PID) tuning in the absence of high-quality demonstration data. First, adversarial imitation learning (AIL) is used to learn an initial control policy using imperfect demonstration data. Subsequently, reinforcement learning (RL) refines this initial policy to derive the optimal PID control policy. SRAIL is designed to ensure stable and safe exploration of the RL agent, even with imperfect demonstration data, thereby enabling efficient and automatic tuning in industrial environments. Conventional PID tuning methods are time-consuming owing to system complexity and lack of prior knowledge. Furthermore, the random exploratory nature of RL presents significant risks during the early training phase, limiting its practical application. SRAIL addresses these challenges by ensuring stable control performance despite the absence of high-quality demonstration data and by enhancing the convergence speed of RL. The experimental results demonstrate that SRAIL minimizes random exploration in the initial learning phase, ensuring safety, and highlighting its potential for implementation of RL in real-world industrial environments. Note to Practitioners—This study is motivated by the necessity of efficiently automating the tuning of PID control systems without requiring direct expert intervention in manufacturing environments. PID tuning is considerably affected by changes in the objectives and targets of the control system, which requires constant adaptation. RL can address these challenges by learning and adapting to environmental changes. However, its inherent trial-and-error approach presents difficulties in an industrial environment, where strict safety constraints must be ensured. To mitigate these limitations, learning from demonstrations (LfD) is often employed; however, this method is limited by the requirement for high-quality demonstration data and data collection challenges. This study presents a novel framework capable of learning effectively even in the absence of perfect data and applies it to real-world manufacturing scenarios. It reduces the risks associated with the tuning process and prevents equipment damage. The experimental results demonstrate that SRAIL exhibits superior stability and performance. Future research will focus on validating the framework’s scalability and performance by utilizing a broader range of incomplete demonstration data.}, archive = {J_TASE}, author = {Gon Kim and Kyuchang Chang and Yunseon Byun and Jun-Geol Baek}, doi = {10.1109/TASE.2025.3600504}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20280--20295}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous PID tuning: Two-phase reinforcement learning through adversarial imitation learning under imperfect demonstrations}, volume = {22}, year = {2025}, }
• (2025). Lyapunov-informed multi-agent reinforcement learning for multi-robot cooperation tasks. TASE, 22, 20263--20279. (www)

  @article{ , abstract = {Multi-Agent Reinforcement Learning (MARL) has shown great potential in solving complex tasks. Despite great success, low training efficiency remains a pervasive and long-standing challenge in MARL. To tackle this issue, it is promising to leverage prior knowledge or environmental properties to inform and improve the MARL. We notice that many multi-agent tasks specify certain goal states where special rewards are granted, guiding agents to achieve the goal. Inspired by the theory of Lyapunov stability, an intuitive optimal policy to the tasks should be able to asymptotically converge to the goal states from any initial, making the goal states stable equilibria. Focusing on this type of tasks, we introduce the concept of Lyapunov Markov game (LMG), a new subclass of the cooperative Markov game, featuring a set of goal states and goal-oriented reward function. We then provide a theoretical bound on scaled value distance as a necessary condition to obtain a stable suboptimal policy in LMG. Motivated by this insight, we further propose the Lyapunov-informed MARL, which leverages a newly-designed Lyapunov-informed reward. Theoretical work is conducted to show that the Lyapunov-informed MARL enjoys a broadened bound, facilitating the training process to find a stable suboptimal policy more easily and then converge to an optimal policy more efficiently. Extensive experiments and real-world multi-robot implementations are conducted to show the superior performance of the proposed approach over advanced baseline models. Note to Practitioners—This paper addresses the training efficiency problem in multi-robot cooperation using MARL. Coordinating multi-robot systems, such as uncrewed ground vehicles and drones, to accomplish tasks efficiently is challenging. While MARL offers potential solutions, its slow training process limits practical applications. Leveraging human expert knowledge can significantly improve this issue compared to relying solely on environment interaction. This paper integrates Lyapunov stability theory with MARL to discover suboptimal stable policies as stepping stones towards the optimal policy. We first find a feasible policy and then refine it to achieve better performance (e.g., shorter paths or lower resource consumption). We provide a rigorous theoretical analysis of the proposed method, including proofs of the suboptimal policy bounds and how the embedded Lyapunov method extends these bounds. We provide a general reward design technique for reinforcement learning applications, and this Lyapunov-informed method effectively enhances MARL performance. The proposed method is validated through both simulation and real-world experiments.}, archive = {J_TASE}, author = {Pu Feng and Rongye Shi and Size Wang and Qizhen Wu and Xin Yu and Wenjun Wu}, doi = {10.1109/TASE.2025.3603024}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20263--20279}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lyapunov-informed multi-agent reinforcement learning for multi-robot cooperation tasks}, volume = {22}, year = {2025}, }
• (2025). Observer-based containment control for multiagent systems by a node-based hybrid triggering mechanism. TASE, 22, 20249--20262. (www)

  @article{ , abstract = {This article considers the observer-based containment control problem of multiagent systems (MASs), with a focus on the scenario where the control task only utilizes intermittent communication among agents and from the sensor to the local state observer. To deal with this problem, we propose a hybrid triggering control strategy, which includes two parts. First, a local observer for each follower is proposed to estimate state under a time-triggered mechanism (TTM). Second, based on local observer, a fully distributed containment controller is proposed under a node-based hybrid triggering mechanism (HTM), which does not require global information. Note that such HTM includes a TTM and a dynamic event-triggered mechanism (DETM). The most significant merit of this article is that the TTM is developed to provide the guarantee of the positive minimum triggering interval time (MTIT) between two triggering actions, which is more desirable than simply excluding Zeno behavior in practical applications. Finally, a simulation example on containment of nonholonomic mobile robots illustrates the control strategy. Note to Practitioners—The observer-based containment control issue of multiagent systems (MASs) is addressed under hybrid triggering communication, which can be implemented in many practical control systems. In general, each agent often has only limited communication resources in engineering applications. Besides, owing to high cost or great difficulty, the full state information of each agent is not always measurable in some situations. Considering that unmeasurable state and limited communication resources, we propose an observer-based hybrid triggering control strategy, which can trigger the observer and the controller, respectively. Thus, the sensor and communication burdens can be significantly reduced. Note that the guarantee of the positive minimum triggering interval time (MTIT) makes the method more suitable for real-world applications than traditional event-triggered protocols. In simulations, the control strategy is illustrated by a practical example on containment of nonholonomic mobile robots.}, archive = {J_TASE}, author = {Wang Yang and Jiuxiang Dong and Ju H. Park}, doi = {10.1109/TASE.2025.3597923}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20249--20262}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based containment control for multiagent systems by a node-based hybrid triggering mechanism}, volume = {22}, year = {2025}, }
• (2025). A high-DOF BCI control strategy mapping discrete commands to continuous motion for a drone. TASE, 22, 20234--20248. (www)

  @article{ , abstract = {Because of the non-stationary nature of electroencephalogram (EEG) signals, traditional non-invasive brain-computer interfaces (BCIs) usually only produce discrete commands, limiting their ability to control external devices continuously. This study proposes a novel BCI control strategy mapping multiple discrete commands to continuous motion, enabling real-time manipulation of a drone in four degrees of freedom (DOF). Our strategy used the fast steady state visual evoked potential (SSVEP) encoding and decoding method to convert user intentions into the drone’s flight status in near real-time. Simultaneously, the drone’s live video was embedded into the SSVEP stimuli, providing users with a first-person perspective control experience. In drone control experiments, participants successfully maneuvered the drone through complex path-following tasks in simulated and physical scenarios. The mean flight trajectory bias ratio was measured as 0.81, with a mean flight smoothness of -3.31 (measured by spectral arc length) and mean Fitts’s throughput of 9.18 bits/min. Notably, the brain-to-hand ratio (BHR) for all metrics approached 1, indicating that our non-invasive control system achieved comparable performance to manual control systems. These results suggest the effectiveness of our proposed BCI control strategy that maps discrete commands to continuous motion and extends the capabilities of non-invasive BCIs in continuous control scenarios. This study significantly advances the applications of BCI and propels human-machine interaction towards a more direct realm. Note to Practitioners—This work is motivated by the real-time and continuous control challenges in using non-invasive brain-computer interfaces (BCIs). Non-invasive BCIs establish a direct communication pathway between the human brain and external devices. Due to their safety and convenience, they are considered a promising human-machine interaction method for future practical applications. However, a major bottleneck is that non-invasive BCIs that generate discrete commands are unable to meet the demands of continuous control for external devices. Additionally, the low accuracy and high latency in decoding commands from human intent further distance current brain-control systems from practical application. In this paper, we propose a BCI strategy that includes a continuous encoding and decoding method for EEG signals. This strategy reduces the latency of command output while maintaining both the size of the command set and decoding accuracy. Furthermore, it introduces a method to map the decoded commands to continuous movements of external devices, effectively bridging the existing gap. We conducted extensive simulations and real-world experiments, demonstrating that the proposed strategy allows users to achieve continuous, real-time, four-degree-of-freedom control of a quadcopter using a non-invasive BCI. The analysis of results shows that, in terms of control accuracy, continuity, and operational efficiency, brain control is comparable to manual control. In the future, we aim to enhance the external devices with artificial intelligence to enable collaborative control between humans and machines, thereby advancing the practical development of BCIs.}, archive = {J_TASE}, author = {Jie Mei and Weize Chen and Ang Li and Yongzhi Huang and Xiaolin Xiao and Kun Wang and Weibo Yi and Tzyy-Ping Jung and Minpeng Xu and Dong Ming}, doi = {10.1109/TASE.2025.3593497}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20234--20248}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A high-DOF BCI control strategy mapping discrete commands to continuous motion for a drone}, volume = {22}, year = {2025}, }
• (2025). PG-ITD3: A potential field-guided deep reinforcement learning approach for UAV path planning after disaster. TASE, 22, 20221--20233. (www)

  @article{ , abstract = {To address the target inaccessibility issue of the Artificial Potential Field (APF) method and the slow convergence of deep reinforcement learning algorithms, we propose a method combining the artificial potential field and an improved twin delayed deep deterministic policy gradient algorithm (PG-ITD3) for UAV path planning in post-disaster scenarios. This approach considers Uncrewed Aerial Vehicles (UAV) kinematic constraints and dynamically adjusts the repulsive gain coefficient through deep reinforcement learning to enhance planning efficiency. The introduction of a virtual obstacle strategy, combining the aftershock probability model with prioritized experience replay (PER), and a novel reward function facilitates real-time reward acquisition and accelerates convergence. Simulation results demonstrate that our proposed algorithm outperforms traditional APF method and other advanced deep reinforcement learning methods in success rate, path length, path reward, and global smoothness. This research offers an effective solution for post-disaster UAV rescue path planning, contributing significantly to enhancing rescue efficiency and safety. Note to Practitioners—This study introduces PG-ITD3, a hybrid path planning solution for UAVs in post-disaster environments with complex obstacles. By integrating APF method with an enhanced deep reinforcement learning algorithm (TD3), the method dynamically optimizes repulsive gain coefficient and accounts for UAV kinematic constraints. Key innovations include: (1) virtual obstacle deployment for proactive collision avoidance, (2) prioritized experience replay to accelerate training, and (3) a refined reward function ensuring real-time feedback. Static and dynamic simulations and field test demonstrate higher success rates and shorter paths compared to conventional methods. Practitioners can implement this framework to enhance mission reliability, reduce response times, and improve safety in disaster zones. For optimal results, prioritize fine-tuning virtual obstacle thresholds and reward coefficients based on specific operational environments.}, archive = {J_TASE}, author = {Xiaohai Ren and Na Geng and Yong Zhang and Lei Xiao and Dunwei Gong}, doi = {10.1109/TASE.2025.3601665}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20221--20233}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PG-ITD3: A potential field-guided deep reinforcement learning approach for UAV path planning after disaster}, volume = {22}, year = {2025}, }
• (2025). Design method of quasilinear nonaffine nonlinear control systems of general structure. TASE, 22, 20208--20220. (www)

  @article{ , abstract = {The paper is devoted to present a new original design method of nonaffine nonlinear control systems of general structure. The originality of the method lies in the application to a nonlinear nonaffine plant of the general structure of an algebraic polynomial matrix design method using quasi-linear models. This method allows finding control by solving a linear algebraic system and an additional nonlinear algebraic equation. In many cases, control can be found analytically, which ensures high efficiency of the method. This research considers nonlinear plants that are nonaffine in control and disturbance. The analytical method for creation of the quasilinear models from the plant differential equations in the Cauchy form based on the theorem of the integral independence on the integration path is given in the article in brief. The design problem of the nonaffine nonlinear system has solution by this method, if the quasilinear model of the plant is controllable on the state and has a controlled output. The article presents the controllability criterion on the state and controllability criterion of the output. The functional parameters of the nonlinear control are defined by solving a system of linear algebraic equations and an additional nonlinear algebraic equation. In the system with this control, the tracking error is zero, if the reference input is stepwise, and it is sufficiently small if the reference input is periodic. The effectiveness of the suggested method is illustrated by three numerical examples of the design of nonaffine nonlinear control systems of the second and third order. Note to Practitioners— This paper focuses on creating nonlinear control systems for nonaffine nonlinear plants, which are common in actual practice. The known methods of designing nonaffine nonlinear systems are highly complex, since their application requires nonlinear transformations of the initial equations. Generally, these transformations are rather difficult to find. The new method presented in this article is based on quasilinear models and algebraic polynomial matrix method for designing nonlinear Hurwitz control systems, and is very simple and easy to implement in practice using modern technical means of automation. It is applied in conditions when the nonlinearities of the plant are differentiable, the state vector is measurable, and the plant’s quasilinear model satisfies the controllability criterion on the state and the controllability criterion of the output. The advantages of the proposed method are that the coefficients of the quasilinear model are easily found from the initial nonlinear equations of the plant. According to this model, a number of polynomials are found, and a system of linear algebraic equations is compiled using the coefficients of these polynomials. The parameters of the control law are determined as the solution of this algebraic system and of a single nonlinear algebraic equation. The three numerical examples of designing control systems for nonaffine nonlinear plants presented in the paper demonstrate the simplicity of the proposed method and the high quality of the control or tracking processes. This method can be applied to create highly efficient nonaffine nonlinear control systems for various industries: shipbuilding, aircraft, automobile, machine tools, agriculture and many others.}, archive = {J_TASE}, author = {Anatoly R. Gaiduk and Mikhail Yu. Medvedev and Viacheslav Kh. Pshikhopov}, doi = {10.1109/TASE.2025.3599894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20208--20220}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design method of quasilinear nonaffine nonlinear control systems of general structure}, volume = {22}, year = {2025}, }
• (2025). An improved iterative learning control approach based on random forest and whale optimization for urban traffic signals. TASE, 22, 20195--20207. (www)

  @article{ , abstract = {Iterative learning control (ILC) has demonstrated effectiveness in urban traffic signal control systems. However, conventional ILC methods typically require infinite iterations to achieve theoretical convergence conditions, leading to neglect of convergence speed in practical applications. As convergence speed primarily depends on the initial control signal selection and learning gain configuration, current applications in urban traffic signal control face two limitations: 1) empirical initialization of control signals often induces significant initial iteration errors, and 2) fixed learning gains prevent real-time adaptation to system errors during iterations. In order to address the aforementioned issues, we propose a novel urban traffic signal iterative learning control method based on random forest and whale optimization, and a faster convergence speed is achieved by integrating the iterative initial value and learning gain optimization algorithms. In the initial iteration of ILC, a trained random forest regression model by collecting the historical iteration data is employed to predict the initial control signals. Subsequently, the whale optimization algorithm is used to determine the learning gain of each iteration, and online optimization is performed based on the error minimum principle. Finally, a rigorous proof of convergence is conducted for the proposed method, and ablation and comparative experiments are conducted using the road network between Changfeng Street and Xuefu Street in Taiyuan to verify the effectiveness of the proposed method. Compared with other control approaches, the proposed method achieves faster convergence speed and lower tracking error under varying disturbances. Note to Practitioners—The inspiration for this study stems from the convergence speed of ILC algorithms in urban traffic signal control. Similarly, this concept can be applied to ILC algorithms in other fields. The convergence proof of the existing ILC algorithms typically rely on an infinite number of iterations, leading to the neglect of the convergence speed. To fill this gap, this study proposes to build a historical iteration information dataset by collecting the historical iteration data, and the dataset is used to train a random forest regression model. Subsequently,the trained random forest regression model is used to predict the superior initial iteration control signal, which can accelerate the convergence speed of ILC. In addition, the whale optimization algorithm is also utilized to optimize the learning gain to further accelerate the convergence speed. Compared with existing methods, our proposed method has faster convergence speed and higher tracking accuracy.}, archive = {J_TASE}, author = {Fei Yan and Xiaohan Zhang and Yingmin Yi and Zixuan Huang}, doi = {10.1109/TASE.2025.3600027}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20195--20207}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An improved iterative learning control approach based on random forest and whale optimization for urban traffic signals}, volume = {22}, year = {2025}, }
• (2025). Cascade control for post-chlorine dosage during drinking water treatment under cyber attacks. TASE, 22, 20182--20194. (www)

  @article{ , abstract = {In the water treatment process, the free chlorine residual (FCR) of treated water is a critical metric for the post-chlorination process. However, the stability of this process is significantly compromised by potential cyber attacks targeting communication channels, as well as fluctuations in influent water quality and flow rates. The stringent accuracy requirements of post-chlorination control necessitate a control scheme with fast system response and high robustness under sudden cyber attacks and disturbances. In this context, traditional single-loop feedback control schemes can no longer effectively stabilize the FCR of treated water. This study proposes a cascade (master/slave) feedback control scheme to address the issue of post-chlorine dosage control. The master controller is developed based on a second-order sliding mode (SOSM) control, which utilizes the FCR of the clear-water reservoir inlet as its reference. An internal model control (IMC) approach is then adopted in the slave controller, enabling the FCR of the clear-water reservoir inlet to track the reference acquired from the master controller. It is shown in simulation results that, compared with conventional single-loop feedback control approaches, the proposed cascade control strategy exhibits superior robustness against cyber threats, disturbances, and model uncertainties, along with rapid system response. The proposed strategy thus provides a reliable and efficient solution for maintaining water quality under complex and varying operational conditions. Note to Practitioners—Ensuring stable control of FCR in water treatment systems is critical, particularly when facing challenges such as cyber attacks and fluctuating water conditions. Traditional control methods often fail to meet the robustness and accuracy required in these dynamic environments. This study introduces a cascade control framework combining a master controller (SOSM) with a slave controller (IMC). This system is designed to address disturbances and cyber threats while maintaining fast response times and high accuracy. Practitioners managing water treatment facilities can benefit from this approach by integrating it into existing control systems to enhance resilience and maintain water quality. The proposed method adapts to fluctuating operational demands, offering a practical solution to modern water treatment challenges.}, archive = {J_TASE}, author = {Dongsheng Wang and Yicong Yuan and Yukun Song and Shunyuan Xiao and Le Chang}, doi = {10.1109/TASE.2025.3599161}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20182--20194}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cascade control for post-chlorine dosage during drinking water treatment under cyber attacks}, volume = {22}, year = {2025}, }
• (2025). Neuromorphic event camera-based object recognition and grasping position detection using a transfer learning-enhanced multi-task model. TASE, 22, 20169--20181. (www)

  @article{ , abstract = {Object recognition and grasping position detection are critical tasks in robotic manipulation, particularly when operating in dynamic and unstructured environments. This paper presents the Channel Sharpening Attention-based Adaptive Inception Network (CSA-AInceptNet), a novel multi-task learning model designed for these tasks using event camera data. The proposed architecture integrates channel sharpening attention with adaptive inception networks to enhance feature extraction and improve robustness. The model’s performance is evaluated on two state-of-the-art event camera datasets, E-Grasp and Neuro-Grasp. On the E-Grasp dataset, CSA-AIncepNet achieves a remarkable accuracy of 99.47% and a mean Intersection over Union (IoU) of 0.9370, significantly surpassing existing methods. On the Neuro-Grasp dataset, leveraging transfer learning, the model attains 98.58% accuracy and a mean IoU of 0.4897, demonstrating strong generalization capabilities across datasets. Comparative analyses and ablation studies further validate the effectiveness of the proposed architecture, highlighting its superiority over conventional models like ConvNeXt, DarkNet, DenseNet, and VGG16. The results establish CSA-AIncepNet as a robust solution for event-based object recognition and grasping detection, paving the way for advancements in human-robot collaboration and dynamic robotic manipulation. Note to Practitioners—This work provides a practical solution for improving object recognition and grasping position detection in robotic systems, particularly in unpredictable and fast-changing real-world environments. By leveraging event camera data, the proposed approach enables robots to efficiently identify objects and determine optimal grasping positions, even under challenging conditions. The results highlight the model’s ability to outperform existing methods, making it highly suitable for applications such as human-robot collaboration and precise object handling. This advancement has significant implications for industries like manufacturing, logistics, and healthcare, where robots must interact with objects quickly and accurately. Practitioners can adopt this method to enhance robotic performance, reduce errors, and improve operational efficiency. Future work could focus on testing the model in more complex environments and adapting it for real-time deployment in dynamic settings.}, archive = {J_TASE}, author = {Muhammad Hamza Zafar and Syed Kumayl Raza Moosavi and Filippo Sanfilippo}, doi = {10.1109/TASE.2025.3598695}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20169--20181}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Neuromorphic event camera-based object recognition and grasping position detection using a transfer learning-enhanced multi-task model}, volume = {22}, year = {2025}, }
• (2025). Event-triggered strategy design for demand response management in heterogeneous agent-based smart grids. TASE, 22, 20160--20168. (www)

  @article{ , abstract = {This paper presents a novel event-triggered approach for regulating energy prices of heterogeneous agent-based demand response including prosumer groups and energy companies. First, a price-based demand response management (PDRM) model is developed using networked evolutionary game theory. Then, a necessary and sufficient criterion is proposed to verify the stabilizability of the controlled PDRM system, significantly reducing iterative complexity compared to existing methods. By utilizing truth matrices, an algorithm is proposed to calculate the event-triggered feedback gain matrices and determine the triggering times. The proposed matrix-based event-triggered control guarantees that the strategies of all participants converge to the target game equilibrium set while reducing the update frequencies of the controller. Finally, the simulation results demonstrate the effectiveness of the proposed method. Note to Practitioners—Existing game-theoretic demand response management models are challenging to analyze quantitatively regarding the decision-making processes within heterogeneous agent-based smart grids. This paper presents a novel PDRM model for heterogeneous agent-based smart grids. The proposed PDRM model effectively simulates the dynamic evolution of decision-making processes for each participant. In addition, this paper presents a matrix-based event-triggered control approach for dynamically regulating energy prices to the desired level. With an arbitrary initial strategy, the proposed control algorithm ensures that the controlled PDRM system converges to the desired energy price, while reducing the number of adjustments required. The proposed method provides flexible adjustments based on actual conditions for demand response scheduling in heterogeneous agent-based smart grids.}, archive = {J_TASE}, author = {Qiliang Zhang and Han Wang and Ya Zhao and Ze Wang and Bo Wei and Guang Wang}, doi = {10.1109/TASE.2025.3601160}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20160--20168}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered strategy design for demand response management in heterogeneous agent-based smart grids}, volume = {22}, year = {2025}, }
• (2025). A production scheme design method for cloud manufacturing service systems with resource uncertainty using transition-rewritable petri nets. TASE, 22, 20146--20159. (www)

  @article{ , abstract = {Cloud manufacturing (CMfg) service systems can provide solutions for large-scale personalized customization. Service composition optimal selection, which is in charge of the selection and allocation of cloud services to multiple tasks, is of great importance for providing significant improvement of CMfg service systems with respect to performance measures such as time and cost. To increase resource utilization and decrease cost consumption, we develop a design method of production schemes for CMfg service systems. Specifically, we define a formalism named transition-rewritable Petri nets (RTPNs) to model and reconfigure the multi-task manufacturing processes in environments allowing resource uncertainty. The reconfiguration process is described by the rewritable rules of RTPNs. On this basis, the performance of the reconfiguration of the manufacturing process is analyzed. In addition, we establish a production strategy considering reconfiguration to obtain the optimized service compositions in terms of time, cost, quality, and reliability. Theoretical comparisons highlight the distinctions between our method and the others in the existing literature. Moreover, the results of simulation experiments and comparative studies show that the method proposed in this paper can obtain better schemes with the lower total completion time and total production cost than a closely-related one presented in a recently published paper. Note to Practitioners—The uncertainty of resources exists when using cloud manufacturing service system based on distributed resources to complete manufacturing tasks. However, for the large-scale orders, the resource uncertainty is rarely considered by existing work in the literature. This paper proposes a production strategy to ensure that the utilized schemes are always feasible and optimal despite resource uncertainty. A type of Petri nets is defined to model the manufacturing process for the purpose of reconfiguration of the production schemes when the some services change due to resource alterations. Total completion time and total cost are significant considerations. A problem decomposition based genetic algorithm is employed to generate production schemes, which can be easily implemented.}, archive = {J_TASE}, author = {Hao Yue and Yingtao Wu and Min Wang and Hesuan Hu and Shanchen Pang}, doi = {10.1109/TASE.2025.3587235}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20146--20159}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A production scheme design method for cloud manufacturing service systems with resource uncertainty using transition-rewritable petri nets}, volume = {22}, year = {2025}, }
• (2025). A human-centered task allocation and scheduling framework for multi-human-multi-robot collaboration in precision agriculture settings. TASE, 22, 20126--20145. (www)

  @article{ , abstract = {Human-multi-robot teaming in precision agriculture represents a promising approach to addressing labor shortages and managing the complexities of agricultural practices. An effective coordination of these teams, including task allocation and scheduling strategies while accounting for the inherent unpredictability of human behavior, is crucial for maximizing system productivity and ensuring user comfort. In this study, we introduce a Mixed-Integer Linear Programming (MILP) approach that aims to minimize workers’ waiting times, robots’ energy consumption during the different phases of the robots’ motions, and the overall makespan. To enhance the robustness of our framework and consider human preferences, a user interface is designed to capture real-time human feedback; then, an adaptive online updating strategy that dynamically adjusts plans responding to variations in human operators’ parameters is devised. To handle large-scale problems, we extend the solution approach by leveraging Constraint Programming (CP) combined with a batch decomposition strategy. The approach is validated through extensive simulations in a Unity-based realistic virtual reality environment and laboratory experiments using two TurtleBot2 robots and two human operators performing grape harvesting tasks. Note to Practitioners—This paper was inspired by the necessity to coordinate a heterogeneous team of humans and robots within an agricultural setting to optimize relevant performance indices. The proposed approach focuses on a scenario where mobile service robots perform assistance activities for human and robotic working agents. It enables the allocation and scheduling of all tasks while accounting for the following key factors: i) the different characteristics of the agents, ii) their variability due to changing environment conditions and human dynamic behavior, also captured through chance-constrained programming, and iii) human preferences and feedback provided in real-time. Furthermore, an extension to handle large-scale systems is proposed. Beyond agricultural applications, this approach applies to various domains where cooperation among heterogeneous agents, including logistics, industry, and search and rescue, can be advantageous. Realistic simulation results and laboratory experiments validate the effectiveness of the approach. As future work, we aim to integrate human intent prediction strategies to proactively adapt plans based on anticipated human actions, further enhancing coordination and efficiency.}, archive = {J_TASE}, author = {Jorand Gallou and Martina Lippi and Jozsef Palmieri and Andrea Gasparri and Alessandro Marino}, doi = {10.1109/TASE.2025.3595413}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20126--20145}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A human-centered task allocation and scheduling framework for multi-human-multi-robot collaboration in precision agriculture settings}, volume = {22}, year = {2025}, }
• (2025). Multi-target coverage trajectory planning for ceiling painting robot chassis via two-stage optimization. TASE, 22, 20112--20125. (www)

  @article{ , abstract = {Ceiling painting is an essential yet labor-intensive task in the construction industry, making automation necessary to overcome issues such as inconsistent manual quality and associated health hazards. This paper addresses the multi-target coverage trajectory planning problem for ceiling painting robots operating in complex indoor environments, aiming to generate efficient, collision-free, and kinematically feasible trajectories that fully cover all designated peeling areas. The planning problem is first formulated as a mixed-integer optimal control model and then discretized into a mixed-integer nonlinear program. Conventional solvers struggle with prohibitive computational complexity due to integer variables and non-convex constraints arising from kinematic feasibility, collision avoidance, and coverage requirements. To overcome these computational challenges, we propose a two-stage planning structure. In the first stage, the planner constructs a coarse trajectory by integrating waypoint clustering, probabilistic roadmap-based collision-free path planning, and Traveling Salesman Problem optimization. The second stage addresses the non-convexity of the coverage constraints through an alternating optimization approach, iteratively fixing integer and continuous variables to achieve convexification and refine the trajectory. To further improve computational efficiency and scalability, the trajectory is segmented and each segment optimized independently. Simulation results demonstrate the effectiveness and reliability of the proposed planner, highlighting significant reductions in computational time compared with baseline methods. Note to Practitioners—This work presents a practical trajectory planning framework for automating ceiling painting tasks. The proposed two-stage approach enables robots to efficiently cover scattered peeling areas in complex indoor settings while avoiding collisions. Key advantages for real-world deployment include: 1) Fast Computation: The algorithm is over $1000\times $ faster than traditional solvers, making it suitable for use on standard industrial computers; 2) Improved Safety and Quality: The smooth planned paths reduce mechanical stress and ensure even, reliable paint coverage; 3) High Flexibility: The method adapts to ceilings of various shapes without needing pre-made templates. Engineers can easily integrate this planner into existing robotic systems with manipulators, helping reduce manual labor and exposure to hazardous environments.}, archive = {J_TASE}, author = {Sheng Yang and Chaoyi Sun and Bai Li and Li Li}, doi = {10.1109/TASE.2025.3599163}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20112--20125}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-target coverage trajectory planning for ceiling painting robot chassis via two-stage optimization}, volume = {22}, year = {2025}, }
• (2025). Plug-and-play distributed fencing of multi-agent systems for an uncooperative target without velocity measurements. TASE, 22, 20098--20111. (www)

  @article{ , abstract = {This paper investigates the target fencing problem of multi-agent systems (MASs) for an uncooperative, freely-moving target whose position is accessible solely to a subset of agents. A plug-and-play distributed target fencing strategy is proposed, comprising a finite-time target acceleration estimator, a distributed target motion observer, and a label-free target fencing controller. Under the proposed fencing strategy, all agents can asymptotically observe the target position and acceleration. Moreover, without needing velocity measurement, the target is asymptotically fenced with the velocity convergence and the avoidance of inter-agent collisions. Furthermore, the plug-and-play feature of the proposed strategy ensures that each agent can join or leave the MAS during the fencing process, and the activated agents can still fence the target via self-organization without the need to adjust any parameters of the activated agents. The results are proved by rigorous theoretical analysis and verified by numerical and real-world experiments. Note to Practitioners—This paper focuses on the target fencing problem of multi-agent systems (MASs), which has potential applications in target protection or assault for multiple vehicles. Unlike existing approaches that concentrate on a target with simple movement patterns, this paper considers a freely moving target whose position is measurable solely by a portion of agents, which is more practical in real fencing scenarios. A distributed fencing strategy is proposed to enable all agents to estimate the target position and acceleration and to cooperatively fence the target without measuring their own velocity. Under the proposed strategy, each agent self-organizes its behavior by following identical rules. Each agent can join or leave the MAS while the target remains fenced via the self-organization of remaining agents without modifying any agent’s parameters. This effectively reduces the complexity of configuring numerous agents, such as multiple vehicles, in engineering applications under dynamic situations. The strategy can be applied to multiple vehicles for uncooperative target assaults in complex situations where vehicles are damaged at unknown instants and supported when needed.}, archive = {J_TASE}, author = {Song Jiang and Pei Chi and Luning Yang and Jiang Zhao and Yingxun Wang}, doi = {10.1109/TASE.2025.3596929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20098--20111}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Plug-and-play distributed fencing of multi-agent systems for an uncooperative target without velocity measurements}, volume = {22}, year = {2025}, }
• (2025). Change-driven analysis of timed security requirements for data-aware workflow processes. TASE, 22, 20083--20097. (www)

  @article{ , abstract = {In recent years, due to the complexity of workflow processes (or business processes) and the fluctuation of business environments, the timed security requirements (or constraints) of workflow processes have attracted more and more attentions from the designers and users of workflow processes. The existing methods cannot analyze the security requirements of data-aware workflow processes with dynamic changes. In this paper, we propose a novel change-driven security analysis approach (CDSA approach) to verify timed security requirements for data-aware workflow processes with dynamic changes. First, we analyze the feasibility of changes on workflow processes. Second, we propose the analysis procedure to automatically determine the affected segments of changes. Third, the timed security requirements affected by changes are obtained and verified without repeatedly analyzing all of them. Compared with the existing works, our approach can detect incorrect changes in advance so as to avoid meaningless analysis, also can reduce the cost and time of enterprises as well as eliminate repeated verification of all timed security requirements. Note to Practitioners—Security analysis of workflow processes (or business processes) is an emerging but crucial area for enterprises and supply chains. In this paper, we present a novel approach to analyze data-aware workflow processes with dynamic changes under timed security requirements. It is structured into three phases: 1) feasibility analysis of changes, 2) determination of affected segments, and 3) verification of timed security requirements. The proposed approach can detect incorrect changes in advance and avoid repetitive analysis, but also effectively and efficiently verify the timed security requirements after changes by determining the affected workflow processes and segments. This can greatly improve the efficiency of verification, so that the cost and time of enterprises can be significantly reduced. A real-life case is used to illustrate the feasibility of the proposed concepts and approach.}, archive = {J_TASE}, author = {Yanhua Du and Gege Mu and Junfen Li and Hesuan Hu}, doi = {10.1109/TASE.2025.3578550}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20083--20097}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Change-driven analysis of timed security requirements for data-aware workflow processes}, volume = {22}, year = {2025}, }
• (2025). Automation-driven robust neural predictive control for DC shipboard microgrids. TASE, 22, 20068--20082. (www)

  @article{ , abstract = {The reliable and stable operation of modern shipboard power systems (SPSs) is crucial for mission success, particularly when faced with high-power pulsed loads and complex system interactions. While nonlinear model predictive control (NMPC) offers effective stabilization and robustness to bounded disturbances via its variant quasi-infinite-horizon NMPC (QIH-NMPC), it imposes a computational burden that limits real-time applicability for mission-critical SPSs at the microseconds level. This paper proposes a robust data-driven predictive control (DDPC) approach that leverages deep neural networks to approximate the QIH-NMPC control law, enabling fast-response and automated power management for medium-voltage direct current (MVDC) shipboard microgrids (MGs) that integrate synchronous generators, batteries, and supercapacitors. The proposed method includes a rigorous control design with theoretical guarantees and is validated under out-of-distribution (OOD) scenarios, aggressive load fluctuations with stochastic noise, stealthy false data injection attacks via numerical simulations, and fault conditions in a real-time simulator. Across diverse unseen disturbances, DDPC yields an average inference time of $6.91~\mu s$ , preserves control robustness by stabilizing the system within 0.05 seconds. It achieves a 97.8% reduction in computational load and improves control smoothness by 31% compared with NMPC, while also converging 50% faster than proportional-integral (PI) control. These findings highlight DDPC’s potential for automated, resilient, and real-time SPS control. Note to Practitioners—While NMPC offers strong performance, its computational demands hinder real-time feasibility, especially in nonlinear, fast-changing systems. This work introduces a practical, real-time control solution by learning robust NMPC policies offline using a data-driven (neural-based) predictive control framework. The trained model was tested beyond its training data to handle load surges, failures, and cyberattacks. The proposed method eliminates online optimization, offering fast control actions suitable for embedded systems. Practitioners can deploy the trained controller on real-time platforms (e.g., OPAL-RT and FPGAs) to manage voltage restoration and stability in shipboard microgrids. This approach is extendable to Uncrewed Aerial Vehicles (UAVs), electric vehicles, or smart grids, wherever fast, stable control with limited onboard computation is needed.}, archive = {J_TASE}, author = {Saskia A. Putri and Xiaoyu Ge and Faegheh Moazeni and Javad Khazaei}, doi = {10.1109/TASE.2025.3600293}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20068--20082}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automation-driven robust neural predictive control for DC shipboard microgrids}, volume = {22}, year = {2025}, }
• (2025). Tractor semi-trailer off-tracking and stability approximate bi-level policy optimization. TASE, 22, 20055--20067. (www)

  @article{ , abstract = {Trajectory tracking control of tractor semi-trailer vehicles poses significant challenges due to inherent off-tracking behavior and roll instability risks. While existing approaches have demonstrated effectiveness, they often rely on computationally intensive numerical solvers and require time-consuming manual tuning of cost function weights. This paper presents an approximate bi-level policy optimization (ABPO) framework that simultaneously optimizes the cost function and synthesizes an explicit control policy to minimize off-tracking while reducing computational complexity. The proposed framework employs a hierarchical structure: the upper level updates cost weights based on the trailer’s stability trajectory, while the lower level derives an approximate optimal policy by solving the tractor’s control problem. By leveraging Pontryagin’s Maximum Principle (PMP), we have developed a novel method to analytically compute cost weight gradients through differentiation of the PMP conditions. This enables the formulation of a related optimal control problem (OCP) whose solutions directly yield gradients for cost parameter updates. The ABPO framework achieves automatic weight coefficient adjustment, enhances trajectory tracking accuracy for both tractor and trailer units, and significantly reduces computational burden. Simulation and experimental validation across 4 classical scenarios demonstrates that the learned policy reduces rearward amplification by 17.82%, lateral tracking errors by 84.15%, and rollover by 64.19%, respectively. Notably, the control policy computation requires less than 10ms, making it suitable for real-time applications. The source code for the algorithms described in this paper is publicly available at https://github.com/TroyResearch/ABPO.git Note to Practitioners—This paper addresses a challenge in the autonomous driving industry: efficient and safe trajectory tracking control for tractor semi-trailer vehicles. Current industrial solutions typically require significant computational resources and time-consuming manual tuning of control parameters, limiting their practical implementation. Our proposed ABPO framework offers a practical solution by automating the parameter-tuning process and reducing computational complexity while maintaining safety standards. The framework can be readily integrated into existing complex control systems with minimal hardware requirements, making it suitable for real-time applications. However, when applying this framework to other control problems, the specific formulation of upper and lower-level problems needs to be carefully redesigned based on the particular problem characteristics to ensure the effectiveness of the control architecture. Future research will focus on extending this methodology to a broader range of control applications, like multi-robot control systems, autonomous fleet navigation, and imitation learning scenarios.}, archive = {J_TASE}, author = {Fawang Zhang and Jingliang Duan and Hui Liu and Xingyu Cao and Shida Nie and Congshuai Guo and Yujia Xie and Jun Ma and Shangli Wang}, doi = {10.1109/TASE.2025.3600553}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20055--20067}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tractor semi-trailer off-tracking and stability approximate bi-level policy optimization}, volume = {22}, year = {2025}, }
• (2025). Finite-time event-triggered h∞ control via predictive observer for state-dependent switched systems. TASE, 22, 20040--20054. (www)

  @article{ , abstract = {This article addresses the asynchronous control issue for a class of switched systems with norm-bounded external disturbances within a finite time interval. Firstly, a predicted model-based observer is constructed for switched systems. In the case of updates without actual output, the output predictor is introduced to compensate for the effects of an event-triggered scheme (ETS), while at triggered instants, the predicted output is reset to the latest triggered output. Secondly, a novel state-dependent switching (SDS) rule with fixed dwell time is developed within the framework of switched systems, integrating estimation performance, system performance, and ETS-designed dynamic variables. Thirdly, a new dynamic ETS related to both the output error and the predicted error is designed to optimize network resource utilization efficiency. Due to the ETS, the asynchronous behaviour among the controller, observer and system modes arises. Then, a novel multiple Lyapunov function is proposed to effectively characterize hybrid switching dynamics with coexisting synchronous and asynchronous modes in the closed-loop switched systems. Based on this framework, a unified finite-time ${\mathcal {H}}_{\infty }$ stabilization criterion is established, explicitly addressing switched systems containing partially or fully unstabilizable subsystems. Further, the Zeno behaviour of the designed ETS is ruled out by proving the existence of a minimum triggered interval. Finally, two practical simulations with comparative studies are conducted to substantiate the effectiveness of the derived results. Note to Practitioners—This work addresses practical challenges in controlling networked switched systems, such as those in autonomous vehicles, industrial automation, and smart grids, where limited network resources and partial state observability are critical concerns. Our approach focuses on three key aspects. Firstly, we develop a predictive observer to compensate for intermittent sensor data caused by event-triggered communication. This ensures reliable state estimation even during non-triggered intervals, enabling robust control without continuous sensing. Secondly, a novel ETS is designed to dynamically adjust data transmission based on real-time output prediction errors and system dynamics. Compared to some published ETS methods, it reduces unnecessary network traffic while maintaining control performance. Thirdly, our state-dependent switching strategy and hybrid Lyapunov framework rigorously guarantee finite-time stability, even when controller and observer modes operate asynchronously. The theoretical analysis and simulation results demonstrate that the proposed control method achieves the expected objectives.}, archive = {J_TASE}, author = {Xinyu Zhao and Hao Chen and Zhenzhen Zhang and Shouming Zhong}, doi = {10.1109/TASE.2025.3600499}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20040--20054}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time event-triggered h∞ control via predictive observer for state-dependent switched systems}, volume = {22}, year = {2025}, }
• (2025). Iterated extended kalman filter based on particle swarm optimization and its application to underwater robot navigation. TASE, 22, 20030--20039. (www)

  @article{ , abstract = {In this paper, the high precision and robustness of underwater robot navigation based on strap-down inertial navigation system/ultra short base line (SINS/USBL) is investigated. The Extended Kalman Filter (EKF) is used to SINS/USBL system. An Iterated Extended Kalman Filter (IEKF) is proposed to enhance the robustness, which is based on Levenberg Marquardt (LM) and Laplace principles. IEKF state vector mean square error matrix is optimized based on LM algorithm. Meanwhile, measurement noise variance matrix is constructed based on Laplace principle. Considering the strong nonlinear problem of system noise and measurement noise, Particle Swarm Optimization (PSO) algorithm is introduced, and an Improved Particle Swarm Optimization (IPSO) algorithm is proposed. The mean-square error fitness function based on the position error of the navigation system is constructed. An improved inertia weight calculation method is proposed to further improve the adaptability of IPSO algorithm. The IPSO algorithm can adaptive estimate the system noise variance matrix and measurement noise variance matrix of IEKF. Finally, the SINS/USBL tightly integrated model is designed for proposed IPSO-IEKF algorithm, which based on slant and azimuth information. Simulation and experiment demonstrate the effectiveness of the proposed IPSO-IEKF algorithm. The experimental results show that the positioning accuracy of the proposed method is improved by 70% compared with EKF algorithm. Note to Practitioners—The motivation of this paper is the combined positioning and anti-interference issues of the SINS/USBL integrated navigation system for underwater robots. EKF is a commonly used underwater integrated navigation algorithm. However, it often fails to address nonlinear constrained issues present in practical systems during application. Therefore, an Iterated Extended Kalman Filter (IEKF) is proposed to enhance the robustness, which is based on LM and Laplace principles. On the other hand, PSO algorithm is introduced in IEKF, and an improved PSO algorithm based on IEKF is proposed. Simultaneously, the proposed IPSO-IEKF algorithm enhances the system’s robustness against disturbances. The proposed method effectively addresses the nonlinear constrained issues in underwater navigation systems, thereby enhancing system stability. Simulation and experimentation demonstrate the feasibility of the method studied in this paper for practical engineering applications.}, archive = {J_TASE}, author = {Di Wang and Junwei Wang and Haoqian Huang and Guangcai Wang}, doi = {10.1109/TASE.2025.3600284}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20030--20039}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Iterated extended kalman filter based on particle swarm optimization and its application to underwater robot navigation}, volume = {22}, year = {2025}, }
• (2025). Finite time model predictive control for mobile manipulators with floating-base. TASE, 22, 20019--20029. (www)

  @article{ , abstract = {This article focuses on the trajectory tracking problem of mobile manipulators (MMs). Firstly, we construct a position and orientation model predictive tracking control (POMPTC) scheme for mobile manipulators. The proposed POMPTC scheme can simultaneously minimize the tracking error, joint velocity, and joint acceleration. Moreover, it can achieve synchronous control for the position and orientation of the end-effector. Secondly, a finite-time convergent neural dynamics (FTCND) model is constructed to find the optimal solution of the POMPTC scheme. Then, based on the proposed POMPTC scheme, a non-singular fast terminal sliding model (NFTSM) control method is presented, which considers the disturbances caused by the floating-base on the manipulator at the dynamic level. It can achieve finite-time tracking performance and improve the anti-disturbances ability. Finally, simulation and experiments show that the proposed control method has the advantages of strong robustness, fast convergence, and high control accuracy. Note to Practitioners—In actual applications, mobile manipulators need to operate on rugged terrains, over wavy seas, or in windy air, which can cause the base to tilt in uncertain directions or experience forces from different directions at any time. The manipulators need to consider these interference factors in a timely manner to adapt to changing environments. This can be seen as the manipulators being mounted on a floating-base. There have been a large number of studies on mobile manipulators, but there is still a lack of research on the control of mobile manipulators under high-frequency and large disturbances. In the face of different levels of disturbance, the manipulators need corresponding time to calculate and adjust their pose. However, for high-frequency and large disturbances, existing methods struggle to achieve stable control of mobile manipulators in a short time. Therefore, this article proposes a new control scheme to address the aforementioned issues. It should be noted that this method requires acquiring pose information of both the floating base and the manipulator’s end-effector. Consequently, sensors must provide the manipulator’s joint angles, angular velocities, accelerations, and torque information. Additionally, due to its substantial computational demands, the method requires a host computer with a CPU frequency greater than 3.5 GHz and memory exceeding 1 GB to ensure smooth system operation.}, archive = {J_TASE}, author = {Tao Su and Shiqi Zheng and Yixuan Guo and Yuanlong Xie and Chenglong Fu and Sheng Quan Xie}, doi = {10.1109/TASE.2025.3599904}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20019--20029}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite time model predictive control for mobile manipulators with floating-base}, volume = {22}, year = {2025}, }
• (2025). X-RepSLAM: VLM-driven adaptive cross-representation visual SLAM. TASE, 22, 20000--20018. (www)

  @article{ , abstract = {Visual Simultaneous Localization and Mapping (VSLAM) is a critical technology for autonomous driving and mobile robotics. Traditional VSLAM methods based on discrete representations, such as point clouds, offer high computational efficiency and excellent localization accuracy, but they exhibit limited robustness. In contrast, methods employing field representations, like Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3D GS), provide greater robustness at the expense of increased computational demands and reduced localization accuracy. Hybrid VSLAM approaches that attempt to combine these representations typically rely on serial, synchronous cascades, which compromise robustness, computational efficiency, and GPU memory usage. This paper introduces a novel adaptive cross-representation VSLAM framework that applies different representation modeling techniques to distinct regions of an image sequence and adopts asynchronous parallel modeling in overlapping regions. A Vision Language Model (VLM) is used to analyze the image sequence, enabling the detection of representation modeling regions and adaptive switching between representations. Cross-representation data association is performed through a coarse-to-fine feature selection process, resulting in a globally consistent map. The proposed method is evaluated on both public and custom-collected datasets, where experimental results show that it surpasses state-of-the-art methods in terms of robustness, computational efficiency, localization accuracy, and GPU memory usage. Note to Practitioners—X-RepSLAM is an adaptive cross-representation VSLAM framework designed to overcome the practical challenges of achieving robust performance, precise localization, and superior computational efficiency in autonomous systems while efficiently utilizing GPU resources. By leveraging the complementary strengths of discrete representations (e.g., point-based methods) and field-based techniques (e.g., NeRF or 3D GS), our approach uses a VLM to analyze incoming image sequences and identify regions with varying information richness. This enables an adaptive switching mechanism that selectively applies the optimal representation for each region, while asynchronous parallel modeling and a coarse-to-fine feature selection process ensure globally consistent mapping. Consequently, X-RepSLAM provides a flexible, efficient SLAM solution that dynamically adjusts to diverse and challenging environments, offering practitioners improved tracking robustness and localization accuracy with controlled computational demands. In particular, this adaptive framework is well-suited for mobile robot autonomous navigation tasks in scenarios with significant scene variation and limited computational resources. Practitioners can further tailor X-RepSLAM to their specific needs by designing VLM prompts that guide the system toward high-level objectives—such as semantic navigation or manipulation—thereby enabling task-driven mapping.}, archive = {J_TASE}, author = {Shilang Chen and Sehua Ji and Lei Zhu and Xuefeng Zhou and Hong Zhang and Weinan Chen and Yisheng Guan}, doi = {10.1109/TASE.2025.3599888}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {20000--20018}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {X-RepSLAM: VLM-driven adaptive cross-representation visual SLAM}, volume = {22}, year = {2025}, }
• (2025). A programmable pneumatic system with novel improved controller enabling adhesion and desorption operations of soft adhesive robots. TASE, 22, 19986--19999. (www)

  @article{ , abstract = {The inherent flexibility and nonlinearity of soft robots bring a challenge to their actuation and controlling. Especially, developing a portable drive and control system for bionic adhesive pneumatic soft robots (BAPSRs), with high accuracy and responsiveness to improve adhesion and desorption operations, remains a difficult task. In this paper, we aim to develop a high-integration programmable pneumatic system (hiPPS) for use in untethered, size-limited, weight-constrained soft robot applications. The complete design and operational mode are demonstrated, specifically, the dynamic electro-pneumatic characteristics of the system are illustrated through the construction of a microcontroller-based pressure control system that achieves high-integration and programmability. Based on the characteristic curve, an improved fuzzy piecewise PID feedback controller is proposed for the hiPPS system to achieve precise and continuous pressure supply and regulation, as well as to adjust the speed of inflation and deflation. Experimental results indicate that the hiPPS with the novel improved controller can achieve a wide range of pressure from -60 to 100 kPa, with pressure errors within 1.5 kPa, and outperformed a regular controller in realizing adhesion and desorption operations for BAPSRs. Our system liberates the pneumatic soft robots (PSRs) from tethered cumbersome air supply and trivial components by making a lightweight, small-scale pneumatic system with fine-grained control, which enables new possibilities for mobile soft robots. Note to Practitioners—The motivations of this article encompass two aspects: firstly, to provide a versatile, small-volume, lightweight, low-cost, and highly integrated pneumatic solution for PSRs that can generate hybrid pressure supplies, rivaling pneumatic systems composed of bulky pressure supplies and costly proportional valves; secondly, to address the issue of oscillation and fluctuation often encountered when adjusting the duty cycle of switching times in existing pneumatic systems composed of air pumps and low-cost on/off valves. In response to these challenges, we have developed the 93 g hiPPS system and proposed an improved control strategy that proactively adapts to the inherent characteristics of the system. This strategy employs special fuzzy piecewise processing of the reference pressure based on the actual tested electro-pneumatic characteristic curves of the system, thereby achieving continuous and precise pressure regulation, as well as controllable inflation and deflation speeds. The hiPPS system can be easily and rapidly applied to drive and control other single- or multi-chamber PSRs. In future research, we will extend the proposed system and strategy to hydraulic systems and the requirements of mobile PSRs.}, archive = {J_TASE}, author = {Xiangli Pei and Ruizhuo Shi and Liuwei Wang and Shuhao Liu and Zhiwei Wu and Zhendong Dai}, doi = {10.1109/TASE.2025.3599688}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19986--19999}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A programmable pneumatic system with novel improved controller enabling adhesion and desorption operations of soft adhesive robots}, volume = {22}, year = {2025}, }
• (2025). Distributed strategy design for autonomous multi-cluster games with high-order nonlinear physical players and its application to clustered electricity markets. TASE, 22, 19974--19985. (www)

  @article{ , abstract = {This paper considers the distributed autonomous multi-cluster games (MCGs) of physical players. In contrast to existing MCGs, the players in our MCG have high-order nonlinear dynamics. Moreover, these players are restricted by three kinds of constraints: local inequality constraints, intra-cluster and inter-cluster coupling inequality constraints. As far as we know, constrained MCGs with high-order nonlinear players have not been investigated. Existing distributed game strategies cannot handle our problem, mainly owing to the simultaneous presence of inequality constraints and high-order nonlinear dynamics. We exploit a distributed strategy, and give its performance analysis rigorously. With our strategy, those heterogeneous high-order nonlinear physical players do carry out MCG tasks autonomously. Also, the exact variational generalized Nash equilibrium (vGNE) of our MCG is sough via the strategy, even if the dynamics contains unknown parameters. Finally, the application of our method to clustered electricity market games (EMGs) is presented. Under our method, boiler-turbine generation systems (BTGSs) take part in clustered EMGs zealously and spontaneously, without the assistance of human beings. Note to Practitioners—The intellectualization of cyber-physical systems (CPSs) stimulates physical systems to execute distributed game tasks autonomously, e.g., the autonomous electricity market games of generators and the autonomous attack-defense games of unmanned aerial vehicles (UAVs). However, many existing works on games with physical systems do not involve constraints, and only consider first- or second-order (linear) dynamics. Besides, most existing methods cannot seek the exact GNE under the simultaneous appearance of constraints and system dynamics. Here our problem involves more complicated dynamics and constraints of players, which covers existing related problems. Moreover, under our method, complicated high-order nonlinear physical players can autonomously perform constrained games tasks, promoting the intellectualization of CPSs.}, archive = {J_TASE}, author = {Zhenhua Deng}, doi = {10.1109/TASE.2025.3599573}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19974--19985}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed strategy design for autonomous multi-cluster games with high-order nonlinear physical players and its application to clustered electricity markets}, volume = {22}, year = {2025}, }
• (2025). Predefined-time optimized tracking control for QUAVs with extended state observer and global prescribed performance. TASE, 22, 19961--19973. (www)

  @article{ , abstract = {This paper investigates the predefined-time optimized control problem for quadrotor uncrewed aerial vehicles (QUAVs) subject to the global prescribed performance constraints and partially unknown states. Under the backstepping framework, a predefined-time global prescribed performance optimized controller is developed through the actor-critic architecture, where the actor executes control actions and the critic evaluates their effectiveness. Furthermore, a neural network-based extended state observer is integrated into the optimized control approach to estimate unknown states and disturbances, enhancing the robustness of the control system, while the performance index is minimized. Notably, unlike most existing results, the requirement for the initial values of prescribed performance functions is removed, thereby reducing conservatism in the optimization of prescribed performance control. Additionally, a second-order command filter is introduced to overcome the “explosion of complexity”. It is demonstrated by the theoretical analysis that all closed-loop signals are bounded, while the position and attitude tracking errors subject to the global prescribed performance constraints converge into a neighborhood near zero within a predefined time. Finally, the feasibility and effectiveness of the proposed control strategy are validated through a simulation example. Note to Practitioners—Trajectory tracking serves as the foundation for QUAVs to accomplish complex flight missions. Among the key considerations in trajectory tracking are the requirements for tracking speed and precision. Moreover, unknown disturbances and partially unmeasurable states frequently occur. To overcome the aforementioned challenges, a predefined-time global prescribed performance optimized controller is developed, and the control cost is minimized. Compared to previous approaches, this paper eliminates the initial condition requirements of the prescribed performance approach. Furthermore, the extended observer is introduced to compensate for the effect of unknown disturbances and partial states. In addition, a second-order command filter is employed, avoiding the “explosion of complexity”. Through the theoretical analysis, all closed-loop signals are proved to remain bounded, and the tracking errors satisfy global prescribed performance, converging near zero within a preset time. A simulation confirms the validity of the proposed control strategy.}, archive = {J_TASE}, author = {Ce Wang and Wei Zhao and Shaoyu Lv and Hao Shen and Ju H. Park}, doi = {10.1109/TASE.2025.3599400}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19961--19973}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined-time optimized tracking control for QUAVs with extended state observer and global prescribed performance}, volume = {22}, year = {2025}, }
• (2025). Multi-kernel correntropy smoother for 6D foot motion tracking with inertial sensors. TASE, 22, 19945--19960. (www)

  @article{ , abstract = {Accurate foot orientation and trajectory estimation are pivotal for advanced gait analysis, yet achieving this with inertial measurement units (IMUs) remains challenging due to their susceptibility to external acceleration, magnetic disturbances, and unbounded position errors. To address these limitations, we propose a multi-kernel correntropy smoother, which effectively mitigates unknown disturbances and enhances orientation accuracy. Furthermore, while the conventional zero-velocity update (ZUPT) method has been widely adopted in IMUs, the impact of incorporating position constraints has been largely overlooked. This paper demonstrates that by incorporating a single loop closure, the maximum positioning error can be reduced by up to 75%, with further reductions achievable through multiple position constraints. Comprehensive theoretical analysis and extensive experimental validation confirm the superior performance of the proposed methods. Note to Practitioners—This paper aims to enhance 6D foot motion tracking accuracy for foot-mounted IMUs. A hierarchical orientation and position estimation scheme are deployed. Specifically, a robust Kalman smoother for orientation, leveraging multi-kernel correntropy, is designed to mitigate the effects of free acceleration and indoor magnetic disturbances. Additionally, a special forward-backward Kalman smoother for position estimation is designed to accommodate zero velocity and loop closure constraints. Our results underscore the importance of disturbance rejection in orientation estimation and the velocity and position constraints in trajectory estimation, offering insights that may inspire broader applications of IMUs.}, archive = {J_TASE}, author = {Shilei Li and Dawei Shi and Yunjiang Lou and Chenglong Fu and Lisheng Kuang and Ling Shi}, doi = {10.1109/TASE.2025.3599065}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19945--19960}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-kernel correntropy smoother for 6D foot motion tracking with inertial sensors}, volume = {22}, year = {2025}, }
• (2025). Model free adaptive predictive iterative learning cooperative control of multiple high-speed trains operation with disturbances. TASE, 22, 19929--19944. (www)

  @article{ , abstract = {For cooperative control of multiple high-speed trains (MHSTs), a new model free adaptive predictive iterative learning cooperative control (MFAPILCC) method is proposed in this paper. Firstly, the unknown nonlinearity of the MHSTs is handled with the full form dynamic linearization (FFDL) technology along the iteration domain. Then with the predictive mechanism, the unknown bounded time-iteration-varying external disturbances are estimated and predicted, and also be incorporated into the controller design, which achieves better disturbance rejection and system stability. Specially, under the proposed MFAPILCC method, apart from the convergence of tracking error of system could be guranteed, the modeling error of system which practically makes a difference in the cooperative tracking performance of MHSTs also could be guranteed to converge to zero. The convergence of the modeling error and speed tracking error of the system under the constructed method could be guranteed through rigorous theoretical derivation, while the distance between adjacent trains is always maintained within a safety range. Finally, a simulation on the Chinese Railway High-Speed Train (CRH-3) operation control system is supplied to validate the efficacy and superiority of the proposed MFAPILCC method for cooperative control of MHSTs. It is obvious that the theoretical justification and simulation effects either reflect efficacy of the proposed MFAPILCC method. Note to Practitioners—In engineering practice, high-speed train (HST) operates in a repeatable and complex environment, and is susceptible to various external disturbances, such as weather, railway lines, and passenger load, etc. Besides, since the speed of HST gradually gets faster, it always needs to keep a safe distance from adjacent trains to ensure safe operation. Based on the above practical engineering considerations, the motivation of this paper is to design a model free adaptive predictive iterative learning cooperative control (MFAPILCC) method for unknown nonlinear and nonaffine Multiple high-speed trains (MHSTs) under time-iteration-varying external disturbances. The historical operation data of MHSTs is used to describe the complex dynamic mapping relationship between input and output of the train and the total data usage can be determined by engineering applicators. Then based on this description, a corresponding predictive iterative learning control method is constructed. Particularly, the proposed control method constructs a predictive mechanism for prediction and compensation of time-iteration-varying external disturbances with the aim of ensuring safe and punctual operation of trains in complex operation environment. Furthermore, the operation information of adjacent trains is introduced into controller design that assures adjacent trains always keep a safety distance from each other. The convergence of the proposed MFAPILCC method is theoretically analyzed by rigorous derivation. Simulation effects on the CRH-3 high-speed trains operation system further indicate the validity of the proposed method. Actually, the proposed control method can be applied to cooperative control tasks for not only MHSTs but also other practical systems since only I/O data information of the system is required for the design of controller.}, archive = {J_TASE}, author = {Qiongxia Yu and Shuaishuai Wu and Xuhui Bu and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3593561}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19929--19944}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model free adaptive predictive iterative learning cooperative control of multiple high-speed trains operation with disturbances}, volume = {22}, year = {2025}, }
• (2025). Dynamic distributed coordination schemes for multi-mobile robot systems under collision avoidance constraints. TASE, 22, 19912--19928. (www)

  @article{ , abstract = {Guaranteeing collision avoidance is of paramount importance in view of accomplishing missions where many agents are involved to share the same space. The main objective of this work is to expand the Turn-Based Command Governor approach in (Tedesco et al., 2019) by taking non-convex Collision Avoidance constraints into account when performing Plug-and-Play (PnP) operations among agents operating in a 2D environment. To deal with such a scenario, formal conditions that guarantee collision-free Plug-and-Play (PnP) operations are given. These conditions are based on the concept of safe areas, which define regions where agents can safely perform PnP operations without the risk of collision. The effectiveness of the proposed strategy is illustrated through various examples, highlighting its potential for mission accomplishment involving multiple agents. Note to Practitioners—This paper presents a novel distributed control strategy designed to coordinate multiple mobile agents safely and efficiently. By incorporating advanced control techniques such as Model Predictive Control and Command Governors, the proposed approach addresses the challenges of collision avoidance and dynamic formation changes. The methodology offers several potential benefits for practical applications, including enhanced safety, increased flexibility through Plug-and-Play capabilities, and reduced computational demands compared to centralized alternatives. Although implementation may require specialized expertise in control engineering, the potential impact of this research on robotics, autonomous vehicles, and other multi-agent systems is significant. This work provides a foundation for future research and development in distributed control and coordination.}, archive = {J_TASE}, author = {Alessandro Casavola and Vincenzo D'Angelo and Ayman El Qemmah and Gianfranco Gagliardi and Francesco Tedesco and Franco Angelo Torchiaro}, doi = {10.1109/TASE.2025.3599556}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19912--19928}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic distributed coordination schemes for multi-mobile robot systems under collision avoidance constraints}, volume = {22}, year = {2025}, }
• (2025). High-dynamic speed control for high-speed turbo molecular pump under impact load disturbances. TASE, 22, 19901--19911. (www)

  @article{ , abstract = {The magnetic levitation turbo molecular pump (MLTMP) is driven by a high-speed permanent magnet synchronous motor to create a high-vacuum environment for semiconductor manufacturing and other industries. When gas pressure changes abruptly, the MLTMP experiences significant speed fluctuations, which would degrade industrial productivity and even destabilize sensorless control. To address this issue, this paper proposes a high-dynamic speed control strategy for high-speed MLTMP under impact load disturbances. First, a speed control method incorporating a load torque observer and an active disturbance rejection controller (ADRC) is designed. Second, based on the adaptive law developed in conjunction with the ADRC process, an adaptive incremental compensation method is proposed. Finally, the proposed high-dynamic control strategy is performed on a high-speed MLTMP. Experimental results show that the strategy achieves significant attenuation of speed fluctuations, shortens the speed convergence time, and improves the dynamic performance of the MLTMP. Note to Practitioners—The motivation of this study is to address the speed control problem of high-speed magnetic levitation turbo molecular pumps (MLTMP) used in semiconductor manufacturing. During processes such as chip bonding, sudden pressure surges in vacuum chambers lead to rapid increases in MLTMP load torque, degrading pumping speed and compression ratio, and destabilizing sensorless control systems. Conventional speed control methods often rely on an accurate mathematical model of the motor, focus on steady-state performance, and have limited dynamic response capability. The proposed high-dynamic speed control strategy combines a Luenberger load torque observer with an active disturbance rejection controller (ADRC), along with an adaptive incremental compensation (AIC) method. It effectively suppresses speed fluctuations, shortens convergence time, and lowers instantaneous motor power to avoid saturation risks. This strategy replaces traditional PI control and can be implemented in real time via digital signal processor, making it suitable for high-speed, complex load scenarios. This strategy is validated on an experimental platform. Future extensions may include dynamic control of other high-speed rotating equipment such as turbo compressors and precision centrifuges.}, archive = {J_TASE}, author = {Shiqiang Zheng and Gengkui Wei and Baodong Chen and Haifeng Zhang}, doi = {10.1109/TASE.2025.3599216}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19901--19911}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-dynamic speed control for high-speed turbo molecular pump under impact load disturbances}, volume = {22}, year = {2025}, }
• (2025). Robust load frequency control for multi-area power systems: A delay-induced-source dependent approach. TASE, 22, 19889--19900. (www)

  @article{ , abstract = {Modern power systems maintain frequency stability through load frequency control (LFC), which is affected by delays of control signals and the volatility of renewable energy sources. This paper investigates the robust problem of LFC for wind power systems with multi-source-induced delays, considering the significant variations in delay magnitudes across different time intervals due to diverse inducing sources. Firstly, the LFC is modeled as a closed-loop system with two switching modes by dividing the delay into two parts with different magnitudes based on the inducing sources. Then, the controller design criteria for a multi-source-induced delay-dependent LFC system are proposed through the switching system theory. Finally, case studies conducted on two-area and three-area LFC system validate that the method proposed in this paper can enhance the robust performance of the system compared to general methods, and it is also capable of handling certain multi-source-induced delay phenomena that are beyond the capabilities of general methods during the operation of LFC systems. Note to Practitioners—Frequency stability in multi-area power systems is facing growing challenges due to delays caused by various sources. Existing methods for delayed power systems often overlook these diverse delay sources, treating them as uniform, which may reduce system robustness. This paper proposes a robust load frequency control (LFC) scheme based on a delay-induced-source dependent approach, which aims to ensure the exponential stability of multi-area LFC systems. By modeling multi-source-induced delays using switching modes and designing controllers through switching system theory, this method effectively addresses the complexity of multi-source-induced delays. Furthermore, it enables the design of controllers tailored to the type of delay-inducing source, enhancing system stability under different delay types. Practitioners can apply this method to enhance the reliability of LFC systems in power systems experiencing diverse delay scenarios. It is especially effective in situations where existing methods struggle to handle complex delays, providing a practical solution for maintaining frequency stability in modern networked power systems.}, archive = {J_TASE}, author = {Zhe-Li Yuan and Chuan-Ke Zhang and Ju H. Park and Yong He}, doi = {10.1109/TASE.2025.3598838}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19889--19900}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust load frequency control for multi-area power systems: A delay-induced-source dependent approach}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based predefined-time tracking control for input-saturated nonlinear systems with performance guarantees. TASE, 22, 19876--19888. (www)

  @article{ , abstract = {This paper addresses the predefined-time tracking control problem for nonlinear systems subject to predetermined performance metrics and input saturation. With the assistance of the reinforcement learning method, the optimization control objectives of nonlinear systems are achieved. Therein, the execution costs and control behaviors are trained effectively under the actor-critic neural network framework. In addition, the unknown nonlinearities and external disturbances are delicately compensated through the neural approximation rule and an adaptive updated law, respectively. To meet the performance requirements, a performance function is designated to acquire the satisfaction of performance metrics. By incorporating an auxiliary control signal, the saturation constraint on control input is satisfied. In particular, under the action of the developed scheme, the boundness of each system state can be ensured in a predefined time interval. Finally, the effectiveness and superiority of the developed scheme are demonstrated via simulation results. Note to Practitioners—This paper aims to achieve the optimized predefined-time tracking control for nonlinear systems with input saturation and performance guarantees. With the improvement of automation, predefined time convergence becomes one significant performance index in many time-critical practical projects, such as teleoperation systems, missile guidance systems, and so on. As a matter of fact, since actuator saturation and external disturbances exist extensively in practice, they constantly impair the operation performance of engineering systems. To deal with the aforementioned challenges, a reinforcement learning-based predefined-time anti-saturation controller is designed in this paper, which ensures that the system converges within a user-specific time while minimizing the cost. Preliminary simulation comparisons demonstrate that this developed scheme applies to practical servo systems with satisfactory control results.}, archive = {J_TASE}, author = {Wei Zhao and Jing Wang and Shaoyu Lv and Lei Su and Hao Shen}, doi = {10.1109/TASE.2025.3598545}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19876--19888}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based predefined-time tracking control for input-saturated nonlinear systems with performance guarantees}, volume = {22}, year = {2025}, }
• (2025). Data-based adaptive asymptotic tracking control for high-speed train: A feedback linearization approach. TASE, 22, 19860--19875. (www)

  @article{ , abstract = {This paper presents a novel data-based adaptive control strategy based on feedback linearization to address the asymptotic tracking control problem of position and speed for underactuated high-speed train (HST). The proposed strategy accounts for basic resistances, in-train forces, multiple unknown disturbances, and unknown system parameters. The key contributions of this paper are threefold. First, the proposed strategy eliminates the reliance on exact model parameters by incorporating adaptive mechanism, which is a significant advancement over traditional feedback linearization method. Second, by combining Lyapunov stability theory with a novel output redefinition approach, the stability of the zero dynamics system for underdriven HST is rigorously demonstrated. Third, an improved equivalent control law is introduced, which not only suppresses unknown disturbances automatically but also mitigates severe chattering phenomenon. Simulation on a HST with 2 motor cars and 6 trailer cars is provided for verifying the theoretical results. Simulation results show that the proposed strategy achieves asymptotic tracking of the locomotive to the desired position and speed trajectories as well as ensures the uniformly ultimately bounded stability of the internal dynamics of all trailers. Note to Practitioners—The underactuated high-speed train (HST) system, as a practical engineering system, is characterized by multi-input multi-output, multivariable coupling, and nonlinearity. During the operation of an underactuated HST, the system is inevitably affected by basic resistances, in-train forces, multiple unknown disturbances, and time-varying system parameters. This paper aims to achieve high-precision tracking of the position and speed of underactuated HST under accounting for these factors. The proposed data-based control scheme enhances practicality by applying the feedback linearization combined with adaptive mechanism to deal with the nonlinearity of the underactuated HST. To address the challenges posed by the complex operating environments of HST, adaptive control is introduced. Additionally, an adaptive sliding mode control method is employed to enhance the robustness and fast convergence of the proposed control scheme, while effectively mitigating severe chattering. The simulation results show that the HST can achieve high-precision tracking of the desired position and speed curves in the presence of unknown system parameters and disturbances.}, archive = {J_TASE}, author = {Ya-Fei Shi and Hui Yang and Dong Liu and Kunpeng Zhang and Chun-Hua Xie}, doi = {10.1109/TASE.2025.3598314}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19860--19875}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-based adaptive asymptotic tracking control for high-speed train: A feedback linearization approach}, volume = {22}, year = {2025}, }
• (2025). Event-based adaptive fault-tolerant control for nonlinear cyber-physical systems via intermittent available signals. TASE, 22, 19850--19859. (www)

  @article{ , abstract = {This research considers the problem of event-based adaptive fault-tolerant control for nonlinear cyber-physical systems with deception attacks and actuator faults via intermittent available signals. Through the application of fuzzy logic systems, the unknown nonlinear functions of the systems are approximated. Then, a novel state observer is designed that can be driven by actuator faults and intermittent available signals arising from triggered attack-state signals, which can realize directly triggering the states after deception attacks and avoid the problem of virtual controller non-differentiability under the backstepping framework. In order to avoid the complexity explosion problem, the dynamics surface control method is introduced. Meanwhile, the dynamics surface filter signals also realize event-triggered control, and it results in substantial savings in communication resources. To verify the validity of the proposed method, two illustrative examples are presented. Note to Practitioners—Cyber-physical systems are frequently applied in modern industrial processes such as smart grids, industrial Internet of Things and so on. Nevertheless, the open network environment makes system components more vulnerable to attacks, posing a significant security threat to the operation of cyber-physical systems. When attackers deliver deceptive information to the sensors, the system state becomes inaccessible, which is a challenging issue based on the signals available after the attacks. Moreover, when actuators are affected by faults, system performance deteriorates, potentially leading to instability. Consequently, the event-triggered control problem of cyber-physical systems, considering both actuator faults and deceptive attacks, presents challenges for controller design. In order to resolve the issues outlined above, a fault-tolerant state observer based on intermittent available signals is developed. This design employs backstepping recursion and adaptive techniques to achieve stability for system under deception attacks and greatly alleviate communication burden, thereby enhancing the practicality of the proposed control strategy.}, archive = {J_TASE}, author = {Jiawei Ma and Huaguang Zhang and Juan Zhang and Xiyue Guo}, doi = {10.1109/TASE.2025.3598307}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19850--19859}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based adaptive fault-tolerant control for nonlinear cyber-physical systems via intermittent available signals}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered optimal consensus fault-tolerant control for unknown stochastic nonlinear multi-agent systems with time-delays. TASE, 22, 19836--19849. (www)

  @article{ , abstract = {The optimal consensus fault-tolerant control (FTC) problem for unknown time-delays nonlinear stochastic multi-agent systems (MASs) with hybrid actuator faults is investigated via adaptive dynamic programming (ADP) under the dynamic event-triggered control (ETC) mechanism in this paper. In control design, FLSs with delayed states are utilized to design an adaptive state identifier for estimating the unknown delayed dynamics in stochastic MASs. The fault estimator is constructed to detect actuator bias fault in real-time. By utilizing the ADP algorithm with identifier-actor-critic structure, a simplified fuzzy adaptive optimal consensus FTC strategy is developed, which does not require persistent excitation (PE) assumption. Furthermore, a dynamic ETC strategy is designed to further optimize the data transmission efficiency, effectively enhancing the system flexibility and resource utilization. Theoretical analysis indicates that the proposed control scheme can ensure that all signals of the nonlinear stochastic MASs are semi-globally uniformly ultimately bounded (SGUUB) in mean square, while achieving consensus between the follower’s states and the state of the leader. Finally, a numerical simulation validates the efficacy and feasibility of the proposed control strategy. Note to Practitioners—With the intensification of the global energy crisis, energy consumption optimization has become a crucial performance indicator of modern control design. In this paper, the optimal consensus FTC strategy is developed for unknown nonlinear stochastic MASs with time delays and actuator failures, which can achieve control requirements while minimizing control costs. The proposed control strategy can be applied to practical scenarios such as aerospace, formation control, and power systems. Compared with the existing control algorithms, the proposed control strategy avoids complex dynamic modeling and strong assumptions of nonlinear functions, and does not require PE assumption. In addition, a novel dynamic ETC strategy has been developed, which effectively improves resource utilization and data transmission efficiency compared with traditional continuous-time control and general ETC strategies. Consequently, the control strategy designed in this paper exhibits broader applicability and stronger practical value.}, archive = {J_TASE}, author = {Fuling Zheng and Guoping Zhang and Chengbin Liang and Jinrong Wang and Quanxin Zhu}, doi = {10.1109/TASE.2025.3598199}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19836--19849}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered optimal consensus fault-tolerant control for unknown stochastic nonlinear multi-agent systems with time-delays}, volume = {22}, year = {2025}, }
• (2025). Automated assembly of magnetic soft microrobots with chopstick-like two-fingered microhand. TASE, 22, 19824--19835. (www)

  @article{ , abstract = {The development of magnetic soft microrobots has been constrained by the lack of precise control of microactuator’s programmability. To address such an issue, we use a piezo-driven two-fingered microhand to selectively position the magnetic microactuators within a soft scaffold. Each microactuator is sequentially oriented to a desired direction through surface rotation by applying magnetic field–based torques, and is then fixed to the scaffold using ultraviolet (UV) cross-linking. A detailed analysis is conducted on the spatial positioning capability of the microhand’s parallel mechanism and the magnetic programming performance of electromagnetic coils’ orientation control. To overcome the inefficiency and inaccuracy of labor-intensive manual assembly, we propose an automated assembly strategy to create magnetic soft-bodied microrobots following our design. The multiple fabricated microrobot prototypes exhibit programmed 2D and 3D shape transformations and various robotic gaits for surface locomotion. This strategy can enable the rapid fabrication of multimaterial 3D magnetic microrobot designs with potential applications in robotics, biomedical engineering, and environmental governance. Note to Practitioners—The motivation of this work is to address the limitations in current fabrication methods for magnetic soft microrobots, which include 3D printing, heating, mold casting, and chemical synthesis. These approaches fall short when creating microrobots with arbitrary structures, multimaterial compositions, and complex magnetization profiles. Micro-assembly offers a potential solution, yet existing techniques depend on either time-consuming manual assembly or costly equipment lacking sufficient flexibility for 3D spatial manipulations. In this work, a 3-degree-of-freedom, high-precision parallel microhand-based assembly method is proposed. The microhand provides high precision and speed for rapid positioning of assembly modules, while electromagnetic coils enable precise orientation in magnetization programming. Experimental results confirm the effectiveness of this method and demonstrate the robustness and adaptability of the automated fabrication strategy. This approach facilitates the fabrication of magnetic soft microrobots with specific deformation and locomotion capabilities according to design intent.}, archive = {J_TASE}, author = {Yue Zhao and Yan Deng and Ruixi Wang and Dan Liu and Masaru Kojima and Qiang Huang and Tatsuo Arai and Xiaoming Liu}, doi = {10.1109/TASE.2025.3597934}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19824--19835}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated assembly of magnetic soft microrobots with chopstick-like two-fingered microhand}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered optimal control for fuzzy multiagent systems with DoS attacks. TASE, 22, 19814--19823. (www)

  @article{ , abstract = {This article devotes to investigate the fuzzy optimal consensus problem for multiagent systems (MASs) subject to denial-of-service (DoS) attacks by using the dynamic event-triggered (DET) strategy. The interval type-2 (IT2) fuzzy systems are employed to model the MASs, which combines with the DET fuzzy controller to establish a closed-loop system. Moreover, two piecewise functions are designed, and the triggered interval in a DoS period is decomposed into some subsets with a sampling interval. Therefore, the closed-loop system is reconstructed to facilitate stability analysis. The piecewise Lyapunov-Krasovskii functional is designed, and the system stability is deduced with and without DoS attacks, respectively. By considering the relationship of the piecewise Lyapunov functional, all agents are ensured to be exponentially stable, and the closed-loop system satisfies an optimal control performance. Two examples show the feasibility of the proposed theory finding. Note to Practitioners—Cyberattacks lead to control failure or data breach of MASs due to data tampering and communication interruption. Besides, the controller based on DET mechanism can save limited network resources. Hence, this research focuses on MASs with parameter uncertainties under DoS attacks in interval type-2 fuzzy form, and a DET fuzzy control scheme is proposed to guarantee system security. The Lyapunov-Krasovskii functional with sampling information is constructed to promote stability verification, and slack matrices are also introduced. This study aims to reveal the design principles of DET fuzzy secure control scheme for MASs under DoS attacks. The conducted stability analysis and DET strategy adapt to the background of engineering application.}, archive = {J_TASE}, author = {Zhenbin Du and Yonggui Kao and Yang Liu and Zhaojing Wu}, doi = {10.1109/TASE.2025.3597879}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19814--19823}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered optimal control for fuzzy multiagent systems with DoS attacks}, volume = {22}, year = {2025}, }
• (2025). AllTact fin ray: A compliant robot gripper with omni-directional tactile sensing. TASE, 22, 19801--19813. (www)

  @article{ , abstract = {Tactile sensing plays a crucial role in robot grasping and manipulation by providing essential contact information between the robot and the environment. In this paper, we present AllTact Fin Ray, a novel compliant gripper design with omni-directional and local tactile sensing capabilities. The finger body is unibody-casted using transparent elastic silicone, and a camera positioned at the base of the finger captures the deformation of the whole body and the contact face. Due to the global deformation of the adaptive structure, existing vision-based tactile sensing approaches that assume constant illumination are no longer applicable. To address this, we propose a novel sensing method where the global deformation is first reconstructed from the image using edge features and spatial constraints. Then, detailed contact geometry is computed from the brightness difference against a dynamically retrieved reference image. Extensive experiments validate the effectiveness of our proposed gripper design and sensing method in contact detection, force estimation, object grasping, and precise manipulation. Note to Practitioners—Fin Ray fingers have been widely adopted for robot grasping and manipulation due to their adaptive nature. However, existing Fin Ray finger designs lack tactile sensing capabilities, or can only perceive contacts at certain regions. The motivation of this work is to enable the Fin Ray gripper to not only detect contacts in all directions but also measure the detailed contact geometry of the grasped object. The omni-directional contact sensing ability is crucial for manipulation in cluttered and dynamic scenarios. And the contact geometry measurement is critical for in-hand pose estimation and precise manipulation. The proposed finger design is only composed of a single transparent elastic silicone cast, a camera and white LED, making the manufacturing process simple and time-efficient. The design files and sensing algorithms will be open-sourced to foster further development. We validate the effectiveness of this sensor design and sensing method across various applications. It can locate contact positions with an error less than 1mm, accurately estimate the force magnitude on the contact surface, and detect contact forces from different directions with a threshold below 2 N. In future research, we will explore advanced manipulation applications including deformable object and bi-manual manipulation and improve the finger’s durability.}, archive = {J_TASE}, author = {Siwei Liang and Yixuan Guan and Jing Xu and Hongyu Qian and Xiangjun Zhang and Dan Wu and Wenbo Ding and Rui Chen}, doi = {10.1109/TASE.2025.3597842}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19801--19813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AllTact fin ray: A compliant robot gripper with omni-directional tactile sensing}, volume = {22}, year = {2025}, }
• (2025). Controllable mixture-of-experts for multivariate soft sensors. TASE, 22, 19789--19800. (www)

  @article{ , abstract = {Multivariate soft sensors are critical in industrial manufacturing for providing precise estimations of multiple quality variables and ensuring data reliability and completeness. Existing methods predominantly focus on parameter-sharing architectures while overlooking two fundamental issues: 1) catastrophic interference, where sharing parameters across all tasks leads to performance degradation; 2) uncontrollable optimization, where the optimizer lacks controllability over task priorities, misleading specific tasks converge to unexpected values. To handle these issues and enhance multivariate modeling, we propose a controllable Mixture-of-Experts (ControlMoE) framework for industrial soft sensors, consisting of a Mixture-of-Sequential-Experts (MoSE) module and a Proportional-Integral-Derivative Calibrating (PIDC) module. The MoSE module integrates sequential expert networks with task-specific gating networks, enabling parameter sharing while preserving task distinctions to mitigate catastrophic interference. The PIDC module employs an innovative PID controller to dynamically calibrate task learning weights, ensuring expected outcomes through feedback control and enhancing optimization controllability. Evaluations on two industrial datasets from Chinese nuclear monitoring stations demonstrate that ControlMoE achieves superior accuracy and controllability in multivariate soft sensor modeling. Note to Practitioners—This paper tackles the challenges in applying multivariate soft sensors within industrial manufacturing processes, particularly in environments such as nuclear monitoring where precision and control over multiple quality variables are critical. Traditional methods often suffer from performance degradation due to parameter sharing across all tasks-known as catastrophic interference–and lack control in optimizing multiple tasks simultaneously, leading to suboptimal outcomes. Our proposed framework, ControlMoE, introduces a novel architecture that combines sequential expert networks with a dynamic weight adjustment mechanism using PID controllers. This enables precise control over task-specific optimizations, ensuring that the desired importance and sequence of tasks are maintained, thereby enhancing both accuracy and operational control. We believe this framework could be widely applicable in various industrial settings where multivariate monitoring and control are necessary, improving not only performance but also the reliability of the data produced for critical decision-making.}, archive = {J_TASE}, author = {Licheng Pan and Hao Wang and Zhichao Chen and Yuxin Huang and Yunlong Niu and Zhaoran Liu and Qunshan He and Xinggao Liu}, doi = {10.1109/TASE.2025.3597838}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19789--19800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Controllable mixture-of-experts for multivariate soft sensors}, volume = {22}, year = {2025}, }
• (2025). A distributed self-triggered control scheme for multi-unmanned aerial vehicle containment control under markov switching topologies and channel fading. TASE, 22, 19775--19788. (www)

  @article{ , abstract = {This article addresses the containment control problem in multi-unmanned aerial vehicles (multi-UAVs) over communication networks subject to Markov switching topologies and channel fading. Firstly, a decoupling method is skillfully proposed to divide the closed-loop global containment error system into multiple local dynamics, simplifying the design of the distributed self-triggered scheme. Then, a mode-dependent distributed containment controller is designed, leveraging fading depths to mitigate the impact of channel fading. During the containment analysis, a self-triggered auxiliary condition is intelligently derived, forming the basis for the subsequent design of the self-triggered mechanism. Additionally, a novel distributed self-triggered mechanism is designed to significantly reduce the frequency of information exchange in the system while eliminating Zeno behavior. Finally, the effectiveness of the proposed distributed self-triggered containment control is validated through a simulation example. Note to Practitioners—In a multi-UAV system, stable operation primarily depends on information transmission between UAVs. Network-induced constraints will severely undermine system stability and jeopardize coordinated control. Accordingly, studying effective solutions to address network-induced constraints in real-world systems is of practical importance. This paper investigates the containment control problem in a multi-UAV system affected by channel fading, random communication topologies, and self-triggered transmission. By incorporating fading depth in the communication channel, the proposed controller adapts to varying channel conditions, ensuring the UAVs maintain coordination even under poor communication conditions. Additionally, given that the communication topology between UAVs may change due to environmental factors in real-world applications, Markov random switching topologies can be used to model this phenomenon. Limited communication resources may lead to transmission delay or even network congestion due to the large amount of information exchanged between UAVs. Nevertheless, blindly increasing network bandwidth not only imposes high demands on sensors but also raises experimental costs. Therefore, reducing the frequency of information transmission is a commonly used approach. Unlike traditional event-triggered scheme, this paper proposes a distributed self-triggered scheme for addressing containment control problems in multi-UAV systems, eliminating the need for continuous event monitoring. From a practical application perspective, the distributed self-triggered approach does not require additional sensors for event monitoring, further reducing costs. Building on this work, future research should also address the formation control of multi-UAV systems, where multiple UAVs cooperate to form a specific formation through information exchange.}, archive = {J_TASE}, author = {Zhiru Cao and Shiyi Wang and Xiaohua Ge and Chen Peng and Haijuan Zhao}, doi = {10.1109/TASE.2025.3597831}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19775--19788}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A distributed self-triggered control scheme for multi-unmanned aerial vehicle containment control under markov switching topologies and channel fading}, volume = {22}, year = {2025}, }
• (2025). A 6-DoF dynamic model for falcon-like flapping-wing aircraft: Virtual environment design and verification. TASE, 22, 19763--19774. (www)

  @article{ , abstract = {Unsteady aerodynamics because of wing motion and underactuated caused by bionic driving mode are among the most notable challenges to accurate dynamic modeling and flight control for flapping-wing aircraft (FWAs). In this paper, we establish a comprehensive 6-degree-of-freedom (DoF) dynamic model to describe the multi-mode motion of a falcon-like FWA. The unsteady vortex lattice method (UVLM) is used to compute the time-varying aerodynamic forces on the main wing, accounting for the effects of flexible deformations and lateral dynamics. Based on a series of wind tunnel experiments, a data-driven model for the V-tail is developed using an artificial neural network (ANN) to capture the relationship between ruddervators’ deflections and aerodynamic forces. A simulation environment, constructed on the foundation of aerodynamic modeling and dynamic analysis, is validated with both open-loop and closed-loop flight data. The virtual system allows for rapid verification and optimization of the model and control parameters, which can enhance the flexibility and reliability of the development process of FWAs. Note to Practitioners—This work addresses the challenges of modeling and controlling FWAs, which are known for their exceptional agility but also pose significant difficulties due to unsteady aerodynamics and underactuated mechanisms. We develop a 6-DoF dynamic model and simulation environment for a falcon-like FWA, capturing multi-mode motion. Unlike dynamic models constrained to small perturbations around stable operating points, our approach provides a more comprehensive description of FWA dynamics across diverse flight modes. Validated with real flight data, this virtual environment is proven to be reliable and capable of supporting advanced control methods such as reinforcement learning and model predictive control. It offers a platform for optimizing FWA design and control, enhancing flexibility and reliability in practical applications while paving the way for further improvements in complex flight scenarios.}, archive = {J_TASE}, author = {Xuena Zhao and Liang Zhang and Zhijie Liu and Wei He}, doi = {10.1109/TASE.2025.3597138}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19763--19774}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A 6-DoF dynamic model for falcon-like flapping-wing aircraft: Virtual environment design and verification}, volume = {22}, year = {2025}, }
• (2025). An extended false data injection attack via deep reinforcement learning: Attack model and countermeasures in cyber-physical power systems. TASE, 22, 19750--19762. (www)

  @article{ , abstract = {False data injection attacks are commonly used to evade the bad data detector in cyber-physical power systems. This paper proposes an extended attack strategy and a deep reinforcement learning-based detection method. Traditional false data injection attacks aim to remain stealthy and avoid detection by conventional detection mechanisms. An extended load attack is introduced to increase the potential for damage. Directly adding an extended component directly to the measurement makes it easily detectable by bad data detector. Accordingly, the extended attack integrates the added component into the state variables to improve stealth. An optimization model for the extended components of the proposed attack is developed, along with a homologous matrix. Additionally, an online attack detection scheme is formulated as a partially observable Markov decision process problem. A deep reinforcement learning-based detection framework is proposed, featuring a compound reward designed to minimize false alarms and time delays. The proposed online detector extracts state features under varying operating conditions and generates a policy to determine whether the power grid is under attack. An extended Euclidean distance indicator and an adaptive weight matrix are also proposed in the dynamic state estimation to improve estimation or detection accuracy. Numerical experiments validate the effectiveness and robustness of the proposed deep reinforcement learning-based detection scheme in power systems. Note to Practitioners—This paper is motivated by the lack of research on modeling the destructive capabilities of cyber-attacks and the inaccuracy of anomaly detection methods for data integrity attacks. Existing approaches to modeling false data injection attacks primarily focus on the hidden features bypassing detection of bad data detection in cyber-physical power systems. This paper proposes a novel false data injection modeling approach, triggered by the spinning reserves of power grids. The proposed false data injection strategy mathematically characterizes extended stealth attack mechanisms and introduces an extended load attack to explore greater destructive potential. Considering the uncertain environments in perspective attackers and defenders, this study formulates the attack detection problem as a partially observable Markov decision process. This study then characterizes how such metrics of detectors can be efficiently computed, this can allow a defender to automatically learn or generate a detection threshold policy using the deep reinforcement learning, and distinguishes real false data injection attacks from system noises. To improve the detection performance, a novel indicator and an adaptive weight matrix are proposed to enhance learning efficiency of detectors. Numerical simulations suggest that the proposed detection scheme is feasible to both traditional and extended false data injection attacks, though it does not yet account for incomplete measurements. Future work will focus on designing attack detection strategies under conditions of incomplete knowledge of network topology and measurements.}, archive = {J_TASE}, author = {Xiaohong Ran and Lei Ma}, doi = {10.1109/TASE.2025.3596563}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19750--19762}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An extended false data injection attack via deep reinforcement learning: Attack model and countermeasures in cyber-physical power systems}, volume = {22}, year = {2025}, }
• (2025). Obstacle avoidance control for unmanned tracked vehicle formation with communication constraints. TASE, 22, 19735--19749. (www)

  @article{ , abstract = {In this paper, aiming at the ground battlefield environment with communication constraints and many unknown obstacles, the autonomous obstacle avoidance control problem of unmanned tracked vehicle formation is studied. Innovative solutions are proposed to address the challenges of poor dynamic adaptability of obstacle avoidance, high communication dependence, and easy interruption of communication links that exist in traditional methods in the dense scene of unknown obstacles. Firstly, the adaptive formation method is designed and the leadership dynamic transfer mechanism is constructed to enhance the system’s responsiveness to unexpected obstacles by reconfiguring the formation topology in real time. Secondly, a state space modelling method based on perceptual field is proposed, which effectively reduces the communication demand within the formation and improves the convergence speed of the algorithm through the encoding of environmental features and fusion of local information. Finally, a communication constraint model is constructed to evaluate the communication link state in combination with the channel quality, and dynamic rewards are generated in real time to motivate the formation to maintain smooth communication. Based on the above innovations, an adaptive formation communication constraint perceptual field-based multi-agent proximal policy optimization is proposed, and the effectiveness of this method is evaluated through simulation experiments. Experimental results show that the proposed method has strong generalization ability and robustness, and can complete the autonomous obstacle avoidance control task of unmanned tracked vehicle formation with communication constraints. The research results provide theoretical support and technical realization paths for unmanned system formations to collaborate in complex environments. Note to Practitioners—The purpose of this paper is to solve the problem of autonomous obstacle avoidance of unmanned tracked vehicle formation with communication constraints, but it can also be applied to other types of unmanned system formation. The existing unmanned system formation obstacle avoidance control methods are usually rule-based and assume that the communication state is good, which limits its wider application. In this paper, a new method of adaptive formation obstacle avoidance control is proposed. The decision of formation is made only according to the finite state in the perceptual field, which greatly reduces the dependence on communication. Furthermore, the motion mode of unmanned tracked vehicle is described mathematically, and the control scheme is designed based on intelligent algorithm. The new control method fully considers the communication constraints and the robustness of the formation. Preliminary simulation experiments show that this method is feasible, but it has not been tested in physical systems. In future studies, we will conduct physical experiments and verify the effectiveness of the proposed method in more types of control tasks.}, archive = {J_TASE}, author = {Jiangyi Yao and Xiongwei Li and Yang Zhang and Jingyu Ji and Guoquan Ren and Zhigang Ren}, doi = {10.1109/TASE.2025.3598140}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19735--19749}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Obstacle avoidance control for unmanned tracked vehicle formation with communication constraints}, volume = {22}, year = {2025}, }
• (2025). Observer-based dynamic event-triggered resilient control for heterogeneous multi-agent systems under DoS attacks. TASE, 22, 19723--19734. (www)

  @article{ , abstract = {This paper studies the issue of dynamic event-triggered (ET) resilient control for heterogeneous multi-agent systems (MASs) under DoS attacks. Most of the existing work only considers ideal linear models and undirected graph communications. However, in practice, both disturbance and noise exist in the system model, and directed graph communication is more common. To solve this issue, a fully distributed dynamic ET control strategy with a prediction-based dynamic compensation algorithm is designed to deal with the difficulty of communication jamming. By using the Lyapunov method, it is proved that bounded consensus can be achievable in heterogeneous MASs under DoS attacks, and Zeno behavior is excluded. Furthermore, the tolerable attack intensity is quantified by attack frequency and duration. Finally, a numerical simulation is conducted to validate the efficacy of the proposed method.Note to Practitioners—In real-world scenarios, numerous complex tasks necessitate collaborative endeavors among heterogeneous intelligent agents, such as robots swarms, multiple vehicles and smart grids. Due to physical or geographical limitations, direct communication between all agents and a leader is not always feasible, thereby restricting individual agents to indirectly accessing the leader’s information. Furthermore, the practical constraints on network resources, coupled with the susceptibility to malicious DoS attacks, impede the completion of tasks by multiple agents. This paper presents a resilient dynamic ET control strategy for the consensus task of heterogeneous agents, aiming to improve system resilience while conserving network resources.}, archive = {J_TASE}, author = {Yuan-Cheng Sun and Zheng Zhang and Lina Yao and An-Yang Lu}, doi = {10.1109/TASE.2025.3598062}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19723--19734}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based dynamic event-triggered resilient control for heterogeneous multi-agent systems under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). A hierarchical patch feature distribution network for industrial multiscale defect detection. TASE, 22, 19709--19722. (www)

  @article{ , abstract = {Visual inspection of surface defects in industrial products is crucial for quality control but remains challenging due to unpredictable and multiscale defects. Unsupervised anomaly detection methods based on feature distribution have rapidly developed, requiring only normal samples and no additional manual labeling. However, these methods still suffer from feature bias and image misalignment, leading to overdetection of multiscale defects, making them impractical for deployment in factory settings. To address these issues, this study proposes a hierarchical patch feature distribution modeling (HPDM) network. Specifically, the new patch feature alignment learning (PFAL) module mitigates image misalignment by enhancing feature dimension similarity via the PFAL loss function. Additionally, the proposed discrete feature memory (DFM) module employs linear mapping to increase the density of the normal feature distribution, addressing feature bias from pretrained networks. Furthermore, the pyramid-shaped Gaussian distribution (PGD) module fits multivariate Gaussian models at different feature scales for defect detection across various scales. On the MVTec2D, MVTec3D, and VISION datasets, the HPDM network achieves pixel-level AUROC scores of 98.83%, 98.30%, and 71.20%, respectively, surpassing those of the existing methods and demonstrating state-of-the-art anomaly detection performance. The experimental results on a real printed circuit board assembly (PCBA) circuit dataset demonstrate a detection speed of 30 frames per second (FPS), meeting the requirements for factory online detection. Note to Practitioners—Previous unsupervised anomaly detection methods have struggled to detect both large and small surface defects simultaneously. However, the proposed multiscale patch feature modeling network effectively identifies various multiscale defects on product surfaces, including PCBA, thin-film transistor liquid crystal displays, and tiles. Additionally, HPDM requires only a small number of normal samples to learn a robust network, which is crucial for industrial applications where identifying and labeling samples is challenging. Furthermore, by leveraging model pruning and quantization techniques, HPDM can be applied to online visual inspection.}, archive = {J_TASE}, author = {Zhou Zheng and Hua Yang and Zhouping Yin}, doi = {10.1109/TASE.2025.3598017}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19709--19722}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical patch feature distribution network for industrial multiscale defect detection}, volume = {22}, year = {2025}, }
• (2025). SIGMA: An agent-based modeling UAV swarm simulator for swarm intelligence algorithms. TASE, 22, 19694--19708. (www)

  @article{ , abstract = {Swarm intelligence for uncrewed aerial vehicles (UAVs) significantly improves the success rate of executing intricate tasks using “distributed platforms and aggregated effects”. However, the high experimental costs and safety risks constrain its development. This paper introduces SIGMA (Swarm Intelligence Generic simulator for Multi-UAVs), a high-fidelity distributed UAV swarm simulator for swarm intelligence algorithms. As an agent-based modeling simulator (ABMS), SIGMA has three key innovations: First, an automatic model tuning method improves aircraft dynamics fidelity. Second, a bidirectional discrete-event simulation (BiDES) architecture resolves the time alignment challenges in distributed systems. Third, a multi-agent learning toolbox ensures algorithm compatibility via an episodic training structure and a memory replay mechanism. In the verification part, the fidelity and scalability of the simulator are verified by quantitative simulations and experiments, and several successful applications demonstrate the practicality of the proposed simulator. Notice to Practitioners—The motivation for this paper stems from the need to develop a scalable and high-fidelity simulator for practical applications of swarm intelligence algorithms. Simulators developed based on game engines are widely used in swarm robotics due to their realistic 3D environments. However, as the number of nodes increases, the real-time performance and scalability of the simulators will decrease significantly. To achieve real-time simulation of large-scale swarms and improve swarm fidelity, SIGMA is proposed. It uses ABMS technology to achieve better performance and can effectively fit the learning tasks of UAV swarms.}, archive = {J_TASE}, author = {Sheng Zhang and Juan Li and Chang Liu and Lei Fu and Zehao Bai and Jie Li}, doi = {10.1109/TASE.2025.3597077}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19694--19708}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SIGMA: An agent-based modeling UAV swarm simulator for swarm intelligence algorithms}, volume = {22}, year = {2025}, }
• (2025). Security control for fuzzy singularly perturbation systems under DoS attacks. TASE, 22, 19685--19693. (www)

  @article{ , abstract = {This study investigates the security control issue for fuzzy singularly perturbed systems (FSPSs) subjected to Denial-of-Service (DoS) attacks. To accurately characterize the dynamics of DoS attacks, an innovative duration-dependent switching rule is proposed, guided by a joint probability distribution function that incorporates both the current attack mode and its dwell time. Unlike traditional Markov/semi-Markov chains, this approach utilizes duration-conditioned sojourn probabilities to eliminate the necessity for explicit transition probabilities estimation, significantly simplifying implementation and reducing computational complexity in multi-mode attack scenarios. Additionally, an attack compensation mechanism explicitly designed to maintain system stability is developed. Furthermore, a non-monotonic Lyapunov function aligned with the proposed switching logic is introduced, resulting in less conservative stability criteria that guarantee the mean-square stability of closed-loop FSPSs. The effectiveness and resilience of the proposed methodology are substantiated through a numerical simulation and a practical case study, highlighting its superior performance in mitigating the adverse effects of DoS attacks compared to existing methodologies. Note to Practitioners—With the continuous advancement in network-based control technologies, singularly perturbed systems have gained significant attention due to their inherent capability to handle dynamics that evolve across multiple time scales. Practical engineering systems frequently encounter uncertainties, including external disturbances, equipment failures, and cyber threats such as Denial-of-Service (DoS) attacks, which can profoundly impact system states and operational stability. Particularly, DoS attacks introduce substantial risks depending on their duration and operational mode. Recognizing these practical concerns, this study proposes an advanced joint probability distribution model that explicitly incorporates both the current attack mode and its duration. By adopting a duration-dependent sojourn probability based on general distributions, the model closely aligns with real-world operational conditions. Additionally, industrial control systems experiencing DoS attacks often face critical information loss, potentially leading to severe instabilities, such as oscillations or divergence. To address these risks, this paper presents a sophisticated gain compensation mechanism designed explicitly for maintaining robust system performance under attack-induced disruptions. This approach ensures that reliable control gains remain available even when essential data is compromised, significantly enhancing system resilience and stability under challenging attack conditions.}, archive = {J_TASE}, author = {Xiaoke Tang and Jun Cheng and Bin Zhang and Leszek Rutkowski and Wentao Huang}, doi = {10.1109/TASE.2025.3596508}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19685--19693}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Security control for fuzzy singularly perturbation systems under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Simultaneous path and motion planning approaches for cooperative cable-driven transportation with mobile robots: A comparative study. TASE, 22, 19667--19684. (www)

  @article{ , abstract = {Cooperative cable-driven transportation with mobile robots connects mobile robots to the payload with cables, rather than attaching mobile robots to the payload. A cooperative cable-driven transportation system can move over low obstacles and reduce the mutual influence of mobile robots through flexible connections. In a large-scale environment, a cooperative cable-driven transportation system could be regarded as a point and sophisticated path planning approaches are available. However, in a small-scale environment with space constraints and obstacles, a cooperative cable-driven transportation system cannot be regarded as a point and the simultaneous path and motion planning of the system is a changeling problem. To this end, this paper proposes a novel Reinforcement Learning (RL)-based planning approach for cooperative cable-driven transportation. The paper also develops sampling-based and optimization-based planning approaches based on the existing planning approaches for Cable-Driven Parallel Robots (CDPRs). The paper applies these planning approaches to cooperative cable-driven transportation scenarios in simulation and in the real world. The time consumption of these planning approaches and the quality of planning results are compared and analyzed. Based on the comparison of these planning approaches, the paper provides a guideline for the selection and development of planning approaches for cooperative cable-driven transportation with mobile robots. Note to Practitioners—This paper was motivated by the limited capability of a single mobile robot to transport a large payload and the advantages of a cooperative cable-driven transportation system in moving over low obstacles and reducing the mutual influence of mobile robots through flexible connections. This paper focuses on the simultaneous path and motion planning of a cooperative cable-driven transportation system with a given path segment in a small-scale environment. Existing studies of sampling-based, optimization-based, and Reinforcement Learning (RL)-based planning approaches for Cable-Driven Parallel Robots (CDPRs) usually avoid highlighting the weaknesses of a proposed planning approach or cannot be applied to a cooperative cable-driven transportation system. This study proposes an RL-based planning approach, for the first time, and develops sampling-based and optimization-based planning approaches for the simultaneous path and motion planning of a cooperative cable-driven transportation system. The implementation of the three planning approaches is presented. The time consumption and length of planned paths achieved by the three planning approaches are systematically compared in simulation and in the real world. With the presented implementation and comparison of three planning approaches, practitioners can select and implement a proper planning approach to a cooperative cable-driven transportation system according to their requirement of time consumption and length of planned paths. In future research, this study will continue to include additional indices, such as the smoothness of planned paths and wrench feasible workspace.}, archive = {J_TASE}, author = {Hantao Jiang and Wenrui Xie and Bolin Zhou and Weifeng Zeng and Hao Xiong and Yunjiang Lou}, doi = {10.1109/TASE.2025.3596494}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19667--19684}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Simultaneous path and motion planning approaches for cooperative cable-driven transportation with mobile robots: A comparative study}, volume = {22}, year = {2025}, }
• (2025). Early grasp prediction with incomplete data via spatial gating and temporal weighting for teleoperation. TASE, 22, 19655--19666. (www)

  @article{ , abstract = {Accurate and prompt speed grasp intention recognition is crucial in online human-robot interaction (HRI). However, dynamic grasping relying on complete motion for high recognition accuracy will lead to an unavoidable delay in real-time prediction. To address this issue, we propose a Spatial Gating and Temporal Weighting Early Grasp Prediction (STEGP) method that utilizes incomplete dynamic grasping data from sliding windows to reduce the time delay for reliable robot teleoperation. The proposed method comprises a synergy-based feature extraction module, a spatial gating classification module, and a time-decay weighting fusion prediction module. The spatial-temporal mechanism with gating units effectively classifies sequential movements, achieving performance comparable to that of Transformers but being much easier to implement. Integrating a time-decay weighting frame enables reliable early prediction even with incomplete data. gMLP is chosen for the classification of hand dynamic grasping because of its high accuracy, realizing 93.83% accuracy for 33 grasping categories. The prediction tests demonstrated 85.4% accuracy, with the first 25% grasp completion across 28 subjects. Online robotic teleoperation grasp experiments achieved a 57.4% reduction in time delay and a 93.3% success rate. Note to Practitioners—Gesture recognition-based robotic hand control offers the advantages of intuitive operation and incremental configuration. However, motion/gesture recognition can cause delays in the movement of the robotic hand due to the lag associated with the operation of hand motion and the recognition time of the model. We propose a grasping prediction framework that leverages incomplete hand movements for recognition to address this issue. We achieved high recognition accuracy at the early stages of hand motion for 33 types of grasps. Experimental results demonstrate that the delay of robotic hand execution relative to the human hand is significantly minimized. The study enhances the efficient control of the robotic hand through the natural interaction of the human hand grasping.}, archive = {J_TASE}, author = {Yanping Dai and Ning Li and Wenxue Wang and Wenyuan Chen and Guangyong Li and Ning Xi and Lianqing Liu}, doi = {10.1109/TASE.2025.3596443}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19655--19666}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Early grasp prediction with incomplete data via spatial gating and temporal weighting for teleoperation}, volume = {22}, year = {2025}, }
• (2025). Distributed practical predefined-time output optimal allocation of resources for heterogeneous multi-agent systems. TASE, 22, 19646--19654. (www)

  @article{ , abstract = {This paper addresses the issue of distributed predefined-time output optimization for heterogeneous multi-agent systems with equality constraints. A novel distributed algorithm is proposed to enable state tracking of compensator variables under regulator equations assumption. By utilizing the properties of compensator variables and strongly convex functions, a predefined-time output optimization algorithm is proposed to achieve optimal output while satisfying given constraints. A key advantage of this algorithm is its flexibility, allowing convergence time is irrespective of control parameter constraints. Furthermore, a simplified segmented predefined-time optimization algorithm is presented to achieve output optimality within the specified time, reducing the computational and transmission information among agents. Finally, the practicality of the theoretical results is validated through a numerical example. Note to Practitioners—This paper is inspired by resource allocation and optimization problems for homogeneous multi-agent systems. Traditional methods often focus on homogeneous systems, which limits their applicability in diverse industrial environments. To address the urgent need for completing tasks within specified time, two distributed optimization algorithms with predefined-time were developed. These algorithms not only deal with systems with different state dimensions but also allow for flexible setting global objective minimization time. This capability is crucial in resource management applications, such as in smart grids and logistics transportation. Currently, our research focuses on resource allocation problems constrained by equality in undirected networks. In the future, we aim to extend these algorithms to more complex scenarios, integrating both equality and inequality constraints in directed networks to enhance their practicality in real-world applications.}, archive = {J_TASE}, author = {Wanli Jin and Huaguang Zhang and Yapeng Yang and Xiyue Guo and Siyu Chen}, doi = {10.1109/TASE.2025.3593336}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19646--19654}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed practical predefined-time output optimal allocation of resources for heterogeneous multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Distributed unknown inputs observer under hybrid communication disturbances. TASE, 22, 19633--19645. (www)

  @article{ , abstract = {In this paper, we focus on the problem of distributed state estimation for a target system driven by continuous-time linear time-invariant (LTI) system with unknown inputs, where the communication channels are disturbed by stochastic noise and subjected to impulsive sequential attacks. In multi-agent systems (MASs), agents usually move along relatively smooth paths, but occasionally take evasive actions or make sudden turns. This pattern of motion results in relatively small communication disturbances most of the time, but large communication disturbances can occasionally occur. First, this communication disturbances are modeled as a hybrid form: stochastic noise and impulsive sequential attacks, where stochastic noise stands for relatively small communication disturbances most of the time, and impulsive sequential attacks stand for occasionally large communication disturbances. Then, a practical/bounded distributed unknown input observer (DUIO) under stochastic noise and impulsive sequential attack is proposed. Furthermore, the sufficient conditions for the realization of the practical/bounded DUIO in the sense of mean square convergence are presented by graph theory and stochastic analysis. The error bound that depends on the coupling gain, stochastic noise intensity, and average impulsive interval is provided. Finally, the validity of the designed practical/bounded DUIO is confirmed by numerical simulations. Note to Practitioners—Information exchange among agents is a critical for achieving coordination in MASs. Most existing works usually assume that the communication is ideal, i.e., each agent is able to accurately obtain information from its neighbors or is only disturbed by stochastic noise. However, agents generally move along relatively smooth paths but may occasionally execute evasive maneuvers or abrupt turns in MASs. This motion pattern usually results in relatively small communication disturbances, though in certain instances, significant disruptions may occur. How to effectively deal with the impact of such non-ideal communication environments on system performance and ensure that agents can still maintain stable cooperation in complex dynamic environments has become one of the critical topics in the field of control. Specifically, this problem involves several challenges: First, the random and non-Gaussian characteristics of communication disturbances render traditional filtering and estimation methods difficult to apply directly; Second, the burstiness and uncertainty of impulsive noise can lead to severe fluctuations in system states, thereby affecting coordination among agents. In light of this, this paper models such communication disturbances as stochastic impulsive noise and proposes a practical/bounded DUIO under stochastic impulsive noise.}, archive = {J_TASE}, author = {Yali Wu and Housheng Su and Tao Liu and Graziano Chesi}, doi = {10.1109/TASE.2025.3597309}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19633--19645}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed unknown inputs observer under hybrid communication disturbances}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time consensus control for nonlinear multi-agent systems with output constraint: A bounded time-varying gain-based method. TASE, 22, 19619--19632. (www)

  @article{ , abstract = {This paper explores the prescribed-time consensus control problem for a class of high-order strict-feedback nonlinear multi-agent systems with output state constraints. A distributed adaptive $\mathit {C^{1}}$ smooth control scheme is developed such that the output-constrained consensus tracking is achieved within a prescribed time regardless of any initial conditions and other design parameters, although in the presence of completely unknown control gains and uncertainties. The prescribed-time control scheme is made possible for high-order systems by constructing a hybrid time-varying gain-based prescribed-time first-order filter and using the time transformation methods. Particularly, a novel uniformly switching mechanism is introduced into the time-varying gains, where the gain function is $\mathit {C^{1}}$ smooth and the switching time is regardless of initial conditions, ensuring that the gains are uniformly bounded for all time and thus avoiding infinite time-varying gain problem. Finally, the benefits and effectiveness of the proposed control scheme are confirmed by a numerical simulation and its application to a 2 degree of freedom robotic manipulator. Note to Practitioners—This paper investigates the distributed adaptive $C^{1}$ smooth control problem for nonlinear multi-agent systems with high-order strict-feedback dynamics, which can model various physical systems, such as flight systems, robotic systems, and autonomous aerial vehicles. Achieving distributed prescribed-time control for high-order MASs with completely unknown control gains, output constraint, and uncertainties poses significant challenges. By introducing a novel $C^{1}$ smooth gain switching mechanism, a hybrid time-varying gain function is constructed to ensure uniform boundedness of the time-varying gain, which effectively avoids the high-gain implementation problem. Moreover, a prescribed-time distributed $C^{1}$ smooth controller, operating smoothly for all time and everywhere is also forwarded, which is more desirable in practice. As a result, the proposed control scheme demonstrates high adaptability to practical engineering requirements and exhibits enhanced generalizability, as finally validated by the numerical simulations including a dual-degree-of-freedom robotic manipulator.}, archive = {J_TASE}, author = {Bing Cui and Yan Peng and Ling Mao and Zhenhua Pan and Yuanqing Xia}, doi = {10.1109/TASE.2025.3596994}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19619--19632}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time consensus control for nonlinear multi-agent systems with output constraint: A bounded time-varying gain-based method}, volume = {22}, year = {2025}, }
• (2025). Development and evaluation of an automated computational approach for the precise placement of pedicle screws in spinal surgery leveraging three-dimensional point cloud registration methods. TASE, 22, 19606--19618. (www)

  @article{ , abstract = {This study focuses on enhancing the precision and efficacy of pedicle screw placement in spinal surgeries, particularly for patients with osteoporosis. It emphasizes the importance of accurate screw positioning to maximize pullout strength and biomechanical efficacy. The research highlights the relationship between CT values, Bone Mineral Density (BMD), and the Young’s modulus of bone tissue, suggesting that higher CT values, indicative of denser bone, lead to stronger mechanical properties. This understanding is crucial for assessing bone health, especially in osteoporosis and fracture risk analysis. The study introduces an innovative automated planning method for pedicle screw insertion in spinal vertebrae, beneficial for osteoporotic patients. This method uses PointNet++ combined with a Siamese network for semi-supervised segmentation of vertebrae in CT images, converting them into point clouds for individualized planning. The approach aims to improve accuracy and efficiency in pedicle screw placement, especially in complex or osteoporotic vertebrae. The method was validated using osteoporotic in vitro models, demonstrating its potential effectiveness for surgeons facing challenges in pedicle screw placement in osteoporotic patients. The complete workflow begins with semi-supervised vertebra segmentation using a Siamese PointNet++ architecture. The extracted vertebra point cloud is then aligned with a predefined pedicle screw model via an adaptively tuned Super4PCS registration algorithm to generate patient-specific trajectories. The paper also reviews traditional surgical path planning, which relied on surgeons’ experience and intuition, and the shift towards Computer-Aided Design (CAD) and Virtual Reality (VR) technologies for pre-operative planning. Despite these advancements, challenges remain in accurately simulating tissue properties and managing physiological variations. The study proposes a bifurcated approach to automated trajectory planning: segmenting individual vertebrae and planning pedicle screw implantation for each segmented vertebra. The method involves converting CT scans into 3D point clouds, using PointNet++ and the Siamese method for vertebra segmentation, and registering the pedicle screw point cloud with the vertebra model for precise surgical planning. The study demonstrates the method’s feasibility through clinical data from Shengjing Hospital, showing high applicability and consistency in surgical path planning, particularly in the lumbar and lower thoracic regions. The planning results were consistent with surgeons’ experience, indicating the algorithm’s adaptability and stability. Note to Practitioners—Pedicle screw placement is a critical procedure in spinal surgery, especially for osteoporotic patients where achieving strong fixation is essential. Traditional methods depend heavily on the surgeon’s expertise and the use of intraoperative fluoroscopy, which can be both time-consuming and susceptible to human error. Recent advancements propose an automated computational approach that combines deep learning techniques, such as PointNet++ and Siamese networks, for semi-supervised vertebral segmentation. This method also employs an enhanced Super4PCS algorithm for registration-based trajectory planning, aiming to provide a more accurate, consistent, and efficient strategy for pedicle screw placement. This innovative approach offers several key benefits: Automated Vertebral Segmentation and Trajectory Planning: By reducing the need for manual intervention, this method delivers reliable and reproducible paths for pedicle screw insertion based on CT imaging. Enhanced Precision and Safety: By identifying optimal screw trajectories in high-density bone regions through CT value analysis, the approach maximizes pullout strength and minimizes the risk of vertebral perforation, which is particularly beneficial for osteoporotic patients. Clinical Validation and Feasibility: Validated using clinical data from Shengjing Hospital, the study demonstrated alignment with expert surgical decisions. The deviation from planned trajectories in actual surgeries was less than 1 degree, confirming high accuracy. Integrating this method into surgical navigation systems can assist orthopedic surgeons in preoperative planning and provide real-time intraoperative guidance. This integration can lead to minimized surgical risks and complications, reduced radiation exposure by decreasing fluoroscopy usage, and improved surgical workflow efficiency, making spinal surgeries faster and more predictable.}, archive = {J_TASE}, author = {Guoli Song and Andi Li and Yuhan Ying and Yiwen Zhao and Xingang Zhao and Lei Zhang}, doi = {10.1109/TASE.2025.3593231}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19606--19618}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development and evaluation of an automated computational approach for the precise placement of pedicle screws in spinal surgery leveraging three-dimensional point cloud registration methods}, volume = {22}, year = {2025}, }
• (2025). Higher-order sinusoidal input describing function analysis and optimal tuning of hybrid integrator-gain systems. TASE, 22, 19594--19605. (www)

  @article{ , abstract = {The hybrid integrator-gain system (HIGS) is a nonlinear control element with significant potential to enhance controller performance in precise motion control. However, existing HIGS-based controller design methods typically rely on describing functions (DFs), which neglect higher-order harmonics. This paper proposes a novel HIGS-based constant in gain lead in phase (CgLp) element, allowing flexible adjustment of the induced phase lead and high-frequency distortion. Higher-order sinusoidal input describing functions (HOSIDFs) are employed as tools for frequency-domain analysis, providing a method to predict the closed-loop behavior of HIGS-controlled systems by connecting open-loop and closed-loop characteristics. Optimal tuning guidelines for HIGS-based controllers are developed using the proposed tools, accounting for the harmonic effects on closed-loop behaviors. Simulation and experimental results on an H-type precise air-floating motion control platform validate the accuracy of the proposed analysis tools and the effectiveness of the optimal tuning method. Note to Practitioners—This paper was motivated by the challenge of closed-loop behavior evaluation and parameter tuning for nonlinear controllers in precise motion control. This paper proposes a nonlinear phase lead element with constant magnitude and adjustable phase lead, which can be easily integrated into linear controllers. In addition, this paper introduces frequency-domain analysis tools based on higher-order describing functions and an iterative closed-loop performance prediction method, which accounts for higher-order harmonics introduced by nonlinearity. An optimization design method, analogous to the loop shaping method in linear control design, is also presented. Application to the H-type precise air-floating motion control platform demonstrates the effectiveness of the proposed analysis tools and controller design methods. Future research will focus on further suppressing high-frequency distortion in nonlinear controllers.}, archive = {J_TASE}, author = {Jun Wu and Jiehui Zhou and Yifeng Shao and Xiaoqing Li}, doi = {10.1109/TASE.2025.3597594}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19594--19605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Higher-order sinusoidal input describing function analysis and optimal tuning of hybrid integrator-gain systems}, volume = {22}, year = {2025}, }
• (2025). Fully-distributed resilient consensus of multi-agent systems with state and input sector nonlinearities under stochastic denial-of-service attacks. TASE, 22, 19582--19593. (www)

  @article{ , abstract = {This paper addresses a fully-distributed consensus control scheme for multi-agent systems (MASs) with state and input nonlinearities under denial-of-service (DoS) attacks. When agents interact with their neighbors, an adversary with malicious intent randomly launches DoS attacks on specific communication channels in the network, resulting in Markovian switching communication topology. Under stochastic DoS attacks, a fully-distributed secure consensus algorithm is proposed using Lyapunov function analysis that is based solely on immediate partial information per node rather than global information. In addition, the developed approach does not require any specific probability or topology pattern under DoS attack. Furthermore, a novel treatment of the input nonlinearity has been established. For this purpose, the input nonlinearity has been transformed, and a condition in terms of a cumulative nonlinearity bound has been formulated via graph characteristics and algebraic properties. Resilient consensus control design conditions have been established by taking into account the input and state nonlinearities. The developed methods only require a spanning tree between MASs for a union of graph topologies under malicious attacks. Moreover, the resultant consensus protocol is fully-distributed by means of an adaptation mechanism, which has been specifically developed for the directed topologies. The efficiency of the consensus approach has been demonstrated using a simulation of robotic systems along with comparison with a closely-related existing work. Note to Practitioners—The purpose of the present study is to provide a control approach for the consensus of nonlinear MASs under DoS cyber-attacks. Nonlinearities in the state equation and control input are considered, which arise owing to the modeling process of physical systems and actuator’s nonlinear characteristics, respectively. A property for dealing with input nonlinearities in MASs has been provided from the consensus control perspective. DoS attacks can be launched by malicious attackers in cyber-physical systems by blocking the communication channels for information flow. A control scheme has been provided by application of time-varying and adaptive parameters to deal with the resilient consensus control of nonlinear MASs under DoS attacks. In addition, a simple design method is established for practitioners to determine the controller gains using numerical routines. The developed distributed consensus method does not require the knowledge of communication topology to find the parameters of the control systems, and such a method is referred to as a fully-distributed consensus approach. The resultant scheme considers the directed graph topologies which do not require bidirectional communication between any two systems. The study has been tested for the consensus of robotic agents under DoS attacks, and the results are compared with the methods available in the literature.}, archive = {J_TASE}, author = {Musawir Hussain and Fatima Tahir and Muhammad Rehan and Choon Ki Ahn and Ijaz Ahmed and Muhammad Khalid}, doi = {10.1109/TASE.2025.3597153}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19582--19593}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully-distributed resilient consensus of multi-agent systems with state and input sector nonlinearities under stochastic denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Hierarchical safe reinforcement learning control for leader-follower systems with prescribed performance. TASE, 22, 19568--19581. (www)

  @article{ , abstract = {This paper proposes a hierarchical safe reinforcement learning with prescribed performance control (HSRL-PPC) scheme to address the challenges of interconnected leader-follower systems operating in complex environments. The framework consists of two levels: at the higher level, the leader agent detects and avoids moving obstacles while planning optimal paths; at the lower level, the follower agent tracks the leader within strict prescribed performance bounds. We formulate the optimal prescribed performance safe control problem and solve it using the Hamilton-Jacobi-Bellman (HJB) equation. Due to system nonlinearity and obstacle complexity, we approximate the leader’s optimal value function using a state-following neural network that efficiently extrapolates training data to neighboring states, while employing a regular critic neural network for the follower’s value function approximation. Lyapunov stability analysis demonstrates the closed-loop system’s theoretical guarantees. Experimental results from two simulation examples and hardware tests with a quadcopter-vehicle system validate the effectiveness of the proposed approach in achieving safe navigation and precise tracking performance in dynamic environments. Note to Practitioners—Challenges exist in unpredictable obstacles and agent limitations for the interconnected leader-follower system. To provide a safe, efficient, and reliable control scheme, hierarchical safe reinforcement learning with prescribed performance control is proposed in this paper. The hierarchical structure is utilized to coordinate the leader and follower agents in the interconnected system, where the leader agent plans the optimal path and avoids obstacles, and the follower agent tracks the leader within prescribed performance bounds. Based on the proposed hierarchical structure, engineers can design efficient and safe control schemes for interconnected leader-follower systems with moving obstacles. In future work, we will address the problem of external disturbances and uncertainties in the interconnected leader-follower system.}, archive = {J_TASE}, author = {Junkai Tan and Shuangsi Xue and Huan Li and Zihang Guo and Hui Cao and Badong Chen}, doi = {10.1109/TASE.2025.3596912}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19568--19581}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical safe reinforcement learning control for leader-follower systems with prescribed performance}, volume = {22}, year = {2025}, }
• (2025). Adaptive backstepping optimal tracking control of interconnected robotic manipulator system based on reinforcement learning. TASE, 22, 19555--19567. (www)

  @article{ , abstract = {In this paper, an optimal control method based on reinforcement learning (RL) is proposed for the tracking control problem of robotic manipulator (RM) system. In contemporary industrial manufacturing and precision technology, RM has gained significant popularity. To solve the optimal tracking control problem for RM, its dynamics are decomposed into interconnected subsystems. Combining the backstepping method and the RL method, both the virtual control signal and the actual control signal are designed as the optimal solution of each error subsystem, while ensuring that the virtual controller and the actual controller of all subsystems are optimal. For the inherent nonlinearity and difficult solution of the Hamilton–Jacobi–Bellman (HJB) equation, an actor-critic neural network (NN) is constructed to approximate the nonlinear term, and a cost function with input gain function is introduced to ensure the tracking effect. Then, based on the Lyapunov stability theory, it is proved that all the error signals of the system are semi-globally uniformly ultimately bounded (SGUUB). Finally, the effectiveness of the proposed method is validated through 2-degree-of-freedom (DOF) and 3-DOF RM simulations, with robustness verification against sudden disturbances. Note to Practitioners—The tracking control problem of RM is a very extensive and important problem in industrial applications, such as delivery, grasping, etc. The optimal control theory can solve the problems of shortest path, least fuel and shortest time. When RM perform different tasks, different cost functions can be constructed according to the control objectives. Most of the existing optimal control methods for RM systems cannot guarantee the optimality of the entire controller. Inspired by this, this paper develops a new RM optimal tracking control strategy, which can ensure that the virtual controller and the actual controller are the optimal solution, so as to ensure the overall optimality. Based on the RL method, the problem of solving the HJB equation is solved, in which the actor-critic NN weights are updated online at the same time. The actor NN is used to execute control actions, and the critic NN evaluates the control actions and feeds back to the actor NN to improve the subsequent control input.}, archive = {J_TASE}, author = {Hang Su and Zilin Zhang and Weihai Zhang}, doi = {10.1109/TASE.2025.3596555}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19555--19567}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive backstepping optimal tracking control of interconnected robotic manipulator system based on reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Neurodynamics-driven model predictive control with soft-measurement for desulfurization system. TASE, 22, 19546--19554. (www)

  @article{ , abstract = {Sulfur dioxide ( $\mathrm {SO_{2}}$ ) emissions are the main problems causing the air pollution and respiratory disease, which makes desulfurization important in the coal combustion and chemical industry processes. The wet-flue-gas desulfurization (WFGD) is currently an effective method to deal with this issue. However, the WFGD system is actually a complex process with multi-variable coupling, high nonlinearity and large time-delay, which brings great challenges for the traditional mechanism-based modeling and control. In this paper, a neurodynamics-driven model predictive control (NDMPC) with soft-measurement is proposed to predict and control the dynamics of $\mathrm {SO_{2}}$ to improve the operational performance of the WFGD system. First, we design a self-organizing fuzzy neural network (SOFNN) as the soft-measuring model, which is driven by the data and neurodynamics to adaptively predict $\mathrm {SO_{2}}$ in the WFGD system. Second, the designed SOFNN is considered as a predictive model that can give the predicted values for the future moments. Third, we also design a loss function where the one-step output of the predictive model and the control law are the independent variables. The resulting rolling optimization can give the optimal control law sequence by minimizing the loss function. Finally, simulation experiments on the practical data from the WFDG system show that the NDMPC outperforms the other methods in terms of the soft-measurement and control performance, especially generally reducing the $\mathrm {SO_{2}}$ emission and economic cost by 60.25% and 22.27%, respectively. Note to Practitioners—The proposed NDMPC framework presents an effective method to solve the optimal feedstock. which can reduce costs while maintaining the operational efficiency of WFGD system. The optimal solution of NDMPC is dominated by soft-measuring model (SOFNN) to some extent. In the practical applications of WFGD systems, it is better for practitioners to obtain more historical data so that the more accurate soft-measuring model is available for the optimal solution of NDMPC.}, archive = {J_TASE}, author = {Gongming Wang and Xinyi Li and Ke Gu and Hong Chen and Honggui Han and Junfei Qiao}, doi = {10.1109/TASE.2025.3594118}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19546--19554}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Neurodynamics-driven model predictive control with soft-measurement for desulfurization system}, volume = {22}, year = {2025}, }
• (2025). Distributed adaptive consensus of physically interconnected multi-agent networks with uncertainties. TASE, 22, 19538--19545. (www)

  @article{ , abstract = {This paper considers the distributed cooperative control of physically interconnected multi-agent networks in the presence of nonlinear and unknown uncertainties that is one of the key problems in robot swarms and autonomous vehicle fleet. A distributed adaptive control framework is established, under which edge-based adaptive algorithms and node-based adaptive algorithms are designed, respectively. In the designed protocols, a linear term is used to force the states of all agents to be the same and a nonlinear term inspired by the sliding-mode control is introduced to suppress the effect of the undesirable uncertainties. Compared with the exiting related works, the algorithms presented here are applicable to physically interconnected systems, and they are scalable without requiring any global information and robust to various uncertainties. The results are also extended to the case of non-matching uncertainties. The effectiveness of the protocols is verified by an example of multiple coupled vehicles. demonstrating their potential to enhance the robustness and adaptability of robotic systems in dynamic and complex operational environments.Note to Practitioners—This research addresses the challenge of achieving consensus in physically interconnected multi-agent networks, such as autonomous vehicle fleets or robotic swarms, particularly in environments with uncertainties. Traditional methods often require global information, which can be impractical. Our study introduces a distributed adaptive control framework that allows these systems to operate effectively without global data. This framework uses edge-based and node-based adaptive algorithms to enhance system stability and adaptability. In this paper, we improve reliability and efficiency in coordinating multi-agent systems. For example, autonomous vehicles can maintain formation and synchronize movements despite environmental unpredictability. Similarly, robotic swarms can adapt to changing conditions during tasks like search and rescue. The fully distributed nature of our algorithms reduces computational and communication overhead, making them scalable for large-scale applications. Future work will explore extending these protocols to directed graphs and integrating event-triggered communication to further optimize performance.}, archive = {J_TASE}, author = {Zhongpan Zhu and Bin Cheng and Pengpeng Zhang}, doi = {10.1109/TASE.2025.3596071}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19538--19545}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed adaptive consensus of physically interconnected multi-agent networks with uncertainties}, volume = {22}, year = {2025}, }
• (2025). A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data. TASE, 22, 19525--19537. (www)

  @article{ , abstract = {Spatio-temporal data like a Dynamically Weighted Directed Network (DWDN) are ubiquitous in bigdata applications like an intelligent recommender system. They commonly illustrate the complex yet dynamic interactions between tremendous nodes, as well as contain rich knowledge regarding the involved nodes’ behavioral patterns with time dynamics. On the other hand, since the nodes constantly increase as the time accumulates, a DWDN becomes High-Dimensional and Incomplete (HDI) due to the limited interactions, widely-spread time slots and huge node count. Moreover, the inner temporal-spatio patterns exhibit strong nonlinearity, making it very difficult to grasp them from HDI data. To address this critical issue, this paper proposes a novel Tensor Causal Convolution Network (TCCN) model with three-fold ideas: a) innovatively building a feedforward tensor neural network with the incorporation of the Latent Factorization of Tensors (LFT) principle to model complex nonlinear interactions in an HDI DWDN efficiently; b) establishing a Tensor Causal Convolution (TCC) structure, which is able to accurately fuse the time-varying information from node interactions with high scalability; and c) developing a neighborhood regularization scheme to boost the local structural representation, thus capturing spatial dependencies among tremendous nodes. Extensively experimental results on ten real DWDNs from real bigdata applications evidently demonstrate that our proposed TCCN model outperforms several state-of-the-art models in both representation learning accuracy and convergence ability. It provides a highly-efficient representation learning approach for diverse bigdata applications. Note to Practitioners—In modern industrial systems such as recommendation platforms, IoT device networks, and financial transaction monitoring systems, tremendous and temporal interactions emerge among numerous nodes over time. This leads to large-scale dynamic networks that are high-dimensional and incomplete (HDI), making it difficult for traditional methods to extract meaningful behavioral patterns. This paper introduces an efficient method for modeling and representing node relationships and temporal patterns in such a network. The proposed approach, namely the TCCN model, is able to automatically fuse the time-varying features to preserve local structural dependencies, hence significantly improving the interpretation of user behavior, device states, and transaction flows. Experimental results across several real-world industrial scenarios demonstrate TCCN’s superior performance in both representation accuracy and model convergence. In future research, we plan to explore the generative modeling frameworks to improve TCCN’s generalization and adaptability to complex temporal dynamics. In addition, TCNN also holds promise for broader applications such as intelligent transportation systems, industrial process monitoring, and energy management in data-sparse automation contexts.}, archive = {J_TASE}, author = {Xin Liao and Hao Wu and Xin Luo}, doi = {10.1109/TASE.2025.3595545}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19525--19537}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data}, volume = {22}, year = {2025}, }
• (2025). Asynchronous multi-agent collaborative framework for integrated production and procurement optimization in refinery. TASE, 22, 19511--19524. (www)

  @article{ , abstract = {The refinery industry operates as a highly complex system characterized by dynamic interactions among numerous processes, resources, and decision-making strategies. Within this context, production planning and crude oil procurement management are critical to ensuring operational efficiency and profitability. However, traditional optimization approaches often neglect the distinct decision-making time scales of these two functions, limiting their ability to address market dynamics and supply chain complexities effectively. To overcome these challenges, this study introduces an asynchronous multi-agent collaborative optimization framework that enhances the coordination between production planning and crude oil procurement in refinery operations. By enabling production and procurement agents to operate on independent time scales, the framework adapts to fluctuations in crude oil prices and variations in product demand. The study further extends the classical multi-agent reinforcement learning algorithm into an asynchronous paradigm, introducing Async-MAPPO, which allows agents to independently optimize decisions using real-time data. This approach mitigates operational delays and enhances adaptability. Experimental evaluations demonstrate that the proposed method significantly improves production efficiency, reduces operational costs, and strengthens the refinery’s resilience to market fluctuations, underscoring the critical role of asynchronous optimization in refinery operations. Note to Practitioners—Refinery operations involve the intricate coordination of production planning and crude oil procurement, both of which operate on distinct decision-making time scales. This disparity in time scales often complicates efforts to achieve seamless coordination, especially in the face of uncertain disturbances such as fluctuating crude oil prices, variable product demands, and supply chain complexities. This paper proposes an advanced multi-agent reinforcement learning (MARL) framework coupled with an asynchronous collaborative decision-making mechanism. By enabling production and procurement agents to make decisions independently yet collaboratively across different time scales, the proposed framework addresses the dynamic nature of refinery operations. The asynchronous collaboration mechanism allows each agent to respond promptly to real-time data, reducing delays and enhancing adaptability to market fluctuations. This approach provides practitioners with a novel tool for managing the complexities of refinery operations in the future era.}, archive = {J_TASE}, author = {Ya Liu and Bo Chen and Kai Wang and Fei Qiao and Hanli Wang}, doi = {10.1109/TASE.2025.3595265}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19511--19524}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous multi-agent collaborative framework for integrated production and procurement optimization in refinery}, volume = {22}, year = {2025}, }
• (2025). Robust fault-tolerant control based on set-membership estimation and tube MPC. TASE, 22, 19499--19510. (www)

  @article{ , abstract = {A robust fault-tolerant control method based on set-membership estimation and tube model predictive control (MPC) is proposed for the constrained linear systems subject to unknown disturbance and noise. Firstly, an augmented observer is used to estimate state and fault simultaneously. A fault-tolerant controller is proposed based on the state and fault estimation. And the estimation and feedback errors are bounded using set-membership estimation method. Then, the obtained error sets are used to design robust output feedback tube MPC strategy. In order to make the estimation error robust against disturbance and noise, the observer gain is designed based on $L_{\infty }$ method. Finally, the effectiveness and real-time performance of the proposed method is verified by a hardware-in-the-loop (HIL) test, which shows its potential application value in aeroengine control systems. Note to Practitioners—Many industrial systems for instance the aircraft and aeroengine control systems have high safety requirements. Fault-tolerant control (FTC) especially the active fault-tolerant (AFTC) is an effective way to improve the system security. In addition, the industrial control systems are usually subject to input and state constraints. Therefore, FTC method based on MPC has received increasing attention for the constrained systems especially the aerospace systems. However, how to deal with the model uncertainties including disturbance and noise is an important problem for FTC method based on MPC, which has not be investigated sufficiently. In this paper, we propose a robust fault-tolerant control method based on set-membership estimation and tube MPC for the constrained systems subject to unknown disturbance and noise. Based on the proposed method, the actual state can be guaranteed to satisfy the constraints while the system is subject to unknown disturbance, noise and actuator fault. The set-membership estimation method is introduced to reduce the computational burden of obtaining robust error sets for tube MPC. Moreover, by using the properties of Pontryagin difference and support function of zonotopes, the computational burden of calculating the tube MPC constraints is further reduced. The hardware-in-loop test of an aeroengine model shows the effectiveness of the proposed method and its potential application in aeronautical engineering.}, archive = {J_TASE}, author = {Wenchao Qin and Wentao Tang and Weiguo Xia}, doi = {10.1109/TASE.2025.3594583}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19499--19510}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust fault-tolerant control based on set-membership estimation and tube MPC}, volume = {22}, year = {2025}, }
• (2025). Automating transfer of robot task plans using functorial data migrations. TASE, 22, 19484--19498. (www)

  @article{ , abstract = {This paper introduces a novel approach to ontology-based robot plan transfer by leveraging functorial data migrations, a structured mapping method derived from category theory. Functors provide structured maps between planning domain ontologies which enables the transfer of task plans without the need for replanning. Unlike methods tailored to specific plans, our framework applies universally within the source domain once a structured map is defined. We demonstrate this approach by transferring a task plan from the canonical Blocksworld domain to one compatible with the AI2-THOR Kitchen environment. Additionally, we discuss practical limitations, propose benchmarks for evaluating symbolic plan transfer methods, and outline future directions for scaling this approach. Note to Practitioners—This research was motivated by the challenge of reusing successful robot task plans in different domains without the need to redesign or recalibrate the plans from scratch. In industries where automation is prevalent, such as manufacturing or logistics, the ability to efficiently transfer a task plan from one setting to another can significantly enhance productivity and reduce operational costs. Our solution utilizes functorial data migrations, a concept from category theory, to enable the structured transfer of task plans between distinct domains. This approach ensures that transferred plans remain valid and applicable. For example, a plan developed to transport items in a warehouse can be adapted to transport medical supplies in a hospital, which leverages the underlying structure of the tasks without altering their essence. While our approach generalizes task plan transfer across domains, it relies on precise domain ontology maps, which can be complex. This method may not capture all nuanced transfers between highly-specialized domains, particularly where domain-specific knowledge is comprehensive. Practitioners should be aware that some manual tuning might still be necessary to fully optimize the plans for specific operational contexts.}, archive = {J_TASE}, author = {Angeline Aguinaldo and Evan Patterson and William Regli}, doi = {10.1109/TASE.2025.3586549}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19484--19498}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automating transfer of robot task plans using functorial data migrations}, volume = {22}, year = {2025}, }
• (2025). Online distributed generalized nash equilibrium seeking of energy sharing markets in distribution networks. TASE, 22, 19470--19483. (www)

  @article{ , abstract = {With the proliferation of distributed generations and energy storage systems, traditional passive consumers in the distribution network have been gradually evolving into “prosumers”, who can both produce and consume energy. To alleviate the energy surplus and shortage problem, energy sharing between prosumers is advocated. This paper investigates an online distributed method to seek the generalized Nash equilibrium (GNE) of the energy sharing market. We first formulate the energy sharing problem as a multi-period generalized Nash game (GNG), which has been proven to have a unique variational GNE (v-GNE) with the same multiplier. Then, the GNG is transformed into an equivalent optimization problem based on the variational inequality, which is further divided into multiple single-period problems under the Lyapunov optimization framework. An online distributed algorithm is introduced to solve the Lyapunov drift problems and consequently seek the v-GNE in a time-varying environment. We prove that the expected performance gap is bounded compared to the deterministic offline v-GNE with perfect uncertainty information, while all the constraints across the entire horizon are maintained. Finally, numerical results with 33 and 123 prosumers validate the performance of the algorithm. Note to Practitioners—In a distribution network with massive prosumers, energy sharing with the main grid and among prosumers is inevitable. With the existence of energy storage systems (ESSs), the energy sharing problem is usually formulated as a multi-period GNG with time-coupling constraints, which needs to be solved online due to the integration of volatile renewable generations. However, the existing online distributed GNE seeking algorithm is usually designed for a single-period problem, which is not applicable to the multi-period GNG. In this paper, we first formulate the energy sharing market as a multi-period GNG, and transform it into an equivalent optimization problem. Then, the online solving algorithm is proposed to solve the equivalent problem under the Lyapunov optimization framework. In this way, the multi-period GNG is finally solved online. Simulations on the IEEE 33-bus benchmark show that the accuracy can be guaranteed in a time-varying environment compared with offline method using full information. Results with the IEEE 123-bus system further validate the scalability of our approach. Besides the application in power systems, our approach is also applicable to other scenarios modeled as multi-period GNGs.}, archive = {J_TASE}, author = {Zhaojian Wang and Wenxian Hao and Wei Wei and Zexin Sun}, doi = {10.1109/TASE.2025.3595443}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19470--19483}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online distributed generalized nash equilibrium seeking of energy sharing markets in distribution networks}, volume = {22}, year = {2025}, }
• (2025). Adaptive-critic-based distributed event-triggered augmented differential game for cooperative guidance systems against maneuvering target. TASE, 22, 19451--19469. (www)

  @article{ , abstract = {This paper investigates the cooperative interception guidance issue with the maneuvering target, which can be described as a distributed augmented differential game under direct topology by decomposing the nonlinear uncertain term related to the maneuvering target into matched term and unmatched term. An adaptive event-triggered distributed augmented differential game method is proposed to obtain the solution of a coupled distributed differential game online. Then, a guidance performance adaptive critic network is employed to calculate the event-triggered augmented cost function. Moreover, the adaptive triggering condition is designed, which decomposes the original continuous system into a continuous/discrete hybrid one consisting of flow dynamics and jump dynamics. Certainly, all error signals have been theoretically ensured to be within an adjustable bounded region while excluding infamous Zeno behavior. Finally, the advantages, practicalities, and potential of the proposed method are exhibited by numerical simulations and experiments. Note to Practitioners—Cooperative guidance, which can effectively increase the interception probability of the target, receives increasing attention. Considering the effects of maneuverability of the target generating great disturbance and uncertainty, the differential game cooperative guidance technique is vital for improving the guidance performance. Besides, cooperative methods based on traditional time-triggered mechanisms need to exchange information between missiles and to update guidance law periodically, which brings great challenges in practical engineering with limited communication and computing resources. The proposed method can effectively achieve cooperative interception for maneuvering targets and has great potential for practical engineering applications.}, archive = {J_TASE}, author = {Jingliang Sun and Mengmeng Wang and Zhenlin Zhou and Jianxin Zhong and Junzhi Li and Teng Long}, doi = {10.1109/TASE.2025.3594043}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19451--19469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive-critic-based distributed event-triggered augmented differential game for cooperative guidance systems against maneuvering target}, volume = {22}, year = {2025}, }
• (2025). Segmentation and quantification of surface defects in 3D reconstructions for damage assessment and inspection. TASE, 22, 19439--19450. (www)

  @article{ , abstract = {Detecting surface defects is crucial for maintaining the integrity of critical infrastructure. Traditional RGB image-based methods are limited by their reliance on 2D information, which impairs accurate damage assessment. This paper introduces a novel approach that enhances defect detection and quantification, utilizing dense 3D reconstructions generated through techniques like photogrammetry or profilometry. We develop an improved robust spline fitting algorithm to estimate the undamaged surfaces from the 3D reconstructions. The residual distances between the observed and fitted surfaces are subsequently used to segment and quantify defects. By leveraging 3D data, our method resolves visual ambiguities and enables damage quantification using physically meaningful metrics. For 3D models based on optical sensing, our method complements RGB image-based defect detectors and classifiers, facilitating the fusion of visual and 3D information for a more comprehensive defect analysis. Validated on both synthetic and real-world datasets, our method demonstrates strong performance and practical feasibility. Note to Practitioners—Our research is driven by the growing potential of drone-based inspections using high-resolution imaging platforms, which offer significant advantages for monitoring remote or hard-to-reach infrastructure. With high-quality images, we can use photogrammetry to reconstruct accurate 3D models directly from inspection images, without requiring additional sensors or manual intervention. We leverage this 3D data to improve the robustness of automated defect detection and enable the precise quantification of defect sizes and material loss. Tracking these metrics over multiple inspections can provide valuable information for preventive and predictive maintenance, moving us closer to efficient, comprehensive structural health monitoring. While our method is highly effective across various surfaces, it is less suited to detecting widespread shallow damage, which could benefit from incorporating additional geometric constraints.}, archive = {J_TASE}, author = {Jonathan Sterckx and Michiel Vlaminck and Hiep Luong}, doi = {10.1109/TASE.2025.3593967}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19439--19450}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Segmentation and quantification of surface defects in 3D reconstructions for damage assessment and inspection}, volume = {22}, year = {2025}, }
• (2025). Order fulfillment in a robotic mobile fulfillment system of a business-to-business environment. TASE, 22, 19428--19438. (www)

  @article{ , abstract = {The robotic mobile fulfillment system (RMFS) is a robotized parts-to-picker material handling system used in warehouses to save traveling and searching time, as well as prevent human errors in the order picking processes. This study focuses on the order fulfillment within the RMFS, which plays a crucial role in determining the system’s overall throughput. Although there have been a few works addressing rack and order allocation in the literature, the specific context of a business-to-business (B2B) environment has not received substantial attention. In B2B scenarios, the volume of ordered items and quantities are higher compared to business-to-customer (B2C) environments. Fulfilling an item of an order require multiple racks due to a large quantity. To address this research gap, a mathematical formulation is proposed. The problem is proved to be an NP-hard problem. Several properties are derived and used for identifying infeasible solutions as well as reducing the feasible solution space. Furthermore, we introduce two lower bound estimation methods to enhance computational efficiency. Leveraging these derived properties and lower bound estimation methods, an efficient extended A-star algorithm is developed to efficiently find the optimal order fulfillment. The experimental results reveal the efficiency of the proposed algorithm. Note to Practitioners—Robotic mobile fulfillment systems are widely used in warehouses to enhance order fulfillment efficiency. This study examines the challenges of fulfilling large orders that require multiple racks to accommodate the numerous items. The key contribution of this research is the development of an efficient algorithm based on the A-star method, designed to help warehouse robots minimize total travel distance. This leads to faster order processing and reduced operational costs. The algorithm has been tested in a real-world setting, demonstrating its ability to improve RMFS performance. Practitioners can adopt this approach to optimize warehouse operations and more effectively manage the complexities of order fulfillment.}, archive = {J_TASE}, author = {Edward Huang and Chun-Chih Chiu and James T. Lin}, doi = {10.1109/TASE.2025.3592618}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19428--19438}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Order fulfillment in a robotic mobile fulfillment system of a business-to-business environment}, volume = {22}, year = {2025}, }
• (2025). Adaptive finite-time sliding optimal tracking control for uncertain linear systems with disturbances. TASE, 22, 19415--19427. (www)

  @article{ , abstract = {This paper proposes an adaptive finite-time sliding optimal tracking control (AFSOTC) for uncertain linear systems by combining finite-time optimal tracking with integral sliding mode technology. The AFSOTC includes two components: an optimal finite-time feedback control based on the finite-time control technique for a nominal model, and an integral finite-time sliding control to attenuate external disturbances. First, a sliding model is constructed to transform the tracking problem into a regular control one. The finite-time optimal controller based on the fixed-time stability technique and adaptive dynamic programming is proposed for the uncertain sliding model system, where an improved finite-time algebraic-Riccati-equation (ARE) with fixed-time convergence is online solved by a novel adaptive law. Then, an integral finite-time sliding control is constructed to resist the uncertain dynamics and interferences. Moreover, Lyapunov theory is applied to prove the parameter convergence and the stability of the closed-loop system with the AFSOTC. Finally, a linear system with comparisons and an autonomous underwater vehicle (AUV) with dynamic uncertainties and disturbance are applied to verify the correctness of the proposed method, and a linearization method for AUV underwater is presented. Note to Practitioners—Many linearized systems in practical engineering require finite-time optimal control to improve the tracking performance and also suffer from unknown disturbances, i.e., aerospace and navigation, F-16 aircraft, and autonomous vehicles. Moreover, the accurate models of these systems are generally difficult to construct because of dynamic uncertainties, such that the control performance will be reduced. The optimal finite-time control problem with disturbance and unknown dynamics for these systems is an open research challenge. Thus, this paper proposes an adaptive finite-time sliding optimal tracking control based on the sliding mode and fixed-time techniques, which can realize a fast optimal performance for uncertain linear systems with disturbances.}, archive = {J_TASE}, author = {Yongfeng Lv and Baixue Miao and Jun Zhao and Fujian Yang}, doi = {10.1109/TASE.2025.3595121}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19415--19427}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive finite-time sliding optimal tracking control for uncertain linear systems with disturbances}, volume = {22}, year = {2025}, }
• (2025). A parallel homotopic optimized control scheme of uncertain nonlinear markov jump systems and its applications. TASE, 22, 19403--19414. (www)

  @article{ , abstract = {This paper investigates the optimized control problem for nonlinear Markov jump systems with uncertain time-varying parameters. First, the Takagi-Sugeno fuzzy model is applied to describe the uncertain nonlinear Markov jump systems, where the identification of the uncertain time-varying parameters is reformulated as zero-sum game problems. Subsequently, the reinforcement learning-based scheme is employed to solve such zero-sum game problems and achieve optimization even when the system dynamics remain unknown. Afterward, a parallel homotopic scheme is first adopted for such systems. The assumption of an initial stabilizing gain in the policy iteration scheme may be removed in this scheme, naturally enhancing the generality of the developed method. Finally, the actual availability of the proposed parallel homotopic scheme is validated through a single-link robot arm and a series DC motor. Note to Practitioners—In actual production processes, nonlinear Markov jump systems are widely applied to describe modern industrial systems subject to parameter jumps. The study of the optimized control for such systems can provide considerable benefits. However, external constraints such as complex physical environments and cost limitations make it difficult to obtain stable control gains when the system dynamics are completely unknown. This challenge has hindered the application of many existing reinforcement learning-based optimized control algorithms that require initial stable control gains. Moreover, the algorithms that do not rely on initial stable control gains frequently encounter issues with long computation times, rendering them unsuitable for industrial environments demanding rapid responses. To overcome these challenges, a parallel homotopic scheme is proposed for the optimized control of nonlinear Markov jump systems. The proposed scheme effectively derives an optimized controller from ensuring relatively fast convergence under weakened conditions, where the system dynamics are unknown and no initial stable gain is available. Additionally, the scheme addresses the uncertain time-varying parameters commonly present in industrial systems by employing the robust control theory. The application of the proposed scheme to a single-link robot arm and a series DC motor demonstrates its effective control performance.}, archive = {J_TASE}, author = {Jing Wang and Zheng Huang and Hao Shen and Ju H. Park}, doi = {10.1109/TASE.2025.3594318}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19403--19414}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A parallel homotopic optimized control scheme of uncertain nonlinear markov jump systems and its applications}, volume = {22}, year = {2025}, }
• (2025). Hierarchical optimization with low-carbon economic dispatch in distribution networks. TASE, 22, 19392--19402. (www)

  @article{ , abstract = {A cloud-based collaborative framework is proposed for the low-carbon economic dispatch problem in distribution networks (DNs). The framework integrates multi-objective optimization and control layers in a hierarchical control approach by coordinating active and reactive power from distributed generation (DG). The upper cloud computing layer comes to multi-objective optimization to minimize the generation cost, line loss and bus voltage deviation under low carbon conditions. The lower control layer regulates the dynamics of the DG using state feedback control combined with a game theory-based $H_{\infty } $ filter. Lastly, the proposed framework effectiveness is verified by StarSim hardware-in-the-loop simulation. Note to Practitioners—This framework addresses practical challenges in optimizing distribution networks for a low-carbon economy. By leveraging cloud computing and advanced control strategies, practitioners can enhance efficiency and reduce the carbon footprint in distribution network operations. Existing research focuses on single-objective optimization or single-layer control structures in distribution networks. In this paper, the upper cloud computing layer performs multi-objective optimization to minimize the generation cost, line losses, and node voltage deviations under low-carbon conditions. The lower control layer regulates the dynamics of the DG using state feedback control combined with a game theory-based $H_{\infty } $ filter. The key application is in improving distribution network performance by simultaneously minimizing costs and environmental impact. Computer simulations and hardware-in-the-loop simulations show that this approach is feasible, but it has not yet been considered for attack defense in distribution networks. In future research, we will address the design of simultaneous attack defense for the cloud computing layer and the control layer.}, archive = {J_TASE}, author = {Cong Cai and Qingyu Su and Jian Li}, doi = {10.1109/TASE.2025.3593484}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19392--19402}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical optimization with low-carbon economic dispatch in distribution networks}, volume = {22}, year = {2025}, }
• (2025). Adaptive neural network unified control for general MIMO underactuated mechatronic systems with disturbances via modified normal forms. TASE, 22, 19377--19391. (www)

  @article{ , abstract = {The control problem of underactuated mechatronic systems is one of the key representatives of complex nonlinear dynamical systems. Starting from the dynamical structures of underactuated systems is usually one of the most direct and effective ways to design controllers. However, controllers developed for specific dynamic models are often difficult to be directly generalized to other underactuated systems. Moreover, due to the lack of control inputs, designing robust controllers for unmatched disturbances (acting on unactuated states) remains a challenging problem. Therefore, based on the Euler-Lagrange dynamics of multi-input-multi-output (MIMO) underactuated systems, this paper gives four coordinate transformations according to different configurations of the inertia matrix, which extends the Olfati transformation to some extent and finally unifies underactuated systems into normal forms. A sliding manifold and an adaptive neural network sliding mode controller are developed with the derived normal forms, which improves transient performance and avoids the chattering problem in traditional sliding mode controllers by combining an estimation error-driven adaptive law, a high-order sliding-mode differentiator, and the super-twisting algorithm. More importantly, the stability is guaranteed by Lyapunov techniques even in the presence of both persistent matched disturbances and asymptotically vanishing unmatched disturbances. Furthermore, the proposed control strategy is applied to overhead cranes and tower cranes, whose superior control performance is verified by hardware experiments. Note to Practitioners—Underactuated mechatronic systems, with fewer independent actuators than degrees of freedom, are common in real-world automation applications, such as cranes, aerial vehicles, and flexible joint robots. However, designing robust and generalizable controllers for such systems remains a significant engineering challenge due to their nonlinear dynamics, structural diversity, and sensitivity to matched and unmatched disturbances. This paper proposes a unified control framework, which is specifically designed to meet the needs of engineers and practitioners working with complex underactuated mechatronic systems. This paper presents a systematic approach to normalizing the dynamics of different underactuated mechatronic systems. Based on the unified model, a robust control strategy is proposed with good applicability. The proposed control strategy improves positioning accuracy and control smoothness, particularly in effectively handling unmatched disturbances (such as sudden impact disturbances on the suspended payload). The effectiveness of the proposed strategy is validated through experiments on two representative underactuated mechatronic systems (tower cranes and overhead cranes), demonstrating its broad potential for automation tasks.}, archive = {J_TASE}, author = {Meng Zhai and Tong Yang and Ming Li and Xuerui Jiao and Yongchun Fang and Ning Sun}, doi = {10.1109/TASE.2025.3593468}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19377--19391}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neural network unified control for general MIMO underactuated mechatronic systems with disturbances via modified normal forms}, volume = {22}, year = {2025}, }
• (2025). Advanced defense strategy for cyber-resilient frequency control in real power grids. TASE, 22, 19366--19376. (www)

  @article{ , abstract = {This paper investigates the load frequency control problem of real power systems equipped with the automatic generation control service under false data injection (FDI) attacks. The available methods either require the exact system model or extensive historical data. Moreover, they cannot distinguish load changes from cyber-attacks. To address these challenges, this paper proposes a model-free defense method comprising an observer and a detector. The observer estimates the targeted signal and compares the difference between the estimated and observed signals as the state estimation error with a predetermined threshold value. When the state estimation error exceeds the threshold value, the detector detects an attack. In this situation, the observer’s output is the state estimation error. This causes the estimated signal to be sent to the secondary controller to mitigate the attack’s effect. If the state estimation error is smaller than the threshold value, no attack has occurred on the system. In such conditions, the observed signal is sent to the secondary controller. The proposed method rebuilds the actual system output in real-time under cyber-attacks. The presented method is independent of the system’s mathematical model and historical data, and it has a simple design procedure. This method can successfully distinguish load changes from FDI attacks. Moreover, it is not dependent on the size and complexity of the power system. The simulation results on the IEEE 14-bus and 39-bus power systems in the DIgSILENT PowerFactory show that the proposed method can timely detect the attacks and completely mitigate the effects of attacks on the system’s dynamic performance. Also, the obtained false alarm and attack detection rates are zero and over 98%, respectively. Note to Practitioners—The second level of the load frequency control scheme, known as the automatic generation control, is significantly dependent on information and communication technologies. This leads to the vulnerability of the LFC application to cyber-attacks, finally compromising the power system’s stability. False data injection (FDI) attacks, as the most common cyber-attacks, falsify signals transmitted during insecure communication networks. Therefore, the LFC performance investigation of power systems under FDI attacks is necessary. The studies related to this research field focused on the LFC problem of linearized power systems under special presumes. The available defense strategies are based on the power system’s precise mathematical model or machine learning. The system model-based methods are not applicable in power systems without a detailed mathematical model. Machine learning-based methods also require extensive historical data on the power system. Moreover, these methods impose a computational burden on the system. The methods presented in the literature are not capable of distinguishing load changes from cyber-attacks. To address these drawbacks, this paper proposes a model-free defense method to cope with FDI attacks. The proposed method is not dependent on the system’s mathematical model and historical data. Also, its design process is simple. The proposed method is independent of the size and complexity of the power system and can be applied to any real power system. The simulation results on the IEEE 14-bus and 39-bus power systems in the DIgSILENT PowerFactory reveal that the proposed defense scheme can timely identify FDI attacks and completely compensate for the effects of attacks on the system’s frequency response.}, archive = {J_TASE}, author = {Soroush Oshnoei and Esmaeil Mahboubi-Moghaddam and Mohammad Hassan Khooban}, doi = {10.1109/TASE.2025.3594891}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19366--19376}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Advanced defense strategy for cyber-resilient frequency control in real power grids}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning event-triggered control with flexible performance assurance for stochastic nonlinear systems. TASE, 22, 19352--19365. (www)

  @article{ , abstract = {This article focuses on an adaptive neural optimal output feedback control strategy for stochastic nonaffine multiple input multiple output nonlinear systems with input saturation. Initially, an emerging local state estimation filter is delicately formulated to identify the unavailable states while economizing the redundant resource usage in the state estimation filter-to-controller channel. Then, the tracking error can be regulated within a flexible envelope range by designing a modified flexible global prescribed time function depending on the intermittent systems at the expense of certain user-prescribed indexes. Technically, an amended nonlinear filter featuring the hyperbolic tangent function is constructed to overcome the curse of dimensionality while compensating for the effect of neglected filter error. Meanwhile, an optimized adaptive event-triggered controller is developed to adjust the triggered threshold online to release networked resources and consumed control expenses employed in the controller-to-actuator channel. Since the trigger criteria of different channels are not correlated with each other, the designer can tune the signal delivery frequency of each channel separately following the practical requirements. The boundedness of all signals is ensured via It $\hat {o}$ ’s differential equation. Herein, two illustrative analyses verify the efficacy and feasibility of the established control algorithm. Note to Practitioners—The motivation stems from the real demand for optimal output feedback control of stochastic nonlinear systems with saturation nonlinearity. Regrettably, current practices are incapable of attaining the anticipated specified behaviors in the context of actuator saturation. To this end, we ingeniously formulate a unique design procedure, distinguished from the prevailing prescribed-based ones, which exerts the flexible envelope range on the tracking error by incorporating the intermittent systems into the modified flexible global prescribed time function to remedy the bottleneck. Moreover, a local state estimation filter is developed to estimate the unavailable states while reducing the utilization of resource usage in the state estimation filter-to-controller channel. On this basis, a dynamic event-triggered protocol is introduced to economize the networked resources in the controller-to-actuator channel. The control solution is more prone to physical application embodied in the two-stage chemical reactors to confirm its viability.}, archive = {J_TASE}, author = {Xiaona Song and Peng Sun and Shuai Song and Choon Ki Ahn}, doi = {10.1109/TASE.2025.3594789}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19352--19365}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning event-triggered control with flexible performance assurance for stochastic nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Model-free adaptive time-delay-based estimation control for input-saturated coupled tank system: Experimental validation. TASE, 22, 19340--19351. (www)

  @article{ , abstract = {Coupled tank systems are widely employed in various industries, including wastewater treatment, water desalination, pharmaceuticals, and petrochemicals. Precise liquid level regulation is a critical challenge in these applications, especially under system uncertainty and disturbance. This paper introduces a novel model-free, robust adaptive time delay-based estimation control to address this challenge, with disturbances and input saturation. The proposed controller employs a time delay-based estimation technique to estimate unknown system dynamics using the immediate past input and output data. This estimated information is then integrated into a composite control scheme to compensate for unknown dynamics and disturbances. A new adaptive-based proportional controller is proposed, which avoids overestimating the adaptive gains. The closed-loop analysis through Lyapunov theory guarantees uniformly ultimately boundedness of all the internal states. Experimental results validate the better performance of the proposed controller compared to existing schemes under different operating conditions. Note to Practitioners—Controlling liquid levels in coupled tank systems is crucial in many industries, from wastewater treatment to pharmaceuticals. This paper presents a new, practical control strategy to address the challenge of precise level regulation, especially when dealing with system uncertainties, disturbances, and limitations on the control input. The proposed approach is “model-free,” meaning it doesn’t require the mathematical model of the tank system, making it easier to implement in real-world scenarios. The controller only requires the past input and output data to estimate the system’s behavior and then incorporates this estimate into a robust control scheme that compensates for unknown dynamics and disturbances. Another key feature is the application of a new adaptive controller, which avoids the problem of overestimating the control gains, thus reducing excessive control effort. The effectiveness of the proposed scheme is experimentally demonstrated by controlling the water levels in a coupled tank system. The experimental results elucidate that the new controller outperforms traditional methods under various operating conditions, offering improved performance for practitioners dealing with coupled tank systems.}, archive = {J_TASE}, author = {Abdullah Abushokor and Syed Muhammad Amrr}, doi = {10.1109/TASE.2025.3594752}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19340--19351}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model-free adaptive time-delay-based estimation control for input-saturated coupled tank system: Experimental validation}, volume = {22}, year = {2025}, }
• (2025). Modular multirotors: From quadrotors to fully-actuated aerial vehicles. TASE, 22, 19328--19339. (www)

  @article{ , abstract = {Traditional aerial vehicles are constrained to perform specific tasks due to their adhoc designs. Based on modularity, we propose a versatile robot, H-ModQuad, that can adapt to different tasks by increasing its load capacity and actuated degrees of freedom. It is composed of cuboid modules propelled by quadrotors with tilted rotors. We present two families of module designs that bring scalable and versatile actuation to the aerial systems. By configuring multiple modules, H-ModQuad can increase its payload capacity and change its actuated degrees of freedom from 4 to 5 and 6. By modeling the actuation capability of H-ModQuad using actuation ellipsoids and wrench polytopes, we find the body frame of a vehicle that maximizes its thrusting efficiency. We also compare the vehicle capabilities against formally defined task requirements. We present the dynamics of H-ModQuad and integrate control strategies despite the vehicle design. The design and model are validated with experiments using actual robots, showing that H-ModQuad vehicles with different configurations provide different actuation properties. Note to Practitioners—This paper presents the H-ModQuad, a modular multirotor system designed to enhance the flexibility and versatility of aerial vehicles. Traditional drones often need to be specifically designed or modified for different tasks, but H-ModQuad allows for quick reconfiguration to meet varying operational needs. By adding or rearranging modules, users can easily increase the vehicle’s load capacity or improve its force and torque generation characteristics. This adaptability makes H-ModQuad particularly useful in industries such as logistics for transporting varying payloads, inspection and maintenance tasks in hard-to-reach areas, and uncrewed tool wielding. Our analytical tools provide pre-assembly vehicle capability assessment, and flight tests demonstrate that different modular configurations grant the vehicle with distinct actuation characteristics suitable for diverse tasks without redesigning the modules. Therefore, H-ModQuad offers a versatile aerial solution, making it a valuable tool for a wide range of autonomous applications.}, archive = {J_TASE}, author = {Jiawei Xu and Diego S. D'Antonio and David Saldaña}, doi = {10.1109/TASE.2025.3594579}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19328--19339}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modular multirotors: From quadrotors to fully-actuated aerial vehicles}, volume = {22}, year = {2025}, }
• (2025). Learning-based parallel control for unknown nonaffine nonzero-sum games. TASE, 22, 19317--19327. (www)

  @article{ , abstract = {In this paper, a novel nonzero-sum game (NSG) method is developed for completely unknown nonaffine nonlinear discrete-time (DT) systems, which is referred to as model-free NSG (MNSG). First, novel dynamic control laws are developed for NSGs using parallel control, namely introducing controls into feedback. Subsequently, an augmented N-player NSG is formulated according to the original N-player NSG to derive the dynamic control laws. Furthermore, we show that the control stabilities of the original and augmented N-player NSGs are equivalent. In the meantime, we prove that optimal control of the augmented N-player NSG is equivalent to near-optimal control of the original N-player NSG, and the Nash equilibrium of the original N-player NSG can be achieved. Then, a model-free learning scheme is developed to obtain the solution of the augmented N-player NSG using online policy iteration, and neither using a model network to predict unknown dynamics nor off-policy reinforcement learning (RL) is needed in the scheme. Lastly, numerical analysis, including the NSG of a DT system with unknown control-nonaffine dynamics and coupled controls, confirms the correctness of our MNSG method. The associated code is available at: https://github.com/lujingweihh/Adaptive-dynamic-programming-algorithms/tree/main/model_free_nonzero_sum_games_discrete_time Note to Practitioners—With the rise of information technologies and infrastructures, many real-world systems are now manipulated by multiple controllers, and the complexity of such systems is increasing. Hence, nonzero-sum games (NSGs) of multi-controller systems with general nonlinear dynamics, i.e., nonaffine dynamics, are an important research topic in academia and industry nowadays. Aiming at NSGs of multi-controller systems with unknown nonaffine dynamics, this paper develops a novel learning-based model-free NSG (MNSG) method using dynamic parallel control and adaptive dynamic programming (ADP). In the meantime, the developed learning-based MNSG method involves neither using a model network to predict unknown dynamics nor off-policy reinforcement learning (RL), contributing to a new and practical paradigm for MNSGs of multi-controller systems.}, archive = {J_TASE}, author = {Jingwei Lu and Qinglai Wei and Lefei Li}, doi = {10.1109/TASE.2025.3593929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19317--19327}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based parallel control for unknown nonaffine nonzero-sum games}, volume = {22}, year = {2025}, }
• (2025). Observer-based resilient adaptive event-triggered control for islanded microgrids under DoSAs and unknown FDI attacks. TASE, 22, 19307--19316. (www)

  @article{ , abstract = {This study investigates the resilient event-triggered load frequency control (LFC) issue of islanded microgrids (MGs) subject to denial of service attacks (DoSAs) and unknown false data injection attacks (FDIAs) simultaneously. An augmented extended observer is established to estimate both unavailable system states and unknown FDIA signals. A resilient adaptive event-triggered (AET) scheme, incorporating DoSA detection, FDIA estimation, and dynamic threshold adjustment, is proposed to reduce network resource consumption while countering cyberattacks. Some sufficient criteria are derived to ensure the asymptotic stability of the augmented estimation error system with $H_{\infty } $ performance. Additionally, the uniformly ultimate boundedness of the system is also guaranteed even under hybrid cyberattacks. Finally, simulation studies are used to illustrate the effectiveness of the proposed method. Note to Practitioners—This study addresses the network-based load frequency control problem in isolated microgrids. Since microgrids typically rely on open communication networks to transmit control commands, they are vulnerable to cyberattacks, such as injected malicious signals or overwhelming request data. These attacks can result in significant frequency fluctuations. To tackle these challenges, this paper proposes an augmented extended observer capable of estimating both unknown injected signals and compromised system states. Furthermore, a resilient event-triggered control strategy is developed, utilizing the TCP protocol and estimated signals to counteract the effects of cyberattacks. This approach not only stabilizes frequency deviations under hybrid attack scenarios but also effectively reduces communication load. Notably, while this study focuses on load frequency control in isolated microgrids, the proposed resilient control algorithm is adaptable to other types of microgrids.}, archive = {J_TASE}, author = {Yajuan Liu and Dong Xu and Sangmoon Lee and Xiaoning Zhang}, doi = {10.1109/TASE.2025.3593376}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19307--19316}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based resilient adaptive event-triggered control for islanded microgrids under DoSAs and unknown FDI attacks}, volume = {22}, year = {2025}, }
• (2025). Nonlinear tracking differentiator-based prescribed performance zeta-backstepping control with its application to a quad-rotor hover. TASE, 22, 19294--19306. (www)

  @article{ , abstract = {This paper proposes a nonlinear tracking differentiator-based prescribed performance zeta-backstepping control method for uncertain nonlinear systems with unmodeled dynamics and external disturbances. Unlike traditional prescribed performance control methods, the proposed approach allows direct adjustment of the system damping ratio through parameter regulation while ensuring that the tracking error remains within prescribed performance boundaries. To address unmodeled dynamics and external disturbances, the developed feedforward-compensated nonlinear tracking differentiator provides uncertainty estimation. The analysis of a second-order linear differential equation inequality rigorously proves that the closed-loop system achieves adjustable damping characteristics and guaranteed error performance. Finally, this control method is applied to a quad-rotor hover system, and experimental results confirm its effectiveness and advantages. Note to Practitioners—The damping ratio of a system is critically important in engineering applications, as it serves as a key parameter for balancing the convergence rate and overshoot. On the other hand, the prescribed performance control (PPC) allows for setting the tracking error boundary and ensures that the tracking error remains within the prescribed range. Inspired by engineering applications, the proposed control method combines an adjustable damping ratio with the prescribed performance control, enabling the damping ratio to be systematically designed via specific parameter selection rules while maintaining the tracking error within the prescribed performance boundary. As a result, both the transient and steady-state responses of the system can be precisely regulated under the proposed controller, which is highly beneficial for practical engineering applications. Specific applications include: automotive suspension control systems (optimizing driving stability and comfort), aircraft attitude control systems (enhancing dynamic stability and flight safety), and robotic control systems (improving motion accuracy and ensuring successful task completion).}, archive = {J_TASE}, author = {Jiaxing Yang and Xuebo Yang and Xiaolong Zheng}, doi = {10.1109/TASE.2025.3591880}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19294--19306}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nonlinear tracking differentiator-based prescribed performance zeta-backstepping control with its application to a quad-rotor hover}, volume = {22}, year = {2025}, }
• (2025). Weighted feature fusion of CNNs utilizing magnitude and phase information for chicken health detection. TASE, 22, 19279--19293. (www)

  @article{ , abstract = {Recent studies have shown the potential of phase features in audio processing, providing distinct insights into signal structure beyond what magnitude features can capture. In chicken health detection, however, existing models often rely heavily on magnitude-based features, neglecting phase-related information that could enhance detection performance. This paper introduces a Weighted Feature Fusion of Convolutional Neural Networks (WFF-CNNs), which integrate magnitude features, namely Mel Frequency Cepstral Coefficient (MFCC), with phase features, specifically Modified Group Delay Cepstral Coefficient (MGDCC), to enhance the identification of chicken-health issues. By integrating phase information, the proposed WFF-CNNs address an acknowledged limitations in current single-stream approaches that predominantly utilize magnitude-only features, hence augmenting the capacity of the model for capturing acoustic signals in chicken vocalizations. Compared with conventional CNNs using only MFCC, our approach overcomes the benchmark challenge of insufficient phase-based feature representation, leading to more robust detection of chicken health states. The efficacy of the proposed method is validated using audio signals from a publicly available database. Experimental results demonstrate that CNNs using MGDCC outperform conventional CNNs using MFCC, highlighting the benefits of using phase information. Furthermore, the WFF-CNNs utilizing MFCC and MGDCC outperform single-feature CNNs and existing standard models, confirming the efficacy of integrating both magnitude and phase information for accurate chicken health assessment. We anticipate that these findings will be useful in the formulation of more complete guidelines for monitoring poultry health, considering both size and phase characteristics. Note to Practitioners—Poultry farming represents a crucial sector in global food security and economics, with significant impact on both large-scale industrial operations and small-scale farmers. Disease outbreaks in chicken populations can lead to devastating economic losses - not just from chicken mortality, but also from decreased productivity, increased treatment costs, and potential trade restrictions. For instance, a single outbreak of avian influenza can result in the loss of entire flocks and disrupt local economies, particularly affecting small-scale farmers who may lack advanced monitoring systems. Our research addresses this critical need by developing an automated health monitoring approach that analyzes chicken vocalizations to detect potential health issues before they become severe outbreaks. Using advanced sound analysis techniques that examine multiple aspects of chicken vocalizations simultaneously, our method achieves 96.92% accuracy in health detection, surpassing traditional analysis methods. This analysis approach has the potential to be developed into an automated detection system suitable for small and medium-sized farms, requiring only basic microphones installed in chicken coops or farming areas, making it a cost-effective and easily implementable solution. However, system performance may be affected by environmental noise in real-world conditions. Future developments will focus on improving robustness across different farm environments and testing with various chicken breeds to validate real-world effectiveness. This development could enhance sustainable healthcare standards in poultry farming.}, archive = {J_TASE}, author = {Wongsathon Pathonsuwan and Khomdet Phapatanaburi and Khwanjit Orkweha and Talit Jumphoo and Patikorn Anchuen and Kasidit Kokkhunthod and Atcharawan Rattanasak and Monthippa Uthansakul and Peerapong Uthansakul}, doi = {10.1109/TASE.2025.3591801}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19279--19293}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Weighted feature fusion of CNNs utilizing magnitude and phase information for chicken health detection}, volume = {22}, year = {2025}, }
• (2025). Iterative learning predictive control method based on reinforcement learning for rubber mixing batch process. TASE, 22, 19266--19278. (www)

  @article{ , abstract = {As a typical batch production process, the rubber mixing process involves complex dynamics and significant fluctuations in results. Since the mixing temperature is key to product quality, this paper proposes an iterative learning model predictive control (ILMPC) strategy that integrates reinforcement learning (RL) for the temperature control model, utilizing the principle of reaction heat balance. The Iterative Learning Control (ILC) enhances performance across repetitive batches. Simultaneously, Model Predictive Control (MPC), leveraging its predictive model, compensates for external disturbances, ensuring primary tracking performance. To address the model accuracy dependence of ILMPC, RL is integrated to enhance the system’s robustness against non-repetitive disturbances and uncertainties through data-driven adaptive learning, optimizing overall performance. Finally, simulation experiments on a rubber mixing temperature control process demonstrate the superior control performance of the proposed method. Note to Practitioners—Precise temperature control in rubber mixing processes remains a significant challenge due to inherent nonlinearity, time delays, and disturbances. This paper presents a method aimed at improving control in repetitive industrial processes by addressing these issues through an adaptive learning mechanism. By leveraging data from previous batches, the proposed approach enhances performance and consistency while effectively handling unexpected disturbances. However, practical implementation may face challenges such as computational complexity, sensor limitations, and variations in material properties. Future research could focus on optimizing the algorithm for real-time applications, improving robustness to external uncertainties, and extending the approach to other batch-based manufacturing processes requiring precise thermal regulation.}, archive = {J_TASE}, author = {Wenmin He and Ying Pan and Cuimei Bo}, doi = {10.1109/TASE.2025.3591540}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19266--19278}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Iterative learning predictive control method based on reinforcement learning for rubber mixing batch process}, volume = {22}, year = {2025}, }
• (2025). Oscillation compensation in on-the-fly laser processing via optical tracking sensors. TASE, 22, 19254--19265. (www)

  @article{ , abstract = {In today’s industrial solar cell production, laser processing is a standard technology and will remain important for future developments, such as silver-free solar cell concepts. High throughput laser processing of more than $10.000\,\text {wafers per hour}$ with a high precision of 10 to $100\, {\mathrm {\mu \text {m} }}$ is required. Current laser systems use complex automation and transport systems for accurate positioning. However, these systems are costly, large, and have limited throughput. On-the-fly laser processing, with wafers processed on a conveyor belt, offers a compact and cost-effective high-throughput approach. However, its accuracy is limited by transport imprecision. To overcome this, we propose a novel system using real-time measurement of speed fluctuations to adjust the beam position via feed-forward control to achieve both high throughput and accuracy. An FPGA-based array of optical tracking sensors was developed and used in an on-the-fly demonstrator setup. The sensor’s accuracy is characterized and demonstrated in a one-dimensional oscillation compensation. Oscillatory deviations of the laser markings are reduced from 82.2 to $13.2\, {\mathrm {\mu \text {m} }}$ ( $1\,\sigma $ ), a six-fold improvement. Thereby, sampling delay limits the control bandwidth. Offline analysis shows that Kalman filtering could compensate for this, and still reduce the estimation uncertainty by $23\,\%$ . Note to Practitioners—We have developed a compact and robust laser processing system that processes solar cells while they are moving. Instead of depending on high-precision transport systems, our system actively controls the laser beam’s position to correct for any speed fluctuations of the moving solar cells. The prototype created uses an array of optical tracking sensors that measure the speed of the solar cell on its surface. This speed information is used to generate an encoder signal for the laser scanner, which then adjusts the laser beam’s position in real-time. Advantageous are the high sampling rate of the sensors and the low latency in processing the measurements. This approach could also be adapted for other applications involving high throughput laser processing.}, archive = {J_TASE}, author = {Georg Hoppe and Jale Schneider and Jan Frederik Nekarda and Ralf Preu and Fabian Meyer and Moritz Diehl}, doi = {10.1109/TASE.2025.3589129}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19254--19265}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Oscillation compensation in on-the-fly laser processing via optical tracking sensors}, volume = {22}, year = {2025}, }
• (2025). Uncertainty-aware parameter optimization for reliable laser powder bed fusion additive manufacturing. TASE, 22, 19238--19253. (www)

  @article{ , abstract = {Laser powder bed fusion (LPBF) is an additive manufacturing process capable of producing intricate structures with high accuracy. Despite this capability, it struggles to achieve the required reliability for mass production—specifically the stability of a production run and repeatability across multiple runs. Parameter optimization, which adjusts process parameters to regulate a specific quantity of interest (QoI), is a crucial means of quality control. Existing methods, however, have not adequately addressed both random and systematic factors in the LPBF process. The stochastic nature of the process is often neglected under the assumption that identical parameter inputs will consistently yield the same QoI. This deviates from reality and is not intended to reduce potential variations in the QoI. Moreover, many studies do not incorporate the systematic neighboring effects between scan tracks into their optimization, so process reliability cannot be guaranteed. To address this issue, this study focuses on optimizing the probability distribution of the QoI. The key idea is not only to increase the likelihood of achieving the ideal QoI but also to reduce its variance. This is achieved by uncertainty-aware modeling and optimization of the LPBF process using machine learning. Specifically, the problem is formulated as maximizing the posterior distribution of scan parameters given an ideal QoI sequence and historical manufacturing data, yielding a large-scale constrained optimization problem. A stochastic, distributed, gradient-based method is proposed to solve this problem, where a coarse-to-fine strategy plays a critical role in accelerating convergence. A case study is then conducted to stabilize the melt pool volume by optimizing laser powers. The solutions are verified in a calibrated finite element-based simulation environment, in which the variations of the melt pool volume are effectively reduced both within a single run and across multiple runs. The implementation of our method is available at https://github.com/qihangGH/uncertainty_aware_param_optim_for_AM. Note to Practitioners—This paper is motivated by the critical reliability issues in laser powder bed fusion (LPBF) additive manufacturing, requiring that the process is stable within a run and repeatable across runs. Parameter optimization is an effective way to control a certain quantity of interest (QoI) to regulate the LPBF process. Existing methods, however, often ignore the stochastic variations and neighboring effects in the LPBF process. The objective solely focuses on minimizing the error between the predicted QoI and ideal one, under the assumption that the environment is deterministic. The optimized parameters, therefore, do not guarantee reliability under the influences of random factors. This paper intends to improve the reliability through uncertainty-aware modeling and optimization of the LPBF process. It focuses on optimizing the probability distribution of the QoI, encouraging the ideal QoI not only more likely to appear but also appear with less fluctuations. By incorporating important neighboring effects and bounded and smooth input constraints, a large-scale constrained optimization problem is formulated. To solve this problem efficiently, a stochastic and distributed gradient-based optimization algorithm is proposed, and a coarse-to-fine strategy is employed to accelerate the optimization process. For validation, a case study is conducted to stabilize the melt pool volume by regulating laser powers. The optimized laser powers are verified in a finite element-based simulation environment. Quantitative and qualitative results show that the variations of the melt pool volume within a single run and across repeated runs are reduced, indicating the potential of the uncertainty-aware optimization for enhancing the reliability of the LPBF process.}, archive = {J_TASE}, author = {Qihang Fang and Gang Xiong and Fang Wang and Zhen Shen and Xisong Dong and Fei-Yue Wang}, doi = {10.1109/TASE.2025.3585575}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19238--19253}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Uncertainty-aware parameter optimization for reliable laser powder bed fusion additive manufacturing}, volume = {22}, year = {2025}, }
• (2025). Positive edge-consensus of uncertain fractional-order networked systems. TASE, 22, 19229--19237. (www)

  @article{ , abstract = {This paper addresses the edge-consensus problem in a directed nodal network, where each edge is described by uncertain fractional-order dynamics subject to positivity constraints. To solve the issues of positivity and consensus in uncertain fractional-order networked systems (UFONSs) with the order $0 \lt q \leq 1$ , a distributed control protocol by virtue of state feedback is designed. Some sufficient criteria are established to ensure positive edge-consensus of UFONSs. Subsequently, to handle scenarios with unmeasurable states in practical systems, an observer-based control protocol is proposed to achieve positive edge-consensus for UFONSs. Furthermore, by leveraging the fractional-order stability theory and the properties of positive systems, some sufficient positive edge-consensus conditions that depend only on the number of edges and nodes are formulated instead of the global information of the network. Finally, two simulation examples are shown to verify the feasibility of the proposed protocols. Note to Practitioners—In intelligent transportation networks, the traffic flow is affected not only by individual road segments but also by the edge flow, that is, the transfer of flow between segments. Similarly, in social networks, the frequency or intensity of interactions between individuals determines the speed and extent of information dissemination. Therefore, edge states play a critical role in describing the evolution of networked dynamics, which must remain positive. Meanwhile, fractional-order models provide higher accuracy than traditional integer-order models in complex environment. To address these requirements, this paper develops an edge-consensus control strategy with positivity constraints to achieve consensus of UFONSs. Considering practical cases with unmeasurable edge states, an observer-based edge-consensus control strategy is proposed. By this strategy, the positive edge-consensus of UFONSs is realized.}, archive = {J_TASE}, author = {Yanyan Ye and Yanfang Rong and Yong Wang and Xia Wu and Housheng Su}, doi = {10.1109/TASE.2025.3592290}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19229--19237}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Positive edge-consensus of uncertain fractional-order networked systems}, volume = {22}, year = {2025}, }
• (2025). Deep recurrent stochastic configuration networks for modelling nonlinear dynamic systems. TASE, 22, 19215--19228. (www)

  @article{ , abstract = {Deep learning techniques have shown promise in various domains. However, traditional methods can only demonstrate their universal approximation capability from an existential perspective, failing to explicitly construct such networks. This limitation is pronounced in nonlinear dynamic modelling, especially when faced with continual changes and uncertainties in the process, making them struggle to achieve accurate predictions and responses. This paper proposes a novel deep reservoir computing framework, termed deep recurrent stochastic configuration network (DeepRSCN) for problem solving. DeepRSCNs are incrementally constructed, with all reservoir nodes directly linked to the final output. The input random weight matrix and a special structure of the random reservoir weight matrix are assigned in the light of a supervisory mechanism, ensuring the universal approximation property. The output weights are updated online using the projection algorithm to handle the unknown dynamics. Given a set of training samples, DeepRSCNs can quickly generate learning representations, which consist of random basis functions with cascaded input and readout weights. Experimental results over a time series prediction, a nonlinear system identification problem, and an industrial data predictive analysis demonstrate that the proposed DeepRSCN outperforms other state-of-the-art methods with sound learning and generalization performance. Note to Practitioners—The motivation of this paper is to develop an advanced deep learning model with universal approximation capabilities for modelling nonlinear dynamic systems. While traditional methods can demonstrate infinite approximation of the target mapping from an existential perspective, they fail to explicitly construct the corresponding network. Furthermore, these approaches often encounter challenges, including unknown dynamic orders, arbitrary structure determination, and sensitivity to learning parameters, which can significantly impact the accuracy and reliability of industrial process analysis. To overcome these issues, the proposed DeepRSCNs are incrementally built using a supervisory mechanism, enhancing feature representation and maintaining universal approximation. Furthermore, DeepRSCNs can dynamically adjust the network parameters in response to real-time data, enabling them to adapt to changing operational conditions. These features are especially valuable in industrial environments, where capturing complex dynamics accurately and efficiently is essential for improving prediction performance and supporting better decision-making.}, archive = {J_TASE}, author = {Gang Dang and Dianhui Wang}, doi = {10.1109/TASE.2025.3592182}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19215--19228}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep recurrent stochastic configuration networks for modelling nonlinear dynamic systems}, volume = {22}, year = {2025}, }
• (2025). CrackSight: An efficient crack segmentation model in varying acquisition ranges and complex backgrounds. TASE, 22, 19197--19214. (www)

  @article{ , abstract = {Accurate crack segmentation in concrete transportation infrastructures is critical for ensuring structural integrity and facilitating timely maintenance interventions. This paper presents CrackSight, an end-to-end deep learning model for precise crack segmentation across varying observational ranges and extremely complex backgrounds. CrackSight seamlessly integrates crack detection and segmentation through two branches. The Detection Feature Extraction Branch (DFEB) provides global context for crack localization in complex backgrounds or at far observation ranges. It guides the segmentation model to focus on regions with the highest crack-prone potential. The segmentation branch leverages the fusion of multi-scale feature maps using dilated convolutions, allowing to capture subtle and complex crack patterns. The branch also incorporates the Dual-Attention Linear Focus Mechanism (DALFM) enhancing crack segmentation through saliency-driven improvements. Finally, CrackSight uses a novel hybrid contextual loss, which dynamically compensates for class imbalance and enhance crack discrimination against complex backgrounds. Our model is also lightweight and can be run in resource-constrained environments, making it suitable for real-world inspection using mobile platforms. Our results demonstrate that it significantly improves segmentation accuracy, setting a new benchmark for crack segmentation. The dataset and additional resources are available on GitHub. Note to Practitioners—CrackSight is a dual-branch deep learning framework designed for accurate and efficient segmentation of concrete cracks under challenging real-world conditions. By combining a detection-guided localization branch with a context-aware segmentation, CrackSight offers enhanced robustness to noise, background clutter, and varying acquisition distances, common challenges in UAV-based infrastructure inspections. Its architecture integrates multi-scale feature fusion and adaptive contextual guidance, enabling reliable detection of both fine and fragmented cracks. With its lightweight design and fast inference time, CrackSight offers practitioners a practical and scalable solution for automating visual inspection tasks, reducing manual effort, and improving safety in structural health monitoring workflows.}, archive = {J_TASE}, author = {Dariush Amirkhani and Mohand Saïd Allili and Jean-François Lapointe}, doi = {10.1109/TASE.2025.3591407}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19197--19214}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {CrackSight: An efficient crack segmentation model in varying acquisition ranges and complex backgrounds}, volume = {22}, year = {2025}, }
• (2025). Data-driven optimal output feedback control of unknown system model via adaptive dynamic programming. TASE, 22, 19187--19196. (www)

  @article{ , abstract = {This paper investigates the linear quadratic optimal output feedback control problem for an unknown linear continuous-time system. Combined with adaptive dynamic programming and optimal control theory, an online data-driven iteration learning algorithm is developed to learn an optimal controller from system data. The main advantage of the proposed algorithm is that it does not require an initial stabilizing control policy, a full-rank condition, or historical data storage to guarantee algorithm convergence. This is fundamentally different from the existing results based on the least-squares method, which requires these conditions. Moreover, the developed algorithm uses only the input and output data of the system, which solves the problem of unmeasurable system states. The simulation results demonstrate the efficacy of the proposed algorithm, and its superiority is demonstrated by comparison with the existing algorithms. Note to Practitioners—In the actual control systems, stability is only the basic condition that should be met. Besides, the control systems usually need to optimize certain indices to achieve the optimal control performance. Therefore, there is a practical research demand for the optimal control problem of the system. The traditional optimal control method requires the knowledge of the accurate system model, which is difficult to establish in actual control system. To this end, combined with the adaptive dynamic programming and optimal control theory, this paper proposes a data-based iteration learning algorithm to adaptively learn an optimal output feedback control policy from system data. Simulation results related to autonomous ground vehicle underscore the effectiveness of our proposed algorithm, and its superiority is further demonstrated through comparisons with existing methodologies.}, archive = {J_TASE}, author = {Yong-Sheng Ma and Jian Sun and Yong Xu and Shi-Sheng Cui}, doi = {10.1109/TASE.2025.3593481}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19187--19196}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven optimal output feedback control of unknown system model via adaptive dynamic programming}, volume = {22}, year = {2025}, }
• (2025). Resilient control strategy for smart grid under delayed sporadic measurements: Attack estimation via hybrid observer. TASE, 22, 19175--19186. (www)

  @article{ , abstract = {This paper investigates the safe control problem for smart grid and proposes a resilient control scheme via attack estimation, in which both the sampled output signals and command signals may be corrupted by false data injection attacks. Meanwhile, the communication channels between power plant and controller suffer from time-varying delay. Firstly, the i-th power area is modeled as a hybrid system model composed of continuous system signals and discrete attack signals. And then, a hybrid observer is constructed for each area to estimate the attacks, by which a modified load frequency control scheme is designed to effectively attenuate the effect of attacks on the actuator signals. In particular, the observer gains for each area are obtained via only local information, which can avoid privacy leakage and reduce computational complexity. Moreover, an optimization algorithm is proposed for each area to search a sub-optimal observer gain. Finally, a three-area power system is utilized to verify the proposed resilient control scheme. Note to Practitioners—It is challenging to realize frequency regulation of power systems via communication networks that are vulnerable to cyber-attacks. Conventional robust control methods from a passive perspective are inefficient against prolonged and intense attacks, thus, this paper proposes an active defense scheme based on attack estimation, in which the attacks are estimated via the designed hybrid observer. Using only delayed sporadic measurements, the developed resilient control scheme can solve the practical problems in existing works from following aspects: 1) Continuous output signals are no longer required for attack estimation, which is of significant importance because frequency control in practical cyber-physical systems typically employs sampled-data control; 2) Model discretization is no longer needed, a discretized model obtained with a relatively large sampling period may lack sufficient accuracy; 3) The attack signals in both measurement and controller channel can be estimated, and the inevitable transmission delay is considered; 4) Centralized estimation is avoided, the designed hybrid observer works in a distributed way, making it more scalable. The above merits make the proposed resilient control strategy more practical to be adopted for improving the resilience in cyber-physical power systems against false data injection attacks.}, archive = {J_TASE}, author = {Kun Xu and Yugang Niu and Zhina Zhang}, doi = {10.1109/TASE.2025.3593480}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19175--19186}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient control strategy for smart grid under delayed sporadic measurements: Attack estimation via hybrid observer}, volume = {22}, year = {2025}, }
• (2025). USDE-based fast sliding mode control for magnetostrictive actuated positioning systems with hysteresis. TASE, 22, 19163--19174. (www)

  @article{ , abstract = {The inherent hysteresis nonlinearity in the magnetostrictive actuators (MAs) restricts their application in high-precision positioning systems. While direct inverse compensation has been widely used to mitigate hysteresis effects, it suffers from complexities in computing the inverse model, and addressing the residual modeling and compensation errors. In this respect, this paper introduces a composite control strategy to enhance the control performance of magnetostrictive actuated positioning system (MAPS) by presenting a new inverse-based compensator for the hysteresis. The key innovation lies in the derivation of an analytical expression for the inverse compensation error based on a generalized Prandtl-Ishlinskii (GPI) model, and the design of an unknown system dynamics estimator (USDE) to estimate this residual error. The USDE has a simple structure and easy parameter tuning, and can be trivially integrated into the closed-loop control system. Furthermore, a fast reaching law (FRL) is developed to design a composite sliding mode control (SMC), which uses both the inverse-based compensator and the USDE. This control with the FRL can achieve both the enhanced control response and the reduced chattering, compared with the standard SMC techniques. Finally, the effectiveness of the proposed approaches is confirmed by comparative simulations and experiments. Note to Practitioners—The accurate control for magnetostrictive actuated positioning system (MAPS) is challenging due to its inherent hysteresis nonlinearity. The traditional approach to mitigate the hysteresis effects is to construct a complex inverse model of hysteresis and then apply the direct inverse model compensation in the control, but this approach suffers from significant compensation error and imposes nontrivial modeling efforts. In this work, we adopt a generalized hysteresis model to form a model based inverse compensation and then derive an analytical expression of the compensation error. Furthermore, an unknown system dynamics estimator(USDE) with a simple structure is used to estimate the compensation error, where the estimated results are incorporated into the control to improve the control performance. Additionally, we introduce a novel sliding mode control that simultaneously achieves fast response and chattering suppression compared to conventional ones. It uses the inverse compensator and the USDE to develop a composite control structure. Compared to previous works, the proposed control scheme is easy to implement and can achieve superior results, as verified by simulation and experimental results.}, archive = {J_TASE}, author = {Shengbin Wu and Jing Na and Xinkai Chen and Yingbo Huang and Guanbin Gao}, doi = {10.1109/TASE.2025.3591702}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19163--19174}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {USDE-based fast sliding mode control for magnetostrictive actuated positioning systems with hysteresis}, volume = {22}, year = {2025}, }
• (2025). Vertical-plane locomotion control of a high-speed robotic tuna via NMPC. TASE, 22, 19151--19162. (www)

  @article{ , abstract = {The development of bionic underwater robots has brought new vitality to ocean exploration. Motion control is crucial for the stability of underwater robots due to significant differences in flow field characteristics at various swimming speeds. This study focuses on vertical-plane motion and proposes a model predictive control method to achieve integrated control of depth position and pitch attitude for bionic robotic fish. First, based on a robotic tuna system, high-maneuverability vertical-plane motion configuration elements are analyzed and summarized, laying the foundation for motion stability and controllability. Second, through hydrodynamic sampling in aquatic environments, a system model covering the range of swimming speeds is established. Regarding the control method, the proposed motion planning approach converts the desired motion sequence into an equivalent “pitch-depth” trajectory curve. A nonlinear model predictive controller (NMPC) is then designed to track the trajectory curve, ultimately achieving the desired vertical-plane motion. Experimental results validate that the proposed method not only ensures control accuracy under both low and high-speed conditions, but also enables the execution of complex motion sequence control. This study provides a fresh perspective on the motion instability analysis of robotic fish at high swimming speed and a novel control framework for regulating continuous posture sequences in the vertical plane. Note to Practitioners—The motivation of this paper is to address the challenges associated with stable motion and control of robotic fish in the vertical plane, given the variability of flow field characteristics at different swimming speeds. Existing methods for controlling pitch attitude and depth in bionic underwater robots are typically designed for stable flow conditions encountered during low-speed swimming. However, the instability and agility of high-speed robotic fish movements have not been adequately considered. Additionally, the coupling between pitch attitude and depth poses challenges for joint control of their combined states. This paper proposes a configuration analysis and control methodology to achieve desired vertical-plane locomotion for robotic fish. Specifically, using a robotic tuna as the research subject, a configuration analysis method for high-maneuverability motion in the vertical plane is presented, providing a foundation for ensuring motion stability and controllability. To accurately evaluate the motion of robotic fish under varying flow speeds, a system model for vertical plane motion is constructed based on hydrodynamic data collected from aquatic environments. A motion planning approach is proposed to convert desired vertical plane motion sequences into controllable “pitch-depth” trajectory curves, and a nonlinear model predictive controller is designed to track these trajectories. Configuration simulations and control experiments validate the effectiveness of the proposed method. Hopefully, our proposed methods can provide valuable insights and support for high-maneuverability motion control and continuous posture sequences tracking of bionic underwater robots in the vertical plane.}, archive = {J_TASE}, author = {Ru Tong and Sijie Li and Di Chen and Zhengxing Wu and Junzhi Yu}, doi = {10.1109/TASE.2025.3592696}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19151--19162}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Vertical-plane locomotion control of a high-speed robotic tuna via NMPC}, volume = {22}, year = {2025}, }
• (2025). Transformer-driven semantic-spatial adaptive fusion representation for object-goal navigation. TASE, 22, 19135--19150. (www)

  @article{ , abstract = {Visual object-goal navigation requires an agent to make decisions to search for specified target objects within an unknown environment. While learning-based approaches have achieved progress, they still face two limitations: 1) visual representation: the lack of a low-dimensional representation that balances object priors with current observations, and the absence of effective modeling of scene spatial structures, which restricts the agent’s ability to perceive its surroundings. 2) Policy learning: existing reward functions often fail to consider the visual essence of object goal-driven tasks, leading to inefficient learning and suboptimal performance. To address these issues, this paper introduces a semantic-spatial fusion representation framework that incorporates both object semantics and the intrinsic spatial structure of the scene. Specifically, a new object context matrix captures semantic relationships between objects while providing distinct low-dimensional representations for different observations, and an episode memory graph is also constructed and dynamically updated based on observation similarity and geodesic distance to represent the spatial structure of the agent’s environment in real-time. Then, a transformer-based adaptive multimodal feature fusion module is proposed to integrate these dual representations. Moreover, a sparse reward function is designed based on the target’s bounding box to guide the agent to learn correctly. The proposed method is the first to be evaluated on both the simulation platform AI2-THOR and the real-world dataset AVD, demonstrating its generalization in unseen environments. The method is also deployed on a physical mobile robot and tested in real-world scenarios, further validating its practical effectiveness. Note to Practitioners—This work aims to enhance the visual object-goal navigation capabilities of mobile robots in unknown environments. To achieve this, we propose a dual visual representation that incorporates an enhanced object context matrix and an episode memory graph, enabling an agent to recognize object semantics in the current view while perceiving the spatial structure of its surroundings. These features are adaptively fused using a multimodal feature fusion module based on a transformer with a variant of attention mechanism. Additionally, a sparse reward function is designed based on the bounding box of the target object in the current observation to better guide the model training. Extensive experiments demonstrate that, benefiting from the unique visual representation and the newly designed sparse reward, the proposed method allows agents to locate target objects more quickly and accurately. It also enables agents to terminate tasks autonomously at a more appropriate position, ensuring closer proximity to the target and optimal observation views. Throughout the navigation process, the method can flexibly avoid obstacles in its path, particularly when entering narrow spaces, and prevents task failure due to the lack of perception of the surrounding structure. However, our experiments are currently limited to small-scale home environments. Future research will extend the testing to more complex and diverse scenarios, such as restaurants, hospitals, and factories, to further evaluate the method’s robustness and adaptability in real-world applications.}, archive = {J_TASE}, author = {Ning Yang and Fei Lu and Xiaolei Li and Guohui Tian and Zhongyang Li and Tengfan Fu}, doi = {10.1109/TASE.2025.3593344}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19135--19150}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transformer-driven semantic-spatial adaptive fusion representation for object-goal navigation}, volume = {22}, year = {2025}, }
• (2025). Wind turbine blade icing predictive fault warning system based on conditional time series denoising diffusion data restoration model. TASE, 22, 19122--19134. (www)

  @article{ , abstract = {Existing warning methods are hindered by the ability to extract subtle early-stage blade icing features from wind turbine (WT) monitoring data, resulting in delayed or unreliable alerts. To address this challenge, this paper proposes a novel diffusion-based normal WT behavior model, Conditional Time Series Denoising Diffusion (CTSDD), and develops an automated predictive fault warning system accordingly. CTSDD restores normal monitoring data from noise, which encodes conditional information generated by a Hybrid Masking strategy to guide the restoration process. Its conditional encoding and noise prediction networks adopt a Transformer-like symmetric structure, integrating Dual-Attention and bidirectional long short-term memory (BiLSTM) modules to enhance feature extraction for multidimensional time series data. During the early stages of blade icing, deviations from learned normal patterns result in a noticeable decline in wind power (WP) restoration accuracy. Leveraging these deviations, this study proposes a composite scoring algorithm and a thresholding scheme to evaluate WT status and enable timely early warnings. Comparative experiments validate CTSDD’s superior restoration performance, with its system triggering the earliest and most stable warnings—approximately five hours before fault escalation. Ablation studies further confirm the contributions of Dual-Attention and BiLSTM modules. The proposed system offers an intelligent and effective solution to enhance the safety and reliability of WT operations. Note to Practitioners—WT blade icing presents a significant challenge for maintenance, particularly in its early stages, where the subtle impact often goes unnoticed, rendering traditional data analysis ineffective for detection. If left unaddressed, icing progressively degrades WT performance, reducing energy efficiency and increasing operational costs. Therefore, this study introduces an innovative icing predictive warning system based on an advanced deep-learning diffusion model, designed to detect early-stage icing indicators with greater sensitivity, ensuring reliable WT operation and economic efficiency. At its core, the deep learning model is dedicated to learning the typical patterns within normal WT monitoring data. Importantly, it can identify minute deviations in early-stage icing monitoring data from established normal patterns. Leveraging these discrepancies, we develop a WT status scoring algorithm and a calibrated warning threshold, enabling precise and timely early warnings for blade icing. In practice, this system autonomously analyzes monitoring data in real time, promptly issuing accurate icing warnings. This empowers operators to take proactive measures before the fault escalates, enhancing the reliability and intelligence of WT fault management. Future work will explore optimizing data utilization by assessing the value of different monitoring data sources.}, archive = {J_TASE}, author = {Qianyue Wang and Gangquan Si and Zihan Shan and Detao Fan and Bo Li and Xin Wang}, doi = {10.1109/TASE.2025.3593296}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19122--19134}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Wind turbine blade icing predictive fault warning system based on conditional time series denoising diffusion data restoration model}, volume = {22}, year = {2025}, }
• (2025). Multi-UAV dynamic task assignment based on event-triggered graph reinforcement learning under weak communication. TASE, 22, 19106--19121. (www)

  @article{ , abstract = {This paper addresses the dynamic task assignment problem for multiple uncrewed aerial vehicles (UAVs) operating under weak communication. Existing learning-based methods face two primary challenges: limited scene generalization and excessive reliance on communication resources. To address these issues, this paper proposes an event-triggered reinforcement learning algorithm based on graph neural networks. First, heterogeneous UAVs and tasks are embedded into a graph to construct a relationship model, which clearly represents the complex constraints between UAVs and tasks. This graph structure overcomes the limitations of existing methods in capturing constraint relationships. Second, the incorporation of heterogeneous graph neural networks and adaptive attention mechanisms enables effective learning of changes in adjacent node information, allowing the model to capture complex constraint relationships and environmental dynamics. This approach also addresses the lack of sensitivity to environmental changes observed in existing methods. Lastly, a dynamic synchronization mechanism is employed to update task assignment statuses in real time, preventing task conflicts and ensuring efficient allocation. Experimental results demonstrate that this method strikes a better balance between task assignment quality and efficiency. It performs well in untrained scenarios and significantly reduces communication resource consumption. These results highlight its promising potential for application in weak communication environments. Note to Practitioners—This paper addresses the dynamic task assignment problem in multi-UAV systems under weak communication conditions. It proposes an event-triggered graph reinforcement learning algorithm combining graph neural networks and an adaptive attention mechanism. This framework models UAVs and tasks as nodes in a heterogeneous graph, capturing their relationships and constraints. The algorithm updates task assignments in real time using a dynamic synchronization mechanism, minimizing conflicts and ensuring optimal task completion despite intermittent communication. Its key advantage is balancing task quality and efficiency. Compared to traditional methods, it performs better in various scenarios, especially in untrained environments, and adapts to dynamic task demands. This approach is particularly useful in defense, logistics, agriculture, and surveillance. However, it has limitations in extreme cases of frequent UAV sabotage, and future work could optimize it for fully distributed environments to improve adaptability and reliability.}, archive = {J_TASE}, author = {Jin Yu and Ya Zhang and Changyin Sun}, doi = {10.1109/TASE.2025.3593125}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19106--19121}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-UAV dynamic task assignment based on event-triggered graph reinforcement learning under weak communication}, volume = {22}, year = {2025}, }
• (2025). MRA-TESO-based high-precision active magnetic compensation method. TASE, 22, 19095--19105. (www)

  @article{ , abstract = {Active magnetic compensation (AMC) is of significant importance in enhancing the signal-to-noise ratio of cardiac magnetic signals. However, in practical environments, multiple magnetic disturbances exist and need to be suppressed to ensure the accuracy of magnetocardiography (MCG) measurements. This paper proposes a model reference adaptive extended state observer with tracking filter (MRA-TESO) based high-precision active magnetic compensation method. It addresses the model errors in the AMC system and the insufficient capability of a typical extended state observer (ESO) in estimating complex disturbances. In the MRA-TESO framework, a tracking filter is employed to increase the system type of the equivalent open-loop transfer function for improved disturbance observation. Additionally, a model reference adaptive law is introduced to reduce internal system disturbances, providing more accurate model information and thereby lowering the bandwidth required for disturbance estimation. Experimental results demonstrate that the proposed method achieves satisfactory disturbance rejection performance. Note to Practitioners—In this paper, we focus on the issue of magnetic field disturbances in a magnetic shielding cylinder, which is a critical factor that limits the accuracy of MCG measurements. The disturbance suppression methods based on the typical ESO have inadequate accuracy in estimating complex disturbances. This paper proposes a high-precision disturbance suppression method based on MRA-TESO, which enhances the disturbance suppression capability. In future work, we will strive to create a stable environment for MCG measurements using this method, thus improving the accuracy of MCG measurements.}, archive = {J_TASE}, author = {Yuanbo Jin and Peiling Cui and Haifeng Zhang and Yanbin Li and Zhouqiang Yang}, doi = {10.1109/TASE.2025.3592993}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19095--19105}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MRA-TESO-based high-precision active magnetic compensation method}, volume = {22}, year = {2025}, }
• (2025). A safety-critical dynamic system framework for high-precision learning from demonstration. TASE, 22, 19081--19094. (www)

  @article{ , abstract = {Stable dynamic systems enable robotic systems to plan and execute complex geometric motions in unstructured environments. However, in certain scenarios, such as precision assembly tasks, the constrained operational workspace of robotic arms, along with the presence of obstacles, may lead to unintended collisions. Furthermore, motion precision plays a crucial role in determining the success rate of such tasks. To address these challenges, we propose a novel Safe-Critical Dynamic System (SC-DS) framework. The SC-DS framework consists of a stable dynamic system and an obstacle avoidance controller. The stable dynamic system is formulated in a parametric nonlinear form, which enhances performance in terms of accuracy. Additionally, control barrier functions (CBF), corresponding to complex constraint spaces, are learned from demonstration data. By utilizing these learned CBFs, an obstacle avoidance controller is designed to ensure that the system trajectory remains within the learned safety boundaries. Moreover, the controller adaptively extends the effective range of control inputs, thus mitigating replication errors due to input limitations. Experimental results, both in simulation and with a physical robot, demonstrate that the SC-DS framework effectively reproduces trajectories with both stability and safety, outperforming existing methods in terms of overall task performance. Note to Practitioners—This study is motivated by the need to develop a safer and more precise skill-learning framework for practical applications, such as service robots and assembly robots. We propose the SC-DS framework, which integrates the challenges of uncertain environments into the learning of precise motion skills, ensuring both accuracy and safety in trajectory generation. This framework is particularly suitable for applications requiring strict performance and safety standards. By incorporating safety constraints directly into the learning process, our method provides a robust solution that effectively addresses robot-environment interactions, including obstacles, disturbances, and varying conditions. Our research enhances the reliability of dynamic system-based learning frameworks and offers practical tools for real-world applications, ensuring robots perform tasks efficiently while maintaining safety. Practitioners can leverage this framework to improve safety and maintain high-precision skill learning, ensuring effective handling of real-world challenges.}, archive = {J_TASE}, author = {Haotian Huang and Jiayun Fu and Zhehao Jin and Andong Liu and Wen-An Zhang and Chenguang Yang}, doi = {10.1109/TASE.2025.3592739}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19081--19094}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A safety-critical dynamic system framework for high-precision learning from demonstration}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered consensus control of multi-agent systems with time-varying delays and semi-markovian switching topology. TASE, 22, 19069--19080. (www)

  @article{ , abstract = {In this paper, we address the consensus problem in multi-agent systems (MASs) under conditions where time-varying delays impact inter-agent transmissions and the communication topology changes according to semi-Markovian rules with partially unknown transition rates. We implement dynamic event-triggering mechanisms (DETMs) on both the sensor-to-observer (S-O) and controller-to-actuator (C-A) channels to minimize unnecessary data transmissions within the network, which involves utilizing locally triggered sampled data in a distributed manner to optimize resource efficiency. In this output-feedback design, each agent constructs distributed observers to predict its own and neighboring agents’ states. In the design phase, we convert the consensus control problem into an asymptotic stability problem. Employing the Lyapunov-Krasovskii approach, we formulate the event-triggering parameters to ensure the stability of the closed-loop system comprising all agents, thereby achieving consensus. Through numerical simulations and experiments, we demonstrate that our approach effectively balances reducing the frequency of inter-agent communication with ensuring that the agents reach consensus. Note to Practitioners—In many industrial applications of multi-agent systems (MASs), such as robotics and sensor networks, achieving coordinated control in the presence of time-varying communication delays and changing topologies is a significant challenge. This paper presents a solution using dynamic event-triggered mechanisms (DETMs) to reduce unnecessary data transmissions between agents, ensuring efficient resource utilization. Unlike traditional methods that rely on constant data flow, DETMs allow communication only when necessary, optimizing bandwidth usage without sacrificing performance. By employing a distributed observer-based approach, each agent estimates its own state and its neighbors’ states, overcoming time-varying delays. The method offers a promising solution for improving efficiency in MASs. Future research could extend this approach to handle larger networks, more complex topologies, and varying operational conditions, with potential applications in autonomous vehicles, drone fleets, and large-scale sensor networks.}, archive = {J_TASE}, author = {Milad Abbasi and Horacio J. Marquez}, doi = {10.1109/TASE.2025.3592691}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19069--19080}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered consensus control of multi-agent systems with time-varying delays and semi-markovian switching topology}, volume = {22}, year = {2025}, }
• (2025). MMD-OPT: Maximum mean discrepancy-based sample efficient collision risk minimization for autonomous driving. TASE, 22, 19051--19068. (www)

  @article{ , abstract = {We propose MMD-OPT: a sample-efficient approach for minimizing the risk of collision under arbitrary prediction distribution of the dynamic obstacles. MMD-OPT is based on embedding distribution in Reproducing Kernel Hilbert Space (RKHS) and the associated Maximum Mean Discrepancy (MMD). We show how these two concepts can be used to define a sample efficient surrogate for collision risk estimate. We perform extensive simulations to validate the effectiveness of MMD-OPT on both synthetic and real-world datasets. Importantly, we show that trajectory optimization with our MMD-based collision risk surrogate leads to safer trajectories at low sample regimes than popular alternatives based on Conditional Value at Risk (CVaR). Note to Practitioners—Autonomous Driving software stacks have dedicated modules for predicting trajectories of the obstacles(neighboring vehicles). Typically, these predictors provide a set of possible future motions for the obstacles, each of which can have different likelihoods of happening. Thus, a key challenge is to reason about collision risk in a given scene based on predicted trajectories, but without being overly conservative. For example, treating each predicted trajectory as a separate obstacle, without any attention to their likelihood, may not allow any feasible motion to the ego-vehicle. Our work addresses this challenge by proposing a probabilistic approach for modeling and minimizing collision risk. Our core impact lies in improving sample efficiency: that is assessing and minimizing collision risk based on just a handful of predicted trajectories for a given obstacle. Our approach can be easily integrated with any deep neural network based trajectory predictors which have become the de facto standard in autonomous driving industry. Our formulation easily extends to applications like indoor navigation with mobile robots, since human trajectory predictors have structural similarity with those deployed in autonomous driving. A practical limitation of our approach is that it requires additional computing power (in the form of GPU accelerators) to achieve real-time performance.}, archive = {J_TASE}, author = {Basant Sharma and Arun Kumar Singh}, doi = {10.1109/TASE.2025.3589941}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19051--19068}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MMD-OPT: Maximum mean discrepancy-based sample efficient collision risk minimization for autonomous driving}, volume = {22}, year = {2025}, }
• (2025). Bipartite output formation tracking control for discrete-time singular heterogenous multi-agent systems. TASE, 22, 19037--19050. (www)

  @article{ , abstract = {In this paper, the distributed bipartite output formation control problem of heterogeneous discrete-time singular multi-agent systems (MASs) is studied under the cooperative-competition network. Different from the classical formation control issue, the leader (also called exosystem) is also described by more general discrete-time singular dynamics. Based on the cooperative output regulation theory, the bipartite output formation with two opposing subgroups is realized by reconstructing bipartite tracking error and singular exosystem. To obtain the bipartite state of exosystem, a new exosystem observer associated with discrete-time singular algebraic Riccati equation (ARE) and signed directed topological matrix is designed. The observer coupling gains are introduced to enable the global error system to be included in the stability region of the unit circle. Then under the singular regulator equations and the separation principle, two classes of distributed observed-based controllers via state and output feedbacks are respectively presented for bipartite formation scenarios where the states of followers are known and unknown. Moreover, by redefining the expected relative dynamic positions, the proposed control strategies can be extended to achieve time-varying bipartite output formation. Finally, simulation results are given to verify the feasibility and effectiveness of the proposed algorithms. Note to Practitioners—Formation control of MASs plays a critical role in achieving effective coordination of distributed units, enhancing operational effectiveness, optimizing resource allocation, and improving overall production efficiency. This approach has broad applications across various industries, including transportation, energy, and both military and civilian applications, thereby fostering social and economic development. However, in practical applications such as autonomous aerial vehicle (AAV) clustering cooperative search and intelligent robot symmetrical formation, there exist several challenges: (1) the absence of stability analysis tools in discrete-time singular MASs; (2) the heterogeneity in dynamic characteristics; (3) the complex cooperative-competitive interactions among agents. This paper therefore studies the bipartite output formation control problem for heterogeneous discrete-time singular MASs. Specially, the autonomous exosystem with singular dynamics is seen as a leader and the corresponding new observer is introduced. By proposing distributed state and output feedback control protocols based on regulator equations, the followers maintain both fixed and time-varying relative positions with the exosystem. Furthermore, the proposed bipartite formation control strategies can be extended to address related bipartite consensus and formation problems.}, archive = {J_TASE}, author = {Jie Zhang and Yao Yao and Jian-An Wang and Xiaolei Li and Da-Wei Ding}, doi = {10.1109/TASE.2025.3584332}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19037--19050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bipartite output formation tracking control for discrete-time singular heterogenous multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). An end-to-end multi-dimensional perception network architecture for robotic grasp detection with target edge collision-aware strategy. TASE, 22, 19025--19036. (www)

  @article{ , abstract = {This paper investigates the feasibility of robotic grasping of various objects in complex scenarios, with our method aiming to achieve grasping capabilities for any scene and any object. Firstly, a Target Edge Collision-Avoidance Strategy is proposed that systematically incorporates the edge features of grasping objects. This strategy is specifically designed to address two critical challenges: bridging the significant performance gap between offline training data and real-world operating conditions, and effectively preventing collision incidents between the robotic end-effector and target objects during physical grasping operations. Furthermore, the Grasp Detection Network based on Global and Local Information Perception (GLIP-Net) is proposed, featuring two intricately designed components: the Global Information Perception and Local Aggregation Module, and the Multi-dimensional Multi-scale Attention and Adaptive Feature Fusion Module. The GLIP-Net enhances the network’s perceptiveness to global information, strengthening the correlation between features and the spatial parameters of grasping. To validate the effectiveness of the presented method, extensive tests and grasping experiments is conducted on the Cornell Dataset and Jacquard Dataset, as well as in practical scenarios. The empirical outcomes indicate an accuracy level of 99.2% on the Cornell Dataset and 96.8% on the Jacquard Dataset, respectively. Furthermore, by employing the Kinova robot in both single-object and multi-object complex scenarios within real-world environments, the grasping success rates of 97.0% and 95.8% is achieved. Note to Practitioners—This paper was inspired by the problem of robotic object grasping in various scenarios, but it is also applicable to tasks such as object grasping, sorting, and transportation in unstructured environments. Existing robotic grasping methods are typically limited to structured scenarios, where robots can only grasp objects at fixed positions. When the object or its pose changes, the entire grasping task is likely to fail. Additionally, robots often fail to consider the edge information of the object when grasping, leading to collisions between the end effector and the object. In this paper, we propose a novel approach that employs a graping detection network to process an input color image containing depth information. The neural network takes both global and local information into account, fuses useful feature data, and adaptively outputs a set of grasp configurations. To address the issue of collisions between the end effector and the object, we design a target edge collision-avoidance strategy, prioritizing regions adjacent to the gripper side. Preliminary experiments demonstrate the feasibility of our method, which has been tested in several real-life scenarios. In future research, we aim to address the robot vision closed-loop control problem, enabling robots to perform grasping tasks in dynamic environments.}, archive = {J_TASE}, author = {Hejia Gao and Junjie Zhao and Yang Yang and Changyin Sun}, doi = {10.1109/TASE.2025.3592755}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19025--19036}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An end-to-end multi-dimensional perception network architecture for robotic grasp detection with target edge collision-aware strategy}, volume = {22}, year = {2025}, }
• (2025). Toward fault tolerance in multi-agent reinforcement learning. TASE, 22, 19007--19024. (www)

  @article{ , abstract = {Agent faults pose a significant threat to the performance of multi-agent reinforcement learning (MARL) algorithms, introducing two key challenges. First, agents often struggle to extract critical information from the chaotic state space created by unexpected faults. Second, transitions recorded before and after faults in the replay buffer affect training unevenly, leading to a sample imbalance problem. To overcome these challenges, this paper enhances the fault tolerance of MARL by combining optimized model architecture with a tailored training data sampling strategy. Specifically, an attention mechanism is incorporated into the actor and critic networks to effectively and automatically detect fault information and dynamically regulate the attention given to faulty agents. Additionally, a prioritization mechanism is introduced to selectively sample transitions critical to current training needs. To further support research in this area, we design and open-source a highly decoupled code platform for fault-tolerant MARL, aimed at improving the efficiency of studying related problems. Experimental results demonstrate the effectiveness of our method in handling various types of faults, faults occurring in any agent, and faults arising at random times. Note to Practitioners—Multi-agent systems based on MARL outperform those using traditional control methods in terms of performance but remain highly vulnerable to unexpected faults. To improve fault tolerance in such systems, we introduce an attention mechanism that enables the neural network to dynamically adjust its focus on fault-related information. Additionally, a prioritization sampling strategy is employed to select critical samples from collected experiences that are most relevant to current training needs. Experimental results across various fault types demonstrate significant improvements in fault tolerance, validating the robustness of our approach. These findings suggest that the proposed method has the potential to be applied to real-world scenarios, such as multi-robot systems and autonomous vehicle fleets.}, archive = {J_TASE}, author = {Yuchen Shi and Huaxin Pei and Liang Feng and Yi Zhang and Danya Yao}, doi = {10.1109/TASE.2025.3592721}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {19007--19024}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward fault tolerance in multi-agent reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Predefined time anti-swing control of quadrotor UAV transportation systems via 3-D spatial decomposition. TASE, 22, 18995--19006. (www)

  @article{ , abstract = {Air transport technology is vital for emergency and transportation tasks, such as disaster relief and crop transportation. This paper proposes a predefined time controller (PTC) for the quadrotor uncrewed aerial vehicle (UAV) transportation system (QUTS) to achieve rapid trajectory tracking while suppressing payload swing under strong coupling and external disturbances. A dynamic model of the QUTS is developed using the Newton-Euler method, and a nonlinear cascade control scheme is established with three loops: an outer loop for position control, a middle loop for payload swing angle control, and an inner loop for attitude control. The boundedness and predefined-time convergence of the QUTS are proven using the Lyapunov method. Compared to the asymptotically stable controller, the fixed-time adaptive controller, and the no-swing angle controller, on one hand, unlike traditional controllers that exhibit longer and condition-dependent settling times, the PTC ensures predefined-time convergence, leading to rapid system stabilization. On the other hand, the PTC achieves higher trajectory tracking accuracy and superior anti-swing performance. Note to Practitioners—In critical missions such as emergency rescue, rapid response and precise control of aerial transportation system are crucial. Unexpected payload swing can lead to safety incidents, and external disturbances further challenge stability. Considering the underactuated nature of the QUTS, the swinging motion of the payload cannot be directly controlled. To address these issues, this paper presents a cascade control scheme featuring predefined time convergence, comprising an outer loop for quadrotor position control, a middle loop for payload swing angle control, and an inner loop for quadrotor attitude control, effectively enhancing the system’s stability and disturbance rejection capabilities. Rigorous mathematical analysis confirms the closed-loop stability of the system, and numerical simulations validate its effectiveness in achieving fast convergence, accurate trajectory tracking, and enhanced anti-swing performance. Future research will focus on applying this control scheme to real-world aerial transportation tasks, particularly in the field of emergency rescue.}, archive = {J_TASE}, author = {Yanhong Luo and Yannan Li and Jinliang Ding}, doi = {10.1109/TASE.2025.3592651}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18995--19006}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined time anti-swing control of quadrotor UAV transportation systems via 3-D spatial decomposition}, volume = {22}, year = {2025}, }
• (2025). Adaptive PI parameter identification and synchronous control of heat equations modeled multi-agent systems. TASE, 22, 18984--18994. (www)

  @article{ , abstract = {This paper addresses the temperature synchronization control challenge in a thermal pipeline system, which is modeled using a networked heat equation. Leveraging adaptive and PID control methodologies, an adaptive PI synchronization consensus protocol is formulated. This protocol generates a target tracking signal for the temperature output of each pipeline, thereby steering the temperature output of the pipelines accordingly. In response to the output tracking error, a synchronization controller and an infinite-dimensional observer are designed to counteract external disturbances and attain output synchronization of the system temperature. Simulation outcomes substantiate the efficacy of the proposed controller. Note to Practitioners—This study centers on the high-temperature sterilization process employed in industrial production. Throughout the heating phase, a uniform heat source supplies an identical quantity of heat. Given that the initial temperatures of individual sterilization units vary, once heating stabilizes after a certain duration, the final temperatures of these units also differ. This discrepancy results in issues such as heat wastage and suboptimal sterilization outcomes. Consequently, this study proposes the implementation of a synchronous controller based on an adaptive PI-form synchronous consensus protocol.}, archive = {J_TASE}, author = {Faqin Zhao and Xingong Cheng and Xiju Zong and Bao-Zhu Guo}, doi = {10.1109/TASE.2025.3592318}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18984--18994}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive PI parameter identification and synchronous control of heat equations modeled multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Exponential attitude tracking and vibration control for 3-D flexible spacecraft with disturbances and quantized inputs. TASE, 22, 18972--18983. (www)

  @article{ , abstract = {This study addresses the problem of attitude tracking and vibration suppression for 3D flexible spacecraft subject to external disturbances and input quantization, which are two critical factors that can significantly degrade control performance in harsh space environments with limited communication capacity. Using Hamilton’s principle, the spacecraft is modeled by coupled ordinary and partial differential equations to accurately characterize its infinite-dimensional dynamics. Nonlinear observers are constructed to exactly estimate unknown boundary disturbances. To deal with the difficulty caused by quantization, a linear time-varying model is introduced to describe hysteretic quantizers, and adaptive laws incorporating exponential functions are developed to estimate resulting unknown terms. Furthermore, a novel disturbance observer-based adaptive quantized control scheme is proposed, which guarantees the boundedness of all closed-loop signals and ensures that both attitude tracking errors and vibrations converge exponentially to zero, whereas existing methods typically achieve only convergence to small residual sets due to quantization effects or disturbances. In particular, the proposed scheme allows the quantizer parameters to be freely adjusted during operation for balancing communication burden and tracking performance. Simulation results show that, compared to traditional PD control, the proposed scheme exhibits faster convergence in attitude tracking and vibration elimination, and achieves higher control accuracy under unknown disturbances and changeable quantizer parameters. Note to Practitioners—This study presents a novel disturbance observer-based quantized control approach to address attitude tracking and vibration control for 3D flexible spacecraft in the presence of external disturbances and input signal quantization. The designed disturbance observers accurately compensate for a large class of disturbances, while the adaptive control scheme allows real-time adjustment of quantizer parameters, offering operational flexibility. By incorporating exponential functions, the controller achieves precise exponential tracking performance and ensures that undesired vibrations are exponentially suppressed to zero. For practical implementation, the proposed scheme is suitable for 3D spacecraft with flexible solar panels and other flexible systems which can be regarded as a rotating Euler-Bernoulli beam in 3D space. Future work will focus on addressing the vibration control problem while tracking time-varying desired attitude signals, with consideration of communication time delays.}, archive = {J_TASE}, author = {Le Li and Zhongkui Li}, doi = {10.1109/TASE.2025.3591968}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18972--18983}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exponential attitude tracking and vibration control for 3-D flexible spacecraft with disturbances and quantized inputs}, volume = {22}, year = {2025}, }
• (2025). Event-triggered impulsive control for switched systems involving stable and unstable modes. TASE, 22, 18963--18971. (www)

  @article{ , abstract = {This paper investigates the Lyapunov stability problem for switched systems involving stable and unstable modes under event-triggered impulsive control. By constructing multiple Lyapunov functions coupled with average dwell-time methods, some sufficient conditions on Lyapunov stability of switched systems involving stable and unstable modes are provided under event-triggered impulsive control. The designed event-triggered impulsive control strategy is mode-dependent, and Zeno behavior is eliminated. Then, with the construction of a connection between average dwell-time scheme and total dwell time of unstable modes, a relationship between mode-dependent event-triggered mechanism, system dynamics, and switching law is established. It plays a key role in designing event-triggered impulsive control with easy implementation. Moreover, this paper applies the theoretical results to the nonlinear switched system involving dual modes, in which an event-triggered impulsive control strategy is designed for the stability of the switched system. Finally, three numerical examples are given to verify the validity of the proposed results. Note to Practitioners—As a scaled-down version of various industrial plant mechanisms, the industrial plant emulator setup consists of components such as servomechanisms, drives, conveyor belts, and assembly machines. This paper investigates the problem of impulsive control of industrial plant emulator setup. To achieve the Lyapunov stability of switched systems involving stable and unstable modes, the event-triggered impulsive control strategy which integrates the advantages of event-triggered control and impulsive control is proposed. Some sufficient conditions have been constructed in the framework of event-triggered impulsive control. Note that the information of the stable and unstable modes is considered and fully extracted into the design of control strategy. Hence, compared with the general event-triggered impulsive control strategies in which the continuous dynamics of stable modes are ignored, the control burden has been further alleviated. Moreover, the proposed theoretical results can be extended to practical applications, such as single-link robot arm systems, switched multi-agent systems, and automotive engine systems. In future, an interesting topic is to design event-triggered impulsive controller for stability of switched delay systems with dual modes.}, archive = {J_TASE}, author = {Qi Fang and Mingzhu Wang and Xiaodi Li}, doi = {10.1109/TASE.2025.3591021}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18963--18971}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered impulsive control for switched systems involving stable and unstable modes}, volume = {22}, year = {2025}, }
• (2025). Fault tolerant control for nonlinear networked stochastic distribution systems with quantized signals and packet dropouts. TASE, 22, 18951--18962. (www)

  @article{ , abstract = {The fault-tolerant tracking control problem is investigated for nonlinear networked stochastic distribution systems (SDSs) under signal quantization, packet loss and system noise. Firstly, a novel logarithmic uniform quantizer is used to quantize the control signal and feedback signal and a data packet loss model is established through Bernoulli random process, which establishes the theoretical foundation for subsequent fault diagnosis (FD). On this basis, a FD observer is designed, which can accurately estimate the fault information under noise and provide reliable fault estimation information for subsequent fault-tolerant control (FTC). Subsequently, to maintain satisfactory tracking performance after fault occurrence, a state feedback fault-tolerant tracking control strategy is proposed based on the parallel distributed compensation technique. Finally, the superiority of the presented FTC method is proven through a numerical example and a practical example. Note to Practitioners—The motivation is to improve the security of SDSs in network environments without changing the system structure or increasing costs. When the fault occurs, quantization and packet loss can result in missing information containing fault characteristics, thereby affecting the real-time and accuracy performance of FD, and also increasing the difficulty of designing fault-tolerant controllers. Regarding to the above problems, a novel FD and FTC strategy is proposed, which can effectively compensate the negative influences of packet loss, actuator fault and quantization error on the system performance.}, archive = {J_TASE}, author = {Lifan Li and Lina Yao}, doi = {10.1109/TASE.2025.3588222}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18951--18962}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault tolerant control for nonlinear networked stochastic distribution systems with quantized signals and packet dropouts}, volume = {22}, year = {2025}, }
• (2025). 3-axial force self fault-tolerant decoupling of surgical forceps integrating step-coated FBG for spinal endoscopic robot. TASE, 22, 18933--18950. (www)

  @article{ , abstract = {This work proposed surgical forceps as a sensor that integrated step-coated Fiber Bragg Grating (FBG) for the 3-axial force sensing in the percutaneous spinal endoscopic robot. The step-coated FBG achieved a reflection spectrum with double wavelength peaks, providing more wavelength signals for fault-tolerant decoupling. The temperature and force sensitivity were regulated by adjusting the coating size to reduce decoupling error to 3.68% F.S, solving the span temperature ( $\ge 20~^{\circ }$ ) disturbance from the operating room to the operative area. The optical fiber metallization and laser welding package was proposed to achieve stable connections between the quartz optical fiber and metal forceps. The innovative package made the forceps tolerate $180~^{\circ }$ dry-heat sterilization and aqueous erosion, allowing for sterilized reuse (11 times with error lower than 4.98% F.S) and long-term stability (one month with error lower than 4.36% F.S). A Wavelet Fuzzy Entropy (WFE) and Extreme Learning Machine (ELM) based dynsamic fault-tolerant decoupling strategy was proposed. The WFE-ELM reduced the sensor error to 4.42% F.S. under the influence of spectrum chirp noise and single-branch FBG breakage, and the fault-tolerant recovery rate within 10% F.S. error was raised to 40.23%. The designed surgical instruments were integrated with a spinal endoscopic surgical robot to conduct the spine model and pig experiments, verifying its force-sensing effectiveness. Note to Practitioners—This paper aims to address clinical applications of optical fiber force sensors in spinal endoscopic surgery, such as preoperative sterilization, aqueous erosion, temperature perturbations, and stability of reused. Most optical fiber sensors in the medical field are often packaged with adhesives, which make it difficult to withstand high-temperature sterilization environments and are prone to debonding under aqueous erosion. Moreover, FBG force sensors increase interaction force sensing errors under large-span temperature perturbations between the operating room and the body. These limitations hinder the development of fiber optic sensors in clinical applications. To address these issues, this work proposed step-coated FBG and their laser welding package method. Unlike conventional temperature compensation methods by increasing FBGs, the step-grid region microstructure is designed to provide extra twinned wavelength signals, improving temperature and force decoupling accuracy. Laser welding enables robust integration of metalized optical fibers with titanium alloy substrates, ensuring the resultant sensor exhibits exceptional resistance to high-temperature sterilization (up to $180~^{\circ }$ C) and prolonged aqueous erosion in biomedical environments. In addition, a fault-tolerant measurement method based on WFE and ELM was proposed to improve long-term service stability, which solves the problem of increased measurement error and failure due to degradation of sensing performance and link breakage failure. The results of sterilization and liquid erosion experiments show that the sensor can be reused 11 times with an error of 4.98%, reducing its cost in medical clinical applications.}, archive = {J_TASE}, author = {Chen Zhao and Changsheng Li and Mingchang Du and Haolei Fan and Jinpeng Diao and Kaifeng Wang and Jun Wang and Weiwei Xia and Xingguang Duan and Haiying Liu and Yuegang Tan and Tianliang Li}, doi = {10.1109/TASE.2025.3585161}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18933--18950}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {3-axial force self fault-tolerant decoupling of surgical forceps integrating step-coated FBG for spinal endoscopic robot}, volume = {22}, year = {2025}, }
• (2025). Event-triggered-based adaptive neural network secure strategy for stochastic networked nonlinear systems under DoS attacks. TASE, 22, 18920--18932. (www)

  @article{ , abstract = {An adaptive neural networks event-triggered secure strategy is investigated to explore stochastic networked nonlinear systems with denial-of-service (DoS) attacks and unknown dead zone. In order to alleviate the negative effects of DoS attacks, a state estimator is constructed to approximate the immeasurable states. And the unknown dead zone input function is described as a bounded disturbance and a time-varying nonlinear function to offset the unknown dead zone effect. To defend against DoS attacks while ensuring system stability, an adaptive event-triggered prescribed performance controller is designed, ensuring that all signals of the closed-loop system are bounded in probability, and the tracking error tends to the designed performance bound within a predefined finite time. Meanwhile, this secure strategy can completely eliminate potential Zeno behavior. Subsequently, the modified average dwell time (ADT) approach was integrated with Lyapunov stability theory to establish the stability of the system. Eventually, two simulation results are utilized to prove the effectiveness of the developed strategy. Note to Practitioners—Networked systems have been widely utilized in various fields, including the smart grid and aerospace. However, the openness of networks provides attackers with numerous opportunities for malicious assaults. These cyber threats can result in a decline in the performance of the systems. Therefore, it is crucial to study the security challenges faced by networked systems when they are subjected to malicious attacks. Hence, considering the aperiodic DoS attacks, the unknown nonlinear functions, and the unmeasured states existing simultaneously in the studied systems, it is a challenging and meaningful task to achieve the desired security control objectives. This paper focuses on the event-triggered secure strategy for stochastic nonlinear systems under DoS attacks. The event-triggered communication mechanism is introduced to reduce the computational burden. A single-link robotic manipulator system is given to validate the effectiveness of the proposed method.}, archive = {J_TASE}, author = {Junsheng Zhao and Xin Zhang and Zong-Yao Sun and Weihai Zhang}, doi = {10.1109/TASE.2025.3592216}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18920--18932}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered-based adaptive neural network secure strategy for stochastic networked nonlinear systems under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Output tracking control for nonlinear affine systems with relative degree two based on control barrier functions. TASE, 22, 18908--18919. (www)

  @article{ , abstract = {Nonlinear affine systems with relative degree two widely exist in the control field, and the unified output tracking control approach of these systems is still an open problem. This paper presents a formal framework for output tracking control design methodology for nonlinear affine systems with relative degree two. Specifically, a control barrier function is constructed to enforce the forward invariance or the stability of a set contained within a 0-sublevel set of the first-order derivative of pre-stablished Lyapunov function, thereby stabilizing the system output to the desired value. Additionally, by taking the first-order derivative of the elaborated control barrier function, control input terms with a nonzero coefficient are introduced into the stability condition, addressing the weakness of controller design using conventional control Lyapunov function approaches, which fails in states where the coefficient of the control input term is zero. Furthermore, the stability conditions are proposed in the form of linear inequality constraints involving control inputs, with which a collection of admissible controllers is provided. Then, by resorting to the pointwise minimum norm (PMN) technique, a numerical solution of an admissible controller can be derived in the context of a quadratic program. We finally demonstrate the effectiveness of the controller design method via two simulations, adaptive cruise control and vehicle lateral control. Note to Practitioners—The motivation for this work originates from the control requirements of nonlinear systems with relative degree two. Many practical systems, such as vehicle motion systems, exhibit characteristics of a relative degree two. To address this, various output tracking control methods have been developed. This paper utilizes control barrier functions to design an optimization-based output tracking control method that does not rely on the structural analysis of the system itself. Additionally, for systems with relative degree two, the traditional control Lyapunov method may struggle to ensure the feasibility of solutions to the quadratic programming problem due to the possibility of control gain coefficients being zero, which in turn cannot guarantee the convergence of system output. The method proposed in this paper focuses on solving this issue, and the two simulation experiments demonstrated that the proposed method can be effectively applied to vehicle control systems with relative degree two.}, archive = {J_TASE}, author = {Changzhu Zhang and Wenjie Mao and Hao Zhang and Zhuping Wang and Hak-Keung Lam}, doi = {10.1109/TASE.2025.3591853}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18908--18919}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Output tracking control for nonlinear affine systems with relative degree two based on control barrier functions}, volume = {22}, year = {2025}, }
• (2025). Adaptive control with fast convergence for state-quantization constrained manipulators under unknown measurement powers. TASE, 22, 18900--18907. (www)

  @article{ , abstract = {This brief focuses on the adaptive control with fast convergence performance for robot manipulators with unknown measurement powers and quantized states. Since there are some uncertain factors in the actual operation of manipulators, which lead to unknown measurement functions. This paper considers a new measurement function, i.e., the powers of the measured values are unknown time-varying functions and some auxiliary variables are introduced to transform control errors, thereby compensating unknown measurement powers. Unlike the traditional adaptive control methods, this paper designs a novel quantized control strategy based on quantized states. Since quantized states are discrete and not derivable, which poses the difficulty in the design of control algorithm. To compensate the effect of quantized states, this paper establishes the relationship between quantized signals and unquantized signals. The input saturation is handled by developing an auxiliary system. By designing a fast convergence function, the control errors under the proposed strategy can quickly enter into a tunable interval within a time set in advance. Finally, a manipulator simulation is performed to verify the validity of the control method. Note to Practitioners—The robotic manipulators are widely used in industrial production and intelligent manufacturing, which better promotes the development of production automation and intelligence. So the stabilization study for the robotic manipulator has some physical significance. In practice, the input powers of measured states may change due to the inaccuracy of measurement tools and external disturbances. In addition, when the input signal of system is too large, it may lead to the limitation of system output, thereby affecting the normal operation of system. At the same time, in industrial production, it is very important to improve control performance and save resource consumption. Therefore, for the robotic manipulator with unknown measurement powers and input saturation, this article develops an adaptive control scheme with fast convergence performance based on quantized states. Compared with traditional control methods, the presented quantized control scheme can reduces resource loss well and enables the control error converge to a preset interval within a predefined time. The presented quantized control scheme has a wide application prospect and practical value in industrial robotic manipulators.}, archive = {J_TASE}, author = {Yadong Yang and Jianye Gong and Qikun Shen}, doi = {10.1109/TASE.2025.3591491}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18900--18907}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive control with fast convergence for state-quantization constrained manipulators under unknown measurement powers}, volume = {22}, year = {2025}, }
• (2025). Dynamic weight-optimized prototypical contrastive network for cross-domain few-shot bearing fault diagnosis. TASE, 22, 18889--18899. (www)

  @article{ , abstract = {Due to the limited sample size caused by preventive maintenance and the variable data distribution caused by environmental factors, the performance of well-trained laboratory models has decreased significantly when confronted with actual industrial bearing fault diagnosis. The existing domain adaptation methods using cross-domain labeled samples for classification make it difficult to resist the influence of outliers on the overall matching, which leads to a negative transfer problem. Given the above issues, a dynamic weight-optimized prototypical contrastive network is proposed. Primarily, domain adaptation is assisted by the discriminative information the classifiers convey during the prediction process. The model is prompted to find the source data that matches the distribution, removing the effect of distribution variability. Furthermore, the sample-aware weighting term evaluates the difficulty of the sample classification and suppresses the performance degeneration of domain adaptation. Subsequently, the in-domain prototypical contrastive learning evaluates dynamically the intensity of feature distribution around each prototype. The features from the same category are inspired to move closer to the prototype to enhance the consistency and discrimination of intra-class features. Meanwhile, cross-domain instance-prototype learning reduces the distribution inconsistency of the corresponding category data in the source and target domains to enable fine-grained inter-domain alignment and mitigate the negative transfer problem. Through numerous comparative experiments, this method shows superior effectiveness and engineering diagnostic feasibility under cross conditions with limited data resources. Note to Practitioners—This article addresses the cross-domain distribution variability and few-shot bearing limitation. Existing research on intelligent bearing fault diagnosis is based on the unattainable assumption that fault samples are sufficient and homogeneously distributed. Moreover, domain-adaptive methods using cross-domain labeled samples for classification are too coarse to resist the effect of the outlier of bearing operational data, leading to the negative transfer problem. In this article, a dynamic weight-optimized prototypical contrastive network is proposed. Primarily, a feature-level cross-domain alignment module pre-trains the feature extractor to generate unique features for disparate bearing fault signals. After pre-training, the model can find source data that matches the distribution, removing the effect of distribution variability. Among it, the sample-aware weighting term, which dynamically weights the samples according to their classification difficulty, prevents the model performance from degrading on the cross-domain task. Then, the prototypical contrastive learning module can construct prototypes, representing typical features of bearing faults. Each category of features is approached with similar prototypes, pushing away from disparate categories of prototypes. Relying on the prototypes to be aligned with disparate domain instances to weed out the effects of outliers. Not only does it achieve cross-domain alignment of underlying features, but it also aligns semantic structures in a shared cross-domain space. Through numerous experiments, it has been proved that the method can not only solve the cross-domain problem and the few-shot problem, but is also more applicable to the fault diagnosis of real industrial systems. Furthermore, its performance is much higher than the existing methods.}, archive = {J_TASE}, author = {Donglin Li and Yuqi Yao and Yuxian Zhang and Jiacan Xu and Qisen Wang}, doi = {10.1109/TASE.2025.3589234}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18889--18899}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic weight-optimized prototypical contrastive network for cross-domain few-shot bearing fault diagnosis}, volume = {22}, year = {2025}, }
• (2025). Optimal consensus control strategy for multi-agent systems under cyber attacks via a stackelberg game approach. TASE, 22, 18875--18888. (www)

  @article{ , abstract = {Based on the Stackelberg game theory, this paper studies the optimal consensus control problem of the multi-agent system (MAS) under multiple network attacks. Multiple attackers interfere with the information interaction between agents by intervening in the transmission channel of the multi-agent wireless network, thereby changing the communication topology of the MAS. Under the limited energy of the system and attackers, the energy allocation of transmission channels is used as the performance index, and the importance of the transmission channel to the MAS is used as an auxiliary indicator to establish a Stackelberg game model between the MAS and attackers under incomplete information. By solving the Stackelberg equilibrium solution of the game model, the optimal attack energy allocation scheme and the optimal energy allocation scheme of the system are obtained, respectively. Then, based on the equilibrium solution, the Hamilton-Jacobi-Bellman equation with coupling terms for the MAS under attacks is given. In view of the broad learning system theory, online algorithms that approximate the optimal performance function and the optimal controller are designed respectively, and sufficient conditions are obtained for the proposed optimal controller to ensure the stability and consistency of the MAS. Finally, the effectiveness of the method is verified by a numerical simulation. Note to Practitioners— This study employs a Stackelberg game model to analyze the optimal consensus control problem in the MAS facing network attacks. By utilizing this framework, effective energy allocation strategies can be devised for both the MAS and potential attackers under resource constraints, demonstrating broad applicability across various fields, including intelligent vehicle fleets and power grids. The simulation section specifically focuses on the attack-defense dynamics of autonomous aerial vehicles, validating energy allocation strategies and control algorithms under different attack scenarios, thus providing effective solutions for real-world applications.}, archive = {J_TASE}, author = {Peng Yu and Yanpeng Hu and Yinghui Wang and Ruizhe Jia and Jin Guo}, doi = {10.1109/TASE.2025.3591858}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18875--18888}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal consensus control strategy for multi-agent systems under cyber attacks via a stackelberg game approach}, volume = {22}, year = {2025}, }
• (2025). Enhancing safety in collaborative cable-driven parallel robots: Contact distinction and management for carrying tasks. TASE, 22, 18860--18874. (www)

  @article{ , abstract = {Cable-Driven Parallel Robots (CDPRs) have shown significant potential in industrial applications due to their large workspace, high payload capacity, and flexibility. Nonetheless, ensuring safe and efficient human-robot collaboration, particularly during carrying tasks, remains an open problem. This paper primarily contributes a cable tension-based metric for detecting contact events, combined with a frequency-domain analysis of estimated external wrenches to distinguish different contact scenarios. Building upon this framework, three common contact scenarios are then addressed: (i) payload placement and removal without rigid attachment to the moving-platform (SC1), (ii) unintentional human-cable collisions (SC2), and (iii) human-platform collisions (SC3). Specific management strategies are then proposed, including real-time mass estimation for payload handling, cable tension release for human-cable collisions, and a compliant trajectory controller for human-platform collisions. Experimental validations on a CDPR prototype demonstrate accurate scenario classification and safe contact handling without compromising overall productivity. By facilitating safer interactions, improved adaptability, and reliable handling of diverse contact events, this work expands the applicability of collaborative CDPRs in real-world industrial and logistic settings. Note to Practitioners—In many industrial applications, robots operating alongside human workers must quickly detect and respond to contact events to maintain safety and efficiency. Traditional methods often rely on force thresholds, which can introduce delays or false alarms, especially in dynamic tasks where rapid collisions must be addressed. Our work proposes an alternative approach that utilizes a frequency-domain analysis of sensed cable tensions and a weight-score system to robustly distinguish among various contact scenarios. By examining tension signals in the frequency domain, more nuanced information can be extracted, enabling faster and more accurate responses than threshold-based methods. From a practical standpoint, this technique is relatively straightforward to implement with various sensors, and it can be adapted to robots composed of different materials and link configurations by tuning the weight-score system for each new setup. Its primary benefit is improved contact distinction with high precision, which can enhance operator safety and reduce downtime without requiring extensive hardware modifications. However, like any sensor-driven control method, performance depends on adequate calibration and consistent signal quality. Future research might focus on integrating additional sensing modalities (e.g., vision or proximity sensors) and exploring broader deployment in diverse collaborative robot systems.}, archive = {J_TASE}, author = {Hanbang Gao and Christine Chevallereau and Stéphane Caro}, doi = {10.1109/TASE.2025.3589966}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18860--18874}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing safety in collaborative cable-driven parallel robots: Contact distinction and management for carrying tasks}, volume = {22}, year = {2025}, }
• (2025). Fixed-time discontinuous control for cluster secure synchronization of interlayer switching networks under attacks. TASE, 22, 18848--18859. (www)

  @article{ , abstract = {This paper mainly concentrates on the issue of fixed-time cluster secure synchronization upon interlayer switching networks with DoS attacks, in which the DoS attacks may cause random damage to the nodes/edges in networks and then generate impulsive disturbances. To resist the attacks, one robust fixed-time discontinuous control strategy is proposed, such that all the nodes in interlayer switching networks can achieve the cluster secure synchronization even when the communication topology are disconnected or changed over time after being influenced by attacks. By using the Lyapunov method, several effective conditions for fixed-time cluster secure synchronization and the maximum convergence time near to the actual value are obtained. In view of the comparison principle, a suitable estimation method upon stability time for the switching networks is established, and the stability time is computable and controllable. Moreover, the correlation among attack intensity, connectivity network, impulsive interval, impulsive disturbance intensity and synchronization time is revealed. Simulation case is finally supplied to manifest the availability of the obtained outcomes. Note to Practitioners—This work aims at devising an efficient discontinuous control strategy within fixed-time for interlayer switching networks under attacks, such as power networks, industrial networks, and etc. In practical applications, the state of attacked nodes may transform unexpectedly and the nodes as well as their links may not operate under certain circumstances, while the classical time-controllable methods usually overlook these practical constraints. Consequently, by applying a discontinuous control with fixed-time strategy, not only the transmission efficiency is improved but also the control cost against attacks is reduced. Additionally, the fixed-time approach is designed such that the destructiveness of attacks and synchronized time are all predictable and controllable, even if the communication topology may be disconnected and changed over time after being affected by attacks, the given scheme is more effective for realistic requirements.}, archive = {J_TASE}, author = {Lili Zhou and Mingzhe Huang and Fei Tan and Guangdeng Zong and Zhen Wang and Guangming Zhuang}, doi = {10.1109/TASE.2025.3589187}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18848--18859}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time discontinuous control for cluster secure synchronization of interlayer switching networks under attacks}, volume = {22}, year = {2025}, }
• (2025). Resilient control in multi-hydrofoil crafts: Tackling actuator faults and false data injection for attitude consensus. TASE, 22, 18835--18847. (www)

  @article{ , abstract = {This paper addresses the attitude consistency problem in multi-hydrofoil crafts, considering actuator faults, false data injection, stochastic ocean wave disturbances, and topological propagation effects. We propose a game-theoretic fault-tolerant control (FTC) framework integrating a composite controller with a specific performance index. In this framework, both local and neighboring node information is considered. The desired controller, along with non-ideal anomalies and disturbances, is treated as participants, transforming the FTC problem into a multi-player non-cooperative game. Moreover, a single critic neural network (NN) is utilized to alleviate the challenges associated with solving a partial differential equation, thereby facilitating the derivation of the ideal control law. Compared to traditional local information-based control methods, the proposed approach incorporates neighborhood information. It integrates faults, disturbances, and propagation effects within a unified framework for optimal control, enhancing FTC effectiveness. We conduct comparative experiments using a four-craft scenario on the dSPACE platform. The results demonstrate that the proposed method reduces the mean absolute deviation (MAD) of local tracking error by at least 28.00% and the standard deviation (STD) by at least 7.76% compared to the other two methods. Note to Practitioners—This work stems from the challenge of maintaining attitude consensus in a hydrofoil fleet operating under real-world disturbances and actuator failures, considering the propagation effects in interconnected systems. The game-theoretic framework models FTC as a competitive interaction between the controller and anomalies in a unified architecture, eliminating the need for separate fault diagnosis and improving efficiency. Additionally, NNs replace complex mathematical solutions, enabling engineers to deploy control laws without specialized knowledge of partial differential equations. Hardware-in-the-loop experiments validate the feasibility and provide a paradigm of resilience control for other interconnected control systems for pertinent practitioners.}, archive = {J_TASE}, author = {Tao Wang and Dezhi Xu and Chengxi Zhang and Bin Jiang}, doi = {10.1109/TASE.2025.3591792}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18835--18847}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient control in multi-hydrofoil crafts: Tackling actuator faults and false data injection for attitude consensus}, volume = {22}, year = {2025}, }
• (2025). Toward turning performance optimization of a multi-flexible robotic fish. TASE, 22, 18821--18834. (www)

  @article{ , abstract = {Over prolonged evolution, natural fish possess exceptional maneuverability. For underwater robotic fish, maneuverability is a crucial performance metric during locomotion. This paper proposes a turning control optimization framework for a multi-flexible-joint bionic robotic fish. Firstly, three distinct turning strategies are devised based on kinematic analysis of the flexible robotic fish’s turning motion. Experimental results indicate that the control strategies need to be adjusted with respect to motion frequency to achieve optimal performance for the flexible robotic fish. Furthermore, an optimization problem incorporating dynamic constraints and cost functions of the flexible robotic fish is constructed. Based on a Constrained Iterative Linear Quadratic Regulator (CILQR), a turning performance optimization method is designed. Subsequently, optimal control strategies under both stationary and motion states are derived, effectively enhancing the turning performance of the flexible robotic fish. Finally, simulation and experimental results validate the effectiveness of the designed method. The developed flexible robotic fish could achieve a maximum swimming speed of 1.63 BL/s (body lengths per second) and a maximum average turning speed of 130.5°/s, offering guidance for the practical application of flexible robotic fish in aquatic environments. Note to Practitioners—This paper was motivated by the problem of improving maneuverability in underwater robotic systems. While traditional underwater vehicles often struggle with agile movement in complex environments, this paper suggests a solution inspired by nature. The key practical contribution is a turning control optimization framework that enhances maneuverability without requiring complex mechanical redesigns. We’ve developed three distinct turning strategies that can be selected and adjusted based on operational requirements. Our research shows that motion frequency significantly impacts turning performance—a critical insight for field deployment. The optimization method we propose can be implemented in existing control systems with minimal computational overhead, making it accessible for various underwater applications. This approach could benefit underwater inspection, environmental monitoring, and search operations where navigation in confined or obstacle-rich environments is required.}, archive = {J_TASE}, author = {Ben Lu and Chao Zhou and Jian Wang and Min Tan}, doi = {10.1109/TASE.2025.3591790}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18821--18834}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward turning performance optimization of a multi-flexible robotic fish}, volume = {22}, year = {2025}, }
• (2025). Enhancing the condensate throttling stability by phase lead error-based active disturbance rejection control under full operating conditions. TASE, 22, 18809--18820. (www)

  @article{ , abstract = {Storage utilization can improve the frequency regulation of thermal power units. Condensate throttling is one of the most popular means, and the water level fluctuation brought by its process affects the unit operation safety. In this regard, a phase lead error-based active disturbance rejection control (PL-EADRC) that enhances the disturbance observation capability of conventional ADRC is designed. The condensate throttling with the PL-EADRC achieves a smooth change of load and suppresses the fluctuation of deaerator and condenser water levels. It shows that the proposed strategy has good disturbance rejection capability and robustness under full operating conditions. In addition, its simplicity of structure and ease of tuning mean that it has strong industrial utility. Note to Practitioners—In this paper, an industrial practical control strategy is proposed to stabilize the process parameters of thermal power units under rapid load change. The improvement does not depend on increasing the complexity of control structure and iterative operation, and each parameter has clear physical meaning, which is easy to be adjusted by engineers. The signals used are all measured from the original, without adding new measuring points, which is conducive to flexible transformation.}, archive = {J_TASE}, author = {Dawei Pang and Tianmu Qin and Hengyi Zhu and Ming Du and Yuguang Niu}, doi = {10.1109/TASE.2025.3591618}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18809--18820}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing the condensate throttling stability by phase lead error-based active disturbance rejection control under full operating conditions}, volume = {22}, year = {2025}, }
• (2025). Learning from fish: A two-stage transfer learning method for a bionic robotic fish. TASE, 22, 18796--18808. (www)

  @article{ , abstract = {Directly learning the swimming behaviors of real fish can significantly enhance the swimming performance of bionic robotic fish. This paper presents a novel transfer learning method based on a dynamic trajectory control approach for the robotic fish to learn swimming skills from real fish. First, we develop a fish motion capture system and a crucial motion extraction approach to realize precise decomposition of fish motions and collect abundant meaningful features from a snakehead fish as pre-training data. Next, we construct a two-stage transfer learning method based on Deep Deterministic Policy Gradient (DDPG), including an offline and an online stage. Specifically, in the offline stage, the obtained pre-training data is processed for experience learning within a DDPG-based network, whereas in the online stage, a dynamic trajectory tracking method is utilized to refine the robotic fish’s motions in real time based on the learned strategies. Experimental results on a self-developed four-joint robotic fish show that the proposed method effectively extracts and transfers biological motion features into the motion control of the robotic fish. Compared to the conventional CPG method, the proposed approach exhibits stronger acceleration capabilities and more efficient swimming, resulting in enhanced maneuverability of the robotic fish. Overall, this approach provides a technical foundation for bionic robotics to learn from nature. Note to Practitioners—The motivation of this work stems from the need for more adaptable control of robotic fish swimming behaviors to improve their maneuverability. Traditional dynamic trajectory tracking methods lack the flexibility to adjust motion characteristics according to specific requirements. In contrast, the swimming motions of real fish offer valuable insights for controlling robotic fish. This paper proposes a two-stage transfer learning framework that extracts action features from biological fish movements across various conditions, and then transfers these features to robotic fish for optimization in real-world environments. The knowledge gained from biological motion characteristics provides effective guidance for controlling robotic fish swimming behaviors. In practical scenarios, the proposed method enables control of robotic fish movements based on factors such as acceleration requirements and oscillation constraints, offering enhanced flexibility in acceleration and cruising performance.}, archive = {J_TASE}, author = {Fuyang Yu and Zhengxing Wu and Jian Wang and Lianyi Yu and Yukai Feng and Min Tan and Junzhi Yu}, doi = {10.1109/TASE.2025.3591314}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18796--18808}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning from fish: A two-stage transfer learning method for a bionic robotic fish}, volume = {22}, year = {2025}, }
• (2025). Synchronization of stochastic complex networks by dynamic periodic adaptive event-triggered control. TASE, 22, 18786--18795. (www)

  @article{ , abstract = {Throughout this paper, we introduce a novel dynamic periodic adaptive event-triggered control (ETC) strategy to explore the exponential synchronization in mean square (ESMS) of stochastic complex networks (SCNs). Firstly, unlike continuous ETCs, periodic ETCs rely on the periodic monitoring of system states, inherently avoiding the Zeno phenomenon. Secondly, by incorporating a dynamic function and an exponential term, the event-triggering frequency can be effectively reduced. Simultaneously, the triggering thresholds within the event-triggering mechanism automatically adjust in response to changes in the error term, further enhancing control efficiency. Subsequently, through the integration of graph-theoretic results and Lyapunov analysis, we obtain sufficient conditions to ensure ESMS of SCNs under the proposed dynamic periodic adaptive ETCs. Lastly, the analytical results are applied to a stochastic single-link robot arm system, and corresponding numerical simulation results are provided to illustrate the effectiveness of our proposed strategy. Note to Practitioners—In industrial network systems, due to various stochastic disturbances, synchronization of SCNs is widely applied across multiple domains, such as power grid systems and coordinated robotic systems. Then, how to design effective controllers for synchronization of SCNs under limited communication resources has become a critical technical challenge, which promotes this paper. Based on periodic observation of system states and parametric adaptation mechanisms, the dynamic periodic adaptive ETC (DPAETC) strategy is proposed to achieve ESMS of SCNs and reduce communication resource consumption at the same time. In addition, from the simulation results, it can be seen that our DPAETC strategy greatly enhances the flexibility and effectiveness of the control strategy compared to those traditional ETC methods in practical application.}, archive = {J_TASE}, author = {Xuetao Yang and Wei Wu and Quanxin Zhu}, doi = {10.1109/TASE.2025.3590440}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18786--18795}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synchronization of stochastic complex networks by dynamic periodic adaptive event-triggered control}, volume = {22}, year = {2025}, }
• (2025). Virtual error-based data-driven P-type adaptive predictive control and its applications. TASE, 22, 18773--18785. (www)

  @article{ , abstract = {In this paper, a P-type adaptive predictive control (PAPC) method is presented for a category of unknown multi-input multi-output (MIMO) discrete-time systems with nonaffine nonlinear dynamics. First, the unknown nonlinear model is altered to a linear form containing an unknown pseudo-partial derivative (PPD) matrix utilizing the partial-form dynamic linearization (PFDL). A predictive model is then established by employing the modified projection algorithm, an auto-regressive model, and an output estimation technique. Based on the predictive model, an adaptive learning law that incorporates estimated tracking error information is used to generate the virtual error. Then, a data-driven PFDL-PAPC algorithm is constructed by replacing the actual tracking error in the P-type controller with the virtual one. The bounded convergence properties of the output estimation and tracking error dynamics are theoretically analyzed using the contraction mapping principle. The effectiveness of the PFDL-PAPC method is demonstrated through coupled tanks and actual data-based blast furnace ironmaking experiments. Note to Practitioners—Model-based control strategies are highly dependent on the model of the controlled plant, which makes it challenging to apply them in complicated industrial processes. In this paper, a P-type adaptive predictive control algorithm is presented. It is directly driven by the virtual error generated through the multi-layer prediction mechanism without requiring any modeling procedure. The operators can flexibly adjust the linearization length according to the system’s dynamic complexity. Furthermore, the proposed algorithm can effectively resist the negative influence of input disturbances. The coupled tanks and actual blast furnace ironmaking data-based experiments are provided to verify the effectiveness of the proposed algorithm.}, archive = {J_TASE}, author = {Guojie Li and Ping Zhou and Yuyang Zhou and Tianyou Chai}, doi = {10.1109/TASE.2025.3590028}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18773--18785}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Virtual error-based data-driven P-type adaptive predictive control and its applications}, volume = {22}, year = {2025}, }
• (2025). IPLAM: A high-dimensional expensive simulation optimization method, with application to design space exploration in processor. TASE, 22, 18761--18772. (www)

  @article{ , abstract = {Design Space Exploration (DSE) in processors is an expensive discrete simulation optimization problem. The data requirements of the regular data-driven methods are so large that it is challenging to converge to a satisfactory solution within a limited simulation budget. Based on binary integer linear programming (BILP), an iteratively piecewise linear approximate method (IPLAM) is proposed for this kind of problems to reduce the dependence of simulation data. IPLAM starts from an initial reference point. Each iteration generates a set of trial points based on the most promising reference point by piecewise shift method. After evaluating the trial points, a local surrogate model is constructed for the unit neighborhood of the reference point. The surrogate model is then used to guide the exploration of the next reference point. In theory, IPLAM can converge to the global optimal point under mild assumptions, which is further verified by the numerical experiments. Meanwhile, the numerical experiments demonstrate that IPLAM outperforms the advanced Bayesian optimization and differential evolution methods on high-dimensional discrete closed-box test functions. Besides, the practical effectiveness of IPLAM is validated by an industrial case for processor DSE. Note to Practitioners—This paper proposes an efficient optimization method, IPLAM, for high-dimensional expensive simulation problems. To cope with the extremely limited simulation data, we construct the surrogate objective for the local space around the reference point by evaluating a set of trail points. The surrogate objective is then based on to locate the most promising point as the next reference point. We also provide theoretical guarantees of convergence for the proposed method. According to the numerical experiments, the proposed method demonstrates a much superior exploration efficiency and a lower simulation data requirement than competing methods. Besides, compared with an advanced Bayesian optimization method (HEBO) in an industrial case of processor DSE, the proposed method achieves a 47% improvement in convergence speed and a 15% reduction in simulation overhead while the difference of simulation score is marginal.}, archive = {J_TASE}, author = {Xiaoliang Lv and Qiaozhu Zhai and Yuhang Zhu and Jianchen Hu and Xiaohong Guan}, doi = {10.1109/TASE.2025.3588217}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18761--18772}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {IPLAM: A high-dimensional expensive simulation optimization method, with application to design space exploration in processor}, volume = {22}, year = {2025}, }
• (2025). Fast unfolding-based indoor space partitioning and rapid complementary search planning for high-rise fire rescue. TASE, 22, 18750--18760. (www)

  @article{ , abstract = {This paper introduces a swift solution for the complementary coverage path planning issue for multiple uncrewed aerial vehicles (UAV) in 3-dimensional, non-convex, and trap-laden indoor environments. A topology graph is constructed by sampling the blueprint. To divide the indoor environment into several convex areas and form tasks for the UAVs, the Louvain (fast unfolding) algorithm is utilized twice for space partitioning and task assigning. Finally, to ascertain the path for each UAV, a tiny-scale traveling salesman problem (TSP) is formed and solved. The simulation result reveals that the proposed strategy has significantly improved efficiency in both 2-dimensional and 3-dimensional path planning, underscoring the algorithm’s practical relevance for high-rise fire rescue operations. Note to Practitioners—This paper is motivated by the requirement of urgent task assignment and route finding for multiple uncrewed aerial vehicles (UAV) in indoor searching, especially for high-rise fire rescue. A rapid task allocation and pathfinding framework that balances efficiency and optimization levels is proposed. Unlike other complex environments, community-detection-based task allocation methods are recommended for actual indoor space rather than clustering-based ones. Additionally, the degree of each node in the sampling graph can be used to rapidly determine the sensor’s detection range at that node and the distance to obstacles. Making reasonable use of this can further enhance the efficiency of task allocation and pathfinding. Extensive experimental results in actual indoor maps show the effectiveness of the proposed task allocation and pathfinding framework. Note that our method is designed only for indoor environments divided into multiple rooms by continuous walls, and the effectiveness in other types of environments cannot be guaranteed.}, archive = {J_TASE}, author = {Jintao Chen and Hongru Ren and Qi Zhou and Hongyi Li}, doi = {10.1109/TASE.2025.3590419}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18750--18760}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast unfolding-based indoor space partitioning and rapid complementary search planning for high-rise fire rescue}, volume = {22}, year = {2025}, }
• (2025). Preference-based multi-objective reinforcement learning. TASE, 22, 18737--18749. (www)

  @article{ , abstract = {Multi-objective reinforcement learning (MORL) is a structured approach for optimizing tasks with multiple objectives. However, it often relies on pre-defined reward functions, which can be hard to design for balancing conflicting goals and may lead to oversimplification. PReferences can serve as more flexible and intuitive decision-making guidance, eliminating the need for complicated reward design. This paper introduces preference-based MORL (Pb-MORL), which formalizes the integration of preferences into the MORL framework. We theoretically prove that preferences can derive policies across the entire Pareto frontier. To guide policy optimization using preferences, our method constructs a multi-objective reward model that aligns with the given preferences. We further provide theoretical proof to show that optimizing this reward model is equivalent to training the Pareto optimal policy. Extensive experiments in benchmark multi-objective tasks, a multi-energy management task, and an autonomous driving task on a multi-line highway show that our method performs competitively, surpassing the oracle method, which uses the ground truth reward function. This highlights its potential for practical applications in complex real-world systems. Note to Practitioners—Decision-making problems with multiple conflicting objectives are common in real-world applications, e.g., energy management must balance system lifespan, charge-discharge cycles, and energy procurement costs; autonomous driving vehicles must balance safety, speed, and passenger comfort. While multi-objective reinforcement learning (MORL) is an effective framework for these problems, its dependence on pre-defined reward functions can limit its application in complex situations, as designing a reward function often fails to capture the full complexity of the task fully. This paper introduces preference-based MORL (Pb-MORL), which utilizes user preference data to optimize policies, thereby eliminating the complexity of reward design. Specifically, we construct a multi-objective reward model that aligns with user preferences and demonstrate that optimizing this model can derive Pareto optimal solutions. Pb-MORL is effective, easy to deploy, and is expected to be applied in complex systems, e.g., multi-energy management through preference feedback and adaptive autonomous driving policies for diverse situations.}, archive = {J_TASE}, author = {Ni Mu and Yao Luan and Qing-Shan Jia}, doi = {10.1109/TASE.2025.3589271}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18737--18749}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Preference-based multi-objective reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Regularization-based constrained distributed bilevel optimization over unbalanced graphs. TASE, 22, 18722--18736. (www)

  @article{ , abstract = {This paper focuses on solving a class of constrained distributed bilevel optimization problems over unbalanced graphs, where all agents are equipped with convex inner objective functions and strongly convex outer ones. The goal of solving the considered bilevel optimization problem is to minimize the global objective functions at both levels. In this paper, we propose a regularization-based distributed projected algorithm with row stochastic matrices and a time-varying regularization parameter $\theta _{t}$ . Furthermore, with the aid of the strong convexity of outer objective functions and the smoothness of two level objective functions, we establish that the proposed algorithm converges to the optimal solution with ${\mathcal {O}}(t^{-a+b})$ ( $a\in (0.5,1)$ and $b\in (0,0.5)$ ) and ${\mathcal {O}}(t^{-b})$ convergence rates from the perspectives of the outer and inner objective functions, respectively. Finally, we illustrate the effectiveness of the proposed algorithm by numerical simulations. Note to Practitioners—Constrained distributed bilevel optimization has garnered increasing attention in automation control and scientific computing, thanks to its wide range of applications in microgrids, wireless sensor networks, artificial intelligence, and other emerging industries. In real-world scenarios, bilevel optimization offers a broader modeling scope than general single-level optimization problems. Bilevel formulations can capture a variety of existing optimization challenges, including constrained nonlinear and ill-posed constrained optimization. However, this generality also introduces substantial analytical complexities. To address these issues, this paper presents an efficient distributed algorithm that leverages a regularization method and a mild assumption on the networked communication topology (specifically, a row-stochastic weighted matrix) for constrained bilevel optimization. Unlike conventional centralized algorithms and their distributed counterparts, which often rely on restrictive assumptions (such as balanced graphs on networked topologies), the proposed algorithm requires only weak conditions to enable each agent to cooperatively achieve the global optimum. Meanwhile, we establish convergence results for the decision variables and convergence rates with respect to the outer and inner objective functions. The validity and correctness of the proposed method are demonstrated through a sensor network problem and a real-world image recovery problem. Our future work will focus on designing distributed algorithms with improved convergence performance using accelerated techniques.}, archive = {J_TASE}, author = {Yongxiang Fu and Yuan Fan and Songsong Cheng}, doi = {10.1109/TASE.2025.3588911}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18722--18736}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Regularization-based constrained distributed bilevel optimization over unbalanced graphs}, volume = {22}, year = {2025}, }
• (2025). Movement primitive categorization balancing the learnability and adaptability. TASE, 22, 18709--18721. (www)

  @article{ , abstract = {Given the rapid advancement of robotic technologies, robots will eventually enter our daily lives, performing complex long-horizon tasks. Complex long-horizon tasks, such as furniture assembly, typically contain dozens of subtasks with various scenes. Although complex, assembly is based on several reusable movements. Movement Primitive (MP) is a promising framework for learning reusable movements from demonstrations and adapting the learned movements to the test scenes. The critical step in employing MP methods is categorizing the unlabeled demonstrations into different MPs. However, current MP methods focus on individual MP learning using manually selected demonstrations, neglecting categorization. Manual categorization of demonstrations is easy to fall into suboptimal. If the demonstrations within the same category are too similar, the learned MP cannot be adapted to task scenes with various obstacles. Conversely, a significant distance between demonstrations leads to the MP’s failure in learning. To this end, we propose the following principle for MP categorization: balance the Learnability and Adaptability. Following this principle, we introduce an optimal transportation (OT)-based theoretical framework and a practical solution utilizing an auto-encoder network. We obtain the lower threshold of learnability by OT. Then we increase the adaptability of MP until it reaches the lower threshold of learnability. For complex long-horizon task learning, we propose a balanced MPs-based learning framework that contains four modules, termed BaMPs. BaMPs achieved success rates of 100% and 80%, respectively, in the 12-step and 20-step tasks. Note to Practitioners—This work aims to learn multiple MPs capable of adaptation to complex long-horizon tasks from unlabeled demonstrations. Such tasks frequently consist of numerous subtasks, each distinguished by unique start, via, and target points, alongside specific obstacles, thereby posing considerable adaptability challenges for MPs. A critical step in learning MPs from unlabeled demonstrations is the categorization of these unlabeled demonstrations. However, current MP methods primarily focus on individual MP learning using manually selected demonstrations, which easily leads to suboptimal outcomes. To this end, we propose a principle for MP categorization: balance learnability and adaptability. In line with this principle, we introduce an OT-based theoretical framework and a practical implementation leveraging an autoencoder network. Furthermore, we introduce a balanced MPs-based complex long-horizon tasks learning framework, termed BaMPs. BaMPs comprises four modules: (a) MP categorization; (b) MP learning; (c) MP selection; and (d) MP editing.}, archive = {J_TASE}, author = {Xuanhui Xu and Mingyu You and Hongjun Zhou and Zhifeng Qian and Jinzhe Xue and Weisheng Xu and Bin He}, doi = {10.1109/TASE.2025.3588839}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18709--18721}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Movement primitive categorization balancing the learnability and adaptability}, volume = {22}, year = {2025}, }
• (2025). Self-attention enhanced dynamics learning and adaptive fractional-order control for continuum soft robots with system uncertainties. TASE, 22, 18694--18708. (www)

  @article{ , abstract = {Dynamics-based control offers a promising approach to exploring the motion potential of soft robots. However, inherently infinite degrees of freedom of these systems pose significant challenges for dynamics modeling, closely followed by the pressing robustness concerns arising from finite-dimensional approximations. This paper addresses these issues by proposing a physics-informed dynamics learning neural network and an adaptive fractional-order control for continuum soft robots. Specifically, a deep Lagrangian neural network is first developed with an embedded self-attention mechanism to enhance learning efficiency, accuracy, and data sensitivity. Subsequently, an adaptive fractional-order sliding mode controller is designed, leveraging the inherent historical memory properties of fractional calculus. This controller not only ensures robust shape control but also improves response speed and tracking accuracy. To further handle model discrepancies in the learned dynamics and external disturbances, a nonlinear disturbance observer is introduced to effectively estimate and compensate for lumped uncertainties, thereby ensuring reliable performance. Theoretical analysis confirms the closed-loop stability, while both simulation and experiment results validate the high dynamics fitting accuracy of the proposed network, as well as the robust and precise tracking capability of the fractional-order controller. Note to Practitioners—Soft robots offer great potential in unstructured or constrained environments owing to their compliance and adaptability. However, their high degrees of freedom and nonlinear behaviors make analytical modeling and robust control particularly challenging. Meanwhile, traditional closed-box learning methods often suffer from limited physical interpretability, reliability and extrapolability. This work presents a physics-informed dynamics learning framework combined with a fractional-order controller for soft robots. The dynamics learning network embeds physical priors to enhance model interpretability and extrapolability, while a self-attention mechanism improves data efficiency and modeling accuracy. Additionally, a disturbance observer is designed to estimate and compensate for model discrepancies and external disturbances, thereby contributing to the system’s robustness. Incorporating the observer’s outputs, the adaptive fractional-order controller further enhances closed-loop behavior by leveraging the memory properties of fractional calculus.}, archive = {J_TASE}, author = {Xiangyu Shao and Linke Xu and Guanghui Sun and Weiran Yao and Ligang Wu and Cosimo Della Santina}, doi = {10.1109/TASE.2025.3590174}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18694--18708}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-attention enhanced dynamics learning and adaptive fractional-order control for continuum soft robots with system uncertainties}, volume = {22}, year = {2025}, }
• (2025). Progressive-resolution policy distillation: Leveraging coarse-resolution simulations for time-efficient fine-resolution policy learning. TASE, 22, 18682--18693. (www)

  @article{ , abstract = {In earthwork and construction, excavators often encounter large rocks mixed with various soil conditions, requiring skilled operators. This paper presents a framework for achieving autonomous excavation using reinforcement learning (RL) through a rock excavation simulator. In the simulation, resolution can be defined by the particle size/number in the whole soil space. Fine-resolution simulations closely mimic real-world behavior but demand significant calculation time and challenging sample collection, while coarse-resolution simulations enable faster sample collection but deviate from real-world behavior. To combine the advantages of both resolutions, we explore using policies developed in coarse-resolution simulations for pre-training in fine-resolution simulations. To this end, we propose a novel policy learning framework called Progressive-Resolution Policy Distillation (PRPD), which progressively transfers policies through some middle-resolution simulations with conservative policy transfer to avoid domain gaps that could lead to policy transfer failure. Validation in a rock excavation simulator and nine real-world rock environments demonstrated that PRPD reduced sampling time to less than 1/7 while maintaining task success rates comparable to those achieved through policy learning in a fine-resolution simulation. Note to Practitioners—This paper is motivated by the issue of computation time in excavation simulation using soil particles. The behavior of real soil is highly complex, and approximating it at high resolution requires enormous computational costs. Therefore, existing soil simulators have focused on improving simulation accuracy while maintaining reduced computation time. This paper takes a different approach by focusing on the learning of control policies in excavation simulators and proposes a framework for reducing calculation time in such use cases. In this framework, a control policy is first learned in a low-resolution simulation, significantly reducing computation time. The learned policy is then transferred to a high-resolution simulation for retraining, thereby achieving an overall reduction in simulation time. Furthermore, to enable robust policy transfer across different resolutions, this paper discusses a stable policy distillation scheme and insights into resolution design. This approach enables the development of autonomous excavation systems without relying on expensive real-world data collection, improving the scalability and adaptability of autonomous excavation. Simulation experiments suggest that this framework significantly reduces training time compared to conventional policy learning approaches. However, real-world validation has so far been limited to simple excavation robots. Future research will explore applications to excavators and other machinery more suitable for real-world operations. Although this paper focuses on autonomous excavation, the proposed approach can also be extended to environments where increased simulation resolution critically impacts computation time, such as liquid and soft object manipulation.}, archive = {J_TASE}, author = {Yuki Kadokawa and Hirotaka Tahara and Takamitsu Matsubara}, doi = {10.1109/TASE.2025.3590068}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18682--18693}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Progressive-resolution policy distillation: Leveraging coarse-resolution simulations for time-efficient fine-resolution policy learning}, volume = {22}, year = {2025}, }
• (2025). Reachable set control of cyber-physical systems with markov jump parameters and hybrid attacks. TASE, 22, 18673--18681. (www)

  @article{ , abstract = {This study explores the reachable set issue for a category of Markov jump cyber-physical systems subjected to hybrid attacks. Firstly, based on the resilient event-triggered condition, a new lemma is proposed, which proves that the reachable set of the system exists under a certain degree of Denial-of-Service (DoS) attack. Secondly, considering the influence of false data injection (FDI) attacks and DoS attacks, a suitable controller is designed. Then, sufficient conditions for the existence of reachable sets under non-zero initial conditions were derived through the Lyapunov-Krasovskii (L-K) functional method and linear matrix inequality (LMI) techniques. Finally, the simulation results of the F-18 aircraft demonstrate the efficacy of the proposed method. Note to Practitioners—As the interaction between network security and physical systems becomes increasingly complex, especially when facing various attacks such as DoS attacks and FDI attacks, ensuring system stability has become an important issue. This study proposes a new method to address the issue of how to ensure that the system can transition from the initial state to the target state (i.e. reachable set problem) under the influence of hybrid attacks. The key contribution of this study is that even in the event of a system attack, a method can still be found to ensure that the system’s operation is not severely disrupted. The purpose of the study is to design a controller and combine it with a mathematical tool (L-K functional) to enable the system to respond effectively to these attacks and avoid overly conservative results caused by the attacks. For practical applications, this research can help fields that rely on networks and physical systems (such as unmanned aerial vehicles, industrial automation equipment, etc.) to improve their robustness against network attacks. For example, aircraft like the F-18 can still ensure safe flight and reach their designated location in the face of network interruptions or data tampering during flight. This method provides a new approach for designing more reliable control systems in practical engineering.}, archive = {J_TASE}, author = {Weihao Liu and Xiuyang Fan and Guoqiang Tan and Wen-Juan Lin}, doi = {10.1109/TASE.2025.3589031}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18673--18681}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reachable set control of cyber-physical systems with markov jump parameters and hybrid attacks}, volume = {22}, year = {2025}, }
• (2025). A security protection scheme for distributed state estimation based on double-watermarks. TASE, 22, 18659--18672. (www)

  @article{ , abstract = {Distributed sensor networks face security challenges since the state estimates between adjacent sensors may be tampered or monitored by attackers. Once the attack bypasses the detection mechanism and is integrated into the distributed state estimation, the performance of the system will be compromised. In this paper, the securities of distributed state estimation under confidentiality and integrity attacks are studied. A loophole in single watermarking protection strategy is revealed, thus the existing solution cannot balance the false alarm rate and missed alarm rate by the aid of the equipped Kullback-Leibler (K-L) divergence detector. To solve this problem, an improved scheme based on double-watermarks protection strategy is proposed. Under the condition that the designed watermarks are secure, it is infeasible for an attacker to launch such a covert injection attack which can evade K-L divergence detector. Therefore, a suitable detection threshold can be determined, to achieve both low false alarm rate and missed alarm rate. Furthermore, considering that the manner in which the tampered data is discarded weakens the performances of distributed state estimation, the feasibilities of extracting the original signals from the polluted ones under sparse attack and uniformity attack are investigated. A method for recovering the real estimates by indirectly reconstructing attack signals is provided, which is realized by solving an optimization problem. The advantage of distributed state estimation is thus preserved. Finally, the simulation experiments verify the effectiveness of the proposed scheme. Note to Practitioners—The motivation for this paper stems from the security challenges faced by distributed sensor networks. In order to identify the forged data shared by different nodes in the distributed network, the existing method is to add a protective watermark into the transmitted signal. This paper reveals a loophole of the single watermarking strategy, and gives a new solution based on double-watermarks framework. In this case, the attack detection performance is significantly improved, and an alarm will be more likely to be triggered by a real attack rather than noise. For the received polluted signals, this paper provides a method for recovering the real data by indirectly reconstructing the attack signals. Therefore, the performance of distributed sensor networks will not be compromised. Preliminary simulation experiments suggest that the proposed solution is feasible but it has not yet been tested in production. In future research, we will further explore the design of batch producing watermarks by chaotic control, to improve the efficiency of implementation.}, archive = {J_TASE}, author = {Qi He and Xiao-Jian Li}, doi = {10.1109/TASE.2025.3586303}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18659--18672}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A security protection scheme for distributed state estimation based on double-watermarks}, volume = {22}, year = {2025}, }
• (2025). TiCoSS: Tightening the coupling between semantic segmentation and stereo matching within a joint learning framework. TASE, 22, 18646--18658. (www)

  @article{ , abstract = {Semantic segmentation and stereo matching, respectively analogous to the ventral and dorsal streams in our human brain, are two key components of autonomous driving perception systems. Addressing these two tasks with separate networks is no longer the mainstream direction in developing computer vision algorithms, particularly with the recent advances in large vision models and embodied artificial intelligence. The trend is shifting towards combining them within a joint learning framework, especially emphasizing feature sharing between the two tasks. The major contributions of this study lie in comprehensively tightening the coupling between semantic segmentation and stereo matching. Specifically, this study makes three key contributions: (1) a tightly coupled, gated feature fusion strategy, (2) a hierarchical deep supervision strategy, and (3) a coupling tightening loss function. The combined use of these technical contributions results in TiCoSS, a state-of-the-art joint learning framework that simultaneously tackles semantic segmentation and stereo matching. Through extensive experiments on the KITTI, vKITTI2, and Cityscapes datasets, along with both qualitative and quantitative analyses, we validate the effectiveness of our developed strategies and loss function. Our approach demonstrates superior performance compared to prior arts, with a notable increase in mean intersection over union by over 9%. Note to Practitioners—TiCoSS is a robust and effective joint learning framework that can simultaneously tackle semantic segmentation and stereo matching tasks. This work aims to improve semantic segmentation performance by exploring the potential complementarity and tightening the coupling between these two tasks. In the future, we plan to further improve the efficiency of the framework, so as to enable its real-time performance on resource-constrained hardware.}, archive = {J_TASE}, author = {Guanfeng Tang and Zhiyuan Wu and Jiahang Li and Ping Zhong and Wei Ye and Xieyuanli Chen and Huimin Lu and Rui Fan}, doi = {10.1109/TASE.2025.3586286}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18646--18658}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TiCoSS: Tightening the coupling between semantic segmentation and stereo matching within a joint learning framework}, volume = {22}, year = {2025}, }
• (2025). Shortcut-enhanced multimodal backdoor attack in vision-guided robot grasping. TASE, 22, 18629--18645. (www)

  @article{ , abstract = {Integrating the Artificial Intelligence (AI) vision module into the robot grasping system can significantly improve its generalizability, thereby enhancing the efficiency of Human-Robot Interaction (HRI). However, the inherent lack of interpretability in AI also opens the gate to external threats. In this work, we reveal a novel safety risk in this vision-guided robot grasping system by proposing the Shortcut-enhanced Multimodal Backdoor Attack (SEMBA), which can manipulate the grasp quality score using the backdoor trigger leading to a misguided grasping sequence. The SEMBA may thus cause potentially hazardous grasping and pose a threat to human safety in HRI. Specifically, we initially present the Multimodal Shortcut Searching Algorithm (MSSA) to find the pixel value that deviates the most from the mean and standard deviation of the multimodal dataset, along with the pivotal pixel position for individual images. This will guarantee that the proposed attack is effective in complex, multi-class object scenarios. Next, based on MSSA, we devise the Multimodal Trigger Generator (MTG) to create diverse multimodal backdoor triggers and integrate them into the dataset, ensuring that our attack has the multimodality attribute. We conduct extensive experiments on the benchmark datasets and a cobot, showing the effectiveness of the proposed method both in the digital and physical worlds. Our demo videos are available in supplementary items. Note to Practitioners—Robot grasping systems are typically designed to be safe and reliable in HRI scenarios. However, integrating an AI-powered vision module into such systems can introduce substantial unpredictability, particularly when relying on third-party data. This work introduces a novel backdoor attack method to reveal a new safety risk in this vision-guided robot grasping system. Our method aims to mislead the robot into performing hazardous grasps by altering the grasping sequence. It emphasizes the attack’s effectiveness in complex, multi-class object scenarios and highlights its multimodal nature. To the best of our knowledge, this is the first study to present backdoor attacks in vision-guided robot grasping. Our approach paves the way toward future AI-powered visual grasping safety studies and provides valuable insights into building a more reliable and trustworthy vision-guided system for cobots.}, archive = {J_TASE}, author = {Chenghao Li and Ziyan Gao and Nak Young Chong}, doi = {10.1109/TASE.2025.3589764}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18629--18645}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Shortcut-enhanced multimodal backdoor attack in vision-guided robot grasping}, volume = {22}, year = {2025}, }
• (2025). Global asymptotic tracking under prescribed performance for a class of uncertain nonlinear strict-feedback systems with actuator faults. TASE, 22, 18617--18628. (www)

  @article{ , abstract = {This paper investigates global adaptive asymptotic tracking with prescribed performance control (PPC) for nonlinear systems subject to unknown nonparametric uncertainties, mismatched external disturbances, and potential actuator faults. The challenge lies in driving the tracking error to a prescribed region within a given time globally, achieving independent performance criteria for all errors under unknown nonlinear functions and any initial condition, while maintaining arbitrary settling time and accuracy specifications. Hence, we propose a robust global fault-tolerant control scheme using a barrier Lyapunov function and bound estimation. Key features of our approach include: 1) the use of heterogeneous tuning functions for independent error adjustment; 2) the ability to preassign global transient performance criteria off-line, independent of initial conditions, and high-order derivatives of the reference trajectory; 3) a constraint analysis scheme on account of proof by contradiction is introduced to demonstrate the inherent robustness to unknown system functions, disturbances and actuator faults. The theoretical results are demonstrated through a comparative experiment on a linear motor platform. Note to Practitioners—The escalating demands for product quality and production efficiency have necessitated stringent requirements for the performance of motion systems in contemporary industrial production equipment. PPC has demonstrated its effectiveness as a highly successful strategy for addressing this problem. However, achieving the objective of driving the tracking error to a prescribed region within a given time globally, satisfying independent performance criteria for all errors under unknown nonlinear functions, actuator faults, and any initial conditions, while simultaneously maintaining arbitrary settling time and accuracy specifications in the design of PPC, is a complex and challenging task. To tackle these challenges, this paper introduces a global asymptotic fault-tolerant tracking control method with prescribed performance. In contrast to existing approaches, this method ensures asymptotic output tracking with predefined settling time and accuracy for any initial states, despite actuator faults, unknown system nonlinearities, and external disturbances. Comparative experimental results validate the applicability of this method to practical gantry platform systems, illustrating its effectiveness and superiority over existing methods.}, archive = {J_TASE}, author = {Jue Wang and Dazhao Wang and Yanbin Liu and Huihui Pan and Weichao Sun}, doi = {10.1109/TASE.2025.3587643}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18617--18628}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Global asymptotic tracking under prescribed performance for a class of uncertain nonlinear strict-feedback systems with actuator faults}, volume = {22}, year = {2025}, }
• (2025). Multiple pursuers versus one evader reach-avoid differential games with asymmetric observations. TASE, 22, 18605--18616. (www)

  @article{ , abstract = {This paper considers a multiple-pursuer-one-evader reach-avoid (RA) differential game in a two-dimensional (2D) plane that is split by a straight line into a goal region and a play region. The evader aims to enter the goal region from the play region without being captured, while the pursuers try to intercept the evader in the play region. In terms of information structures, we assume that the evader has perfect information about the game, while the pursuers have uncertain observations of the evader’s location and speed. First, a robust estimation of the safe region (RESR) is constructed by computing the safe region boundary beliefs with respect to these uncertainties. Based on the RESR, the winning condition for the one-on-one RA game is analyzed, and an optimal interception strategy is derived. A proximity strategy is also proposed when the winning condition is not satisfied. Then, the concept of effective coalition is studied and its properties are proposed to enhance the robustness of the strategy. Additionally, a receding horizon cooperation algorithm with a strategy switching mechanism is designed for the pursuers to cooperatively capture the evader under distance-related observation uncertainties. Finally, simulations and experiments are conducted to validate the effectiveness of the proposed strategies. Note to Practitioners—This paper is motivated by practical considerations of the reach-avoid game, in which players have different capacities for observation. In most related works, pursuers and evaders are assumed to have perfect information about the game. Optimal strategies are derived from safe regions (SR) determined by perfect information. Our objective is to challenge this assumption and study the robust pursuit strategy from the perspective of the pursuers under observational uncertainty. We approach observational uncertainty in two ways. On the one hand, to obtain the SR of the evader, we construct the robust estimation of the safe region (RESR) and identify its $3\sigma $ principle boundary with theoretical guarantees. On the other hand, to mitigate the impact of observational uncertainty on pursuit strategies, we propose the concept of effective coalition, which can compress the RESR of the evader and capture the evader farther if acting individually. By combining these contributions, we design a receding horizon strategy for the pursuit coalition that is tailored for the autonomous multi-agent system in robotics, defense, and surveillance. At present, this method cannot be used directly in the multi-pursuer versus multi-evader game scenario. Future studies will investigate the scalability of this method.}, archive = {J_TASE}, author = {Hongwei Fang and Peng Yi and Di Deng and Bing He}, doi = {10.1109/TASE.2025.3587433}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18605--18616}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multiple pursuers versus one evader reach-avoid differential games with asymmetric observations}, volume = {22}, year = {2025}, }
• (2025). Universal low-complexity safe tracking control and its application to time-independent path following of nonholonomic mobile robots. TASE, 22, 18594--18604. (www)

  @article{ , abstract = {The existing shared control schemes and automatic control schemes are almost designed separately. Based on the designed constrained function and shared-constrained function, this paper proposes a universal control framework for high-order strict-feedback systems that can achieve safe tracking control with very low algorithm complexity regardless of whether the reference signal exceeds safety constraints, combining the flexibility of manual operating with the persistence and security of automatic control. A universal safe tracking control theorem is presented to prove the security of system operation. Furthermore, based on the proposed universal safe tracking control method, a novel time-independent path following control method with lane constraints is proposed for nonholonomic mobile robots by converting lane constraints into heading angle constraints, thus achieving automatic/auxiliary adaptive cruise control and lane keeping. Simulation and experimental results are provided to demonstrate the practical applicability of this method. Note to Practitioners—This work addresses the issue of the reference signal exceeding the safety constraint due to planner errors (in autonomous control) or human operator mistakes (in shared control). Compared to existing control methods, the proposed architecture is not only applicable to safe tracking control in both autonomous and shared control scenarios, but also features extremely low algorithmic complexity, making it highly suitable for practical deployment. Unlike traditional shared control systems, where human operators directly manipulate system inputs, the proposed method allows operators to control the system by adjusting the reference signal. This ensures that feedback control is always present, thereby reducing the difficulty of manipulation. Moreover, in contrast to common trajectory tracking methods, the proposed time-independent path following approach holds greater practical value, ultimately achieving adaptive cruise control and lane-keeping for mobile robots. To ensure repeatability, our codes are open sourced on github: https://github.com/Mudianrui/USC-and-app-in-NMRs.git. Experimental videos can be accessed through https://youtu.be/VrsISKkJOCA}, archive = {J_TASE}, author = {Dianrui Mu and Changchun Hua and Jiannan Chen and Yafeng Li and Lingchen Zhu and Fucai Liu}, doi = {10.1109/TASE.2025.3586138}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18594--18604}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Universal low-complexity safe tracking control and its application to time-independent path following of nonholonomic mobile robots}, volume = {22}, year = {2025}, }
• (2025). Estimated informed anytime search for sampling-based planning via adaptive sampler. TASE, 22, 18580--18593. (www)

  @article{ , abstract = {Path planning in robotics often involves solving continuously valued, high-dimensional problems. Popular informed approaches include graph-based searches, such as A*, and sampling-based methods, such as Informed RRT*, which utilize informed set and anytime strategies to expedite path optimization incrementally. Informed sampling-based planners define informed sets as subsets of the problem domain based on the current best solution cost. However, when no solution is found, these planners re-sample and explore the entire configuration space, which is time-consuming and computationally expensive. This article introduces Multi-Informed Trees (MIT*), a novel planner that constructs estimated informed sets based on prior admissible solution costs before finding the initial solution, thereby accelerating the initial convergence rate. Moreover, MIT* employs an adaptive sampler that dynamically adjusts the sampling strategy based on the exploration process. Furthermore, MIT* utilizes length-related adaptive sparse collision checks to guide lazy reverse search. These features enhance path cost efficiency and computation times while ensuring high success rates in confined scenarios. Through a series of simulations and real-world experiments, it is confirmed that MIT* outperforms existing single-query, sampling-based planners for problems in $\mathbb {R}^{4}$ to $\mathbb {R}^{16}$ and has been successfully applied to real-world robot manipulation tasks. A video showcasing our experimental results is available at: https://youtu.be/30RsBIdexTU Note to Practitioners—The motivation for this work stems from the challenges faced by existing informed path planners in high-dimensional, continuously valued environments, particularly when an initial feasible solution is difficult to find. Traditional asymmetric bidirectional planners rely on the best current solution to define problem subsets. When a lazy path has been found through lazy reverse search, these planners tend to re-sample and explore the entire problem space, which could hinder the path planning process. Our proposed MIT* algorithm addresses this issue by constructing an estimated informed set based on prior admissible solution costs before finding the initial solution. This estimated set helps to narrow the search area, thereby accelerating the initial convergence rate. MIT* also integrates an adaptive sampling strategy that dynamically adjusts based on the ongoing exploration process, enhancing the planner’s ability to efficiently navigate through challenging spaces. Furthermore, MIT* employs adaptive sparse collision checks, which guide the lazy reverse search that balances computational efficiency with accuracy in pathfinding. The proposed algorithm can be applied to industrial robots, humanoid robots, or service robots to achieve efficient path planning.}, archive = {J_TASE}, author = {Liding Zhang and Kuanqi Cai and Yu Zhang and Zhenshan Bing and Chaoqun Wang and Fan Wu and Sami Haddadin and Alois Knoll}, doi = {10.1109/TASE.2025.3590084}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18580--18593}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Estimated informed anytime search for sampling-based planning via adaptive sampler}, volume = {22}, year = {2025}, }
• (2025). Displacement control method of hydraulic servomotor based on swashplate moment disturbance compensation. TASE, 22, 18570--18579. (www)

  @article{ , abstract = {Hydraulic servomotor is a common energy-saving hydraulic power component that can switch between pump and motor operating conditions. Precise displacement control of hydraulic servomotors is crucial for minimizing throttling losses and enhancing overall efficiency. Nonetheless, the accuracy of hydraulic servomotor displacement control is negatively impacted by high-frequency uncertainty disturbances in swashplate moment. Traditional adaptive robust control methods have overlooked this factor, resulting in suboptimal displacement control effectiveness. To address this issue, a swashplate disturbance compensation robust control approach is introduced in this article, aiming to enhance the control accuracy of hydraulic servomotor displacement under high-frequency uncertainty disturbances. In the article, a mathematical model for the servomotor displacement adjustment system is formulated, and an average disturbance model for the moment of a swashplate is developed to account for periodic disturbances and uncertainties. Comparing experimental results of sine commands at different frequencies reveals that the proposed method achieves a substantial enhancement in displacement control accuracy. Specifically, the proposed method demonstrates a maximum improvement of 39.2% and 27.4% in displacement control accuracy compared to the PID method and traditional adaptive robust control, respectively. Note to Practitioners—The motivation of this article is to solve the displacement control problem of axial piston hydraulic servo motors, but it is applicable to other axial piston variable pumps and variable motors. There is a high frequency of pressure pulsation during the operation of variable hydraulic pumps and variable motors, which significantly affects the accuracy of the displacement control. However, traditional displacement control methods do not account for the pulsation inside the casing, leading to lower control accuracy. To address this issue, this article proposes a new control method aimed at enhancing the accuracy of displacement control in hydraulic servo motors. The proposed method involves modeling the disturbance torque experienced during the swashplate movement of hydraulic servo motors, analyzing the variable mechanism, and incorporating compensation for this disturbance into the control signal. This approach ensures precise displacement control during the operation of hydraulic servo motors and enhances the operational efficiency. The newly proposed control method has undergone rigorous stability testing, and initial experiments have demonstrated its feasibility. In comparison to traditional control methods, the accuracy of displacement control has been enhanced by approximately 10% to 40%. However, a limitation of the proposed control method lies in the restricted number of sensors located on the housing of the hydraulic servo motor. As a result, the method lacks full sensor information and can only make use of estimated values from the state observer in place of actual measurements. In future studies, efforts will be directed towards resolving the sensor acquisition challenge to further augment the accuracy of displacement control.}, archive = {J_TASE}, author = {Zhenyu Wang and Xiaochao Liu and Zhongyi Qiu and Rui Nie and Xinghua Chen and Zongxia Jiao}, doi = {10.1109/TASE.2025.3586067}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18570--18579}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Displacement control method of hydraulic servomotor based on swashplate moment disturbance compensation}, volume = {22}, year = {2025}, }
• (2025). Real-time trajectory tracking of a piezoelectric microrobot using asynchronous fusion of vision cameras and laser sensors. TASE, 22, 18560--18569. (www)

  @article{ , abstract = {This study developed a multisensor asynchronous fusion method for vision cameras and laser sensors for real-time tracking of a millimeter-scale piezoelectric microrobot with high sampling rates. The design, manufacture, and realization of the proposed microrobot using a monolithic integrated manufacturing process is realized with a size of $36 \times 36 \times 18$ mm and a weight of 4.9 g. Two local estimators for the vision camera and laser sensors are established independently for 2-D tracking of the millimeter-scale microrobot. An asynchronous fusion method for the camera and laser sensors is developed by optimally weighting the two local estimators in a linear minimum variance framework using the Lagrange multiplier method at the synchronized update time. High-precision 2-D trajectory tracking of the proposed microrobot was achieved with an average velocity of 301.6 mm/s and a sampling rate of up to 4000 Hz. Experiments tracking different trajectories of the microrobot at different velocities and sampling rates demonstrated the effectiveness of the proposed fusion method. Note to Practitioners—This paper aims to enhance the real-time trajectory tracking accuracy of microrobots. The parallel microrobot is fabricated by a monolithic integrated manufacturing process, and has a small dimension of $36 \times 36 \times 18$ mm. An asynchronous fusion method of the camera and laser sensors is proposed by optimally weighting in a linear minimum variance framework using the LaGrange multiplier method at the synchronized update time. Furthermore, high-precision 2-D trajectory tracking of the proposed microrobot is realized with an average velocity of 301.6 mm/s and a sampling rate of up to 4000 Hz. Experiments demonstrate the effectiveness of the proposed fusion method. It is of great importance to researchers and engineers in trajectory tracking and positioning control of microrobots with high precision and sampling rates.}, archive = {J_TASE}, author = {Junqiang Lou and Kantao Zhang and Jiaxu Shen and Tehuan Chen and Yuguo Cui and Chao Xu and Haojian Lu}, doi = {10.1109/TASE.2025.3585494}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18560--18569}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time trajectory tracking of a piezoelectric microrobot using asynchronous fusion of vision cameras and laser sensors}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy resilient control against unknown deception injection attacks for pure-feedback nonlinear switched systems with intermittent actuator faults. TASE, 22, 18549--18559. (www)

  @article{ , abstract = {This paper is concerned with the adaptive resilient control problem of the pure-feedback nonlinear switched systems with intermittent actuator faults. First, by employing the mean value theorem, the pure-feedback systems under consideration are converted into a category of strict-feedback nonlinear switched systems. Furthermore, by designing new coordinate transformations and Lyapunov functions, an improved adaptive resilient control design is developed, which effectively addresses the issues stemming from sensor deception attacks that obscure the original system information and involve unknown time-varying attack gains. To compensate for the intermittent failures of the actuator, a new adaptive fuzzy resilient controller is developed by co-designing the last virtual control function. Compared to the existing results, the proposed control method not only addresses the resilient control issue of pure-feedback nonlinear switched systems with deception injection attacks but also tackles the control problem of infinite actuator failures. At last, the results from the simulation substantiate the efficiency of the constructed control algorithm. Note to Practitioners —In various industries, the challenge of adaptive fuzzy control in nonlinear systems is present in systems like chaotic systems, polarization processes, mechanical systems, diffusion waves, electrical circuits, and more. As these systems function within a network environment, security issues are a critical concern. Given the presence of unknown deception injection attacks, unknown nonlinear functions, and intermittent actuator faults all at once within these systems, achieving the desired security control objectives becomes a complex and significant task. To address this, a new adaptive fuzzy resilient controller is developed by co-designing the last virtual control function. This approach provides a practical solution for industrial applications.}, archive = {J_TASE}, author = {Jipeng Zhao and Guang-Hong Yang}, doi = {10.1109/TASE.2025.3589211}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18549--18559}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy resilient control against unknown deception injection attacks for pure-feedback nonlinear switched systems with intermittent actuator faults}, volume = {22}, year = {2025}, }
• (2025). Cooperative path planning for heterogeneous UAV swarms: A stackelberg game approach. TASE, 22, 18531--18548. (www)

  @article{ , abstract = {The coordinated operations of Stealth Unmanned Aerial Vehicle (SUAV) and Swarming Drones (SD) have demonstrated formidable power in the military domain. Efficient path planning is a critical technology that enhances combat effectiveness. This paper proposes a game-theoretic optimization approach to achieve cooperative penetration and target search path planning for swarm UAVs. Multi-task and multi-objective optimization models are developed for complex scenarios whose optimal solutions are NP-hard. Consequently, SUAV and SD are defined as leader and followers, respectively. The formulated Stackelberg game model enables distributed intelligent decision-making for SUAV and SD. We theoretically prove that by selecting an appropriate potential function, subgames within the leader-level and followers-level become an Ordinal Potential Game (OPG) with a Nash equilibrium, thereby ensuring the existence of a Stackelberg Equilibrium (SE) through leader-follower interactions. We propose a Gradient-based Hierarchical Learning and Optimization Algorithm (GHLOA) to achieve SE. At the leader-level, a Stochastic Gradient Ascent (SGA) algorithm optimizes the penetration path for SUAV, while in the followers-level, we demonstrate that the designed Hybrid Learning-based Multimodal Adaptive Pigeon-Inspired Optimization (HLMAPIO) algorithm converges with probability one to the suboptimal solution for each SD. Numerical results indicate that our approach is suboptimal, scalable, and fast adaptable to dynamical scenarios, and it outperforms the state-of-the-art techniques. Note to Practitioners—With the rapid advancement of artificial intelligence technologies, the coordinated operations of stealth unmanned aerial vehicle (SUAV) and swarming drones (SD) have demonstrated immense potential in modern high-tech warfare. This paper investigates a heterogeneous UAV swarm system comprising a single SUAV and several, or even dozens of, SDs deployed in complex, dynamic scenarios with moving threats during the execution of diverse tasks. Our core concept is to propose a game-theoretic swarm intelligence decision-making framework that optimizes the rewards and costs of swarm members. This approach effectively mitigates potential conflicts, ensures suboptimal and collision-free paths, and enhances the scalability, adaptability, and robustness of the swarm system in dynamic environments. Notably, our method theoretically guarantees system stability and solution suboptimality. Numerical results show that the proposed approach significantly improves the swarm system’s overall utility while ensuring the safety of individual UAVs. Even when the number of UAVs and targets increase by a factor of five, the computational time remains within an acceptable range. Future work will focus on integrating game theory with reinforcement learning to explore decision-making and autonomous planning technologies for multi-objective and multi-task scenarios across larger-scale scenarios.}, archive = {J_TASE}, author = {Zhe Zhang and Ju Jiang and Keck Voon Ling and Xinhua Wang and Wen-An Zhang}, doi = {10.1109/TASE.2025.3585946}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18531--18548}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative path planning for heterogeneous UAV swarms: A stackelberg game approach}, volume = {22}, year = {2025}, }
• (2025). Feasibility of optimal formation under multiplicative noise with energy and time constraints. TASE, 22, 18522--18530. (www)

  @article{ , abstract = {We investigate the problem of distributed optimal formation control in Multi-Agent Systems (MAS) affected by multiplicative noise, with energy and time constraints. We develop a globally optimal distributed formation control algorithm to achieve the expected formation while minimizing a comprehensive cost function that includes formation errors, energy consumption, and network costs. Additionally, we derive lower bounds for the achievable completion time and the required energy levels, expressed in terms of the convergence error, and the second smallest and largest eigenvalues of the Laplacian matrix. These bounds set practical criteria for formation tasks with constraints on energy and time. Finally, we provide simulation results to validate the effectiveness of the proposed approach. Note to Practitioners—In applications of autonomous aerial vehicles (AAVs), autonomous ground vehicles (AGVs), and maritime robotics, maintaining formation while balancing energy and time constraints is critical. This challenge is further complicated in dynamic environments with noise and disturbances. This paper presents a distributed formation control algorithm to address these challenges. Moreover, we derive the minimum energy and time required for achieving formation, providing a practical framework for implementation. Practitioners can use the proposed results to assess whether the initial energy reserves of all vehicles or robotic systems are sufficient to achieve the task within the prescribed time frame and ensure that the formation task is completed with optimal energy efficiency. This method is particularly useful for energy-sensitive applications, such as long-duration AAV missions and robotic search-and-rescue.}, archive = {J_TASE}, author = {Chunxiang Jia and Fei Chen}, doi = {10.1109/TASE.2025.3587726}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18522--18530}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Feasibility of optimal formation under multiplicative noise with energy and time constraints}, volume = {22}, year = {2025}, }
• (2025). ESSDM: An enhanced sparse swarm decomposition method and its application in multi-class motor imagery-based EEG-BCI system. TASE, 22, 18509--18521. (www)

  @article{ , abstract = {Electroencephalogram (EEG)-based motor imagery (MI) (MI-EEG) decoding has established a novel experimental paradigm in brain-computer interface (BCI) applications that offer effective treatment for stroke paralyzed patients. However, existing MI-EEG-based BCI systems introduce deployment issues because of nonstationary EEG signals, suboptimal features, and limited multiclass scalability. To tackle these issues, we propose an enhanced sparse swarm decomposition method (ESSDM) based on selfish-herd optimization and sparse Fourier transform (FT) spectrum to solve the issue of choice of uniform decomposition and hyperparameters in swarm decomposition, and further applied to enhance MI-EEG classification. ESSDM adopts improved swarm filtering to automatically deliver optimal frequency bands in the sparse FT spectrum with optimized hyperparameters to extract dominant oscillatory components (OCs) that significantly enhance MI activation-related sub-bands. In addition, new fitness criteria has been designed based on the Kullback-Leibler divergence distance from the spectral kurtosis of the obtained modes to select hyperparameters that optimize decomposition effect, avoid excessive iterations, and provide fast convergence with optimal modes. Further, fused time-frequency graph (FTFG) features have been derived from computed time-frequency representation to find cross-channel mutual spectral information. The experimental results on the 2-class BCI III-4a, 2-class OpenBMI, and 4-class BCI IV-2a datasets reveal that the proposed framework based on FTFG features with capsule neural network (CapsNet), ESSDM-FTFG-CapsNet outperformed other existing methods in specific-subject, cross-subject, and cross-session scenarios. Note to Practitioners—This paper addresses the challenges in developing practical BCI systems using MI-EEG signals, specifically focusing on the accurate extraction of OCs corresponding to MI activation patterns from any multichannel MI-EEG signals and the subsequent design of optimal features for automated MI-EEG classification. This paper suggests a novel decomposition technique to extract optimal OCs for efficient feature extraction, and an automated BCI framework is further developed for multiclass MI tasks classification by considering practical testing scenarios. The proposed BCI frameworks enhance MI-EEG task detection accuracy and demonstrates promising improvement in practical BCI system for clinical practices. However, the presented BCI system approach involve limited number of MI-EEG task classes in the cross-subject, cross-dataset, and cross-session validation. In future studies, the presented approach can be studied on more number of MI-EEG task problems in diverse cross-day datasets for achieving generalized MI-EEG BCI model.}, archive = {J_TASE}, author = {Shailesh Vitthalrao Bhalerao and Ram Bilas Pachori}, doi = {10.1109/TASE.2025.3585778}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18509--18521}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ESSDM: An enhanced sparse swarm decomposition method and its application in multi-class motor imagery-based EEG-BCI system}, volume = {22}, year = {2025}, }
• (2025). Multi-objective charging optimization of lithium-ion batteries considering electrical, thermal, and aging behaviors using deep reinforcement learning. TASE, 22, 18493--18508. (www)

  @article{ , abstract = {Lithium-ion battery charging involves many factors such as electricity, heat, aging, etc. Shortening the charging time of lithium-ion batteries while limiting aging and temperature rise is an urgent issue that needs to be addressed. In order to solve this problem, the pseudo-two-dimensions (P2D) model is used to describe the electrical behavior of batteries. Then, a parameter identification method is proposed using a deep deterministic policy gradient for P2D model. Besides, a multi-objective charging model considering time, health loss, and temperature rise is constructed by combining P2D, equivalent heat, and aging mechanism models. A multi- objective charging optimization strategy is proposed using Deep Q-Network algorithm. Under the condition of strong coupling of multiple parameters in the P2D model, the identification accuracy is improved by more than 20% compared to traditional optimization algorithms. Under fast charging conditions, the charging time decreases by 19%, and the health loss only increased by 2.5%. Finally, the impact of multi-stage current variance on battery health loss is discussed. Note to Practitioners—The core of this work is to solve the charging problem of lithium batteries. This paper proposes a new intelligent charging algorithm based on deep reinforcement learning. This method uses a multi-stage constant current charging algorithm as the framework, taking into account the strong adaptability of charging time, temperature, and battery aging to different charging scenarios. In this work, a multi-objective optimization function is constructed using the thermal model, aging model, and electrochemical model of the battery. Then, deep reinforcement learning is used to identify the parameters of the electrochemical model of the battery, which ensures the accuracy of subsequent calculations. Finally, a series of solution sets are obtained by combining neural networks with multi-objective optimization functions. Preliminary comparative experiments have shown that this method is feasible and has higher advantages in reducing charging time and battery aging compared to traditional charging methods, but its adaptability to battery aging is poor. In future research, we plan to propose an adaptive charging scheme by changing the objective function.}, archive = {J_TASE}, author = {Xiang Dong and Huahong Xv and Tianhong Pan and Xinghua Liu and Jiaqiang Tian and Peng Wang}, doi = {10.1109/TASE.2025.3588660}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18493--18508}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-objective charging optimization of lithium-ion batteries considering electrical, thermal, and aging behaviors using deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Deep-BrownConrady: Prediction of camera calibration and distortion parameters using deep learning and synthetic data. TASE, 22, 18481--18492. (www)

  @article{ , abstract = {This research addresses the challenge of camera calibration and distortion parameter prediction from a single image using deep learning models. The main contributions of this work are: (1) demonstrating that a deep learning model, trained on a mix of real and synthetic images, can accurately predict camera and lens parameters from a single image, and (2) developing a comprehensive synthetic dataset using the AILiveSim simulation platform. This dataset includes variations in focal length and lens distortion parameters, providing a robust foundation for model training and testing. The training process predominantly relied on these synthetic images, complemented by a small subset of real images, to explore how well models trained on synthetic data can perform calibration tasks on real-world images. Traditional calibration methods require multiple images of a calibration object from various orientations, which is often not feasible due to the lack of such images in publicly available datasets. A deep learning network based on the ResNet architecture was trained on this synthetic dataset to predict camera calibration parameters following the Brown-Conrady lens model. The ResNet architecture, adapted for regression tasks, is capable of predicting continuous values essential for accurate camera calibration in applications such as autonomous driving, robotics, and augmented reality. Note to Practitioners—This paper introduces a deep learning approach to predict camera calibration and distortion parameters directly from a single image, addressing the limitations of traditional methods that require structured calibration objects and multiple images. Using synthetic datasets generated with a simulation platform, the model predicts essential parameters such as field of view, principal points, and Brown-Conrady distortion coefficients. A key innovation is incorporating image size into the learning process, enabling the model to generalize well to real-world scenarios. This approach simplifies the calibration process, making it suitable for dynamic and unstructured environments, such as autonomous driving and robotics, where traditional calibration methods are not feasible. As an important result, the proposed method enables the use of synthetic data to overcome data scarcity in real-world applications by adapting to the camera parameters of the underlying physical system. By leveraging synthetic data and deep learning, the method offers a modern, flexible alternative that practitioners can adopt to enhance camera calibration workflows in real-world applications. Extensive experiments are reported that showcase a successful solution based on the famous Brown-Conrady camera lens model and are validated on a real-world dataset. We believe a similar methodology can be used and extended as future work to enable camera parameter estimation and the use of synthetic data for other camera models.}, archive = {J_TASE}, author = {Faiz Muhammad Chaudhry and Jarno Ralli and Jerome Leudet and Fahad Sohrab and Farhad Pakdaman and Pierre Corbani and Moncef Gabbouj}, doi = {10.1109/TASE.2025.3588584}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18481--18492}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep-BrownConrady: Prediction of camera calibration and distortion parameters using deep learning and synthetic data}, volume = {22}, year = {2025}, }
• (2025). A mechano-electromagnetic hybrid actuation system for multimodal control of magnetic microrobots. TASE, 22, 18470--18480. (www)

  @article{ , abstract = {The motion of magnetic microrobots relies on an external magnetic field. Current electromagnetic propulsion systems are often complex in structural design, challenging in control strategies, and associated with high energy consumption. In this study, we introduce a mechano-electromagnetic hybrid actuation system that employs only a pair of Helmholtz coils to generate an oscillating uniform magnetic field, combined with a rotating platform to modify the field’s direction, thereby enabling the control of microrobot speed and direction. The properties of this hybrid actuation system were modeled and validated to analyze the magnetic field distribution within the workspace. The design and actuation of two microrobots are described in detail. Furthermore, a visual feedback closed-loop control strategy for the system was developed. A series of navigation control experiments were conducted, demonstrating that the proposed hybrid actuation system achieves effective microrobot actuation and precise control while reducing manufacturing costs and simplifying control strategies. The proposed novel mechano-electromagnetic hybrid structure contributes to the field of electromagnetically driven systems and magnetically controlled microrobots. Note to Practitioners—This paper was motivated by the importance of electromagnetic field in the actuation and control of magnetic microrobots which have shown significant potential in medical and bioengineering applications. In this paper, we propose a mechano-electromagnetic hybrid actuation system that utilizes only a pair of Helmholtz coils to generate an oscillating uniform magnetic field, with a rotating platform to alter the magnetic field’s direction, thereby controlling the motion speed and direction of microrobots. The results demonstrate that the proposed system achieves effective microrobot actuation and precise control with reduced manufacturing costs and simpler control strategies.}, archive = {J_TASE}, author = {Hao Yang and Dongqin Xu and Ying Li and Qingwei Li and Yixin Liu and Fuzhou Niu and Lining Sun}, doi = {10.1109/TASE.2025.3588504}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18470--18480}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A mechano-electromagnetic hybrid actuation system for multimodal control of magnetic microrobots}, volume = {22}, year = {2025}, }
• (2025). Automatic generation control based on brain-inspired continuous return distribution algorithm. TASE, 22, 18458--18469. (www)

  @article{ , abstract = {The increased penetration of renewable energy has triggered significant random disturbances that affect the performance of automatic generation control, posing a challenge to the stability of novel power systems. Reinforcement learning based on Markov stochastic processes has advantages in obtaining stochastic optimal solutions and has gradually been explored and applied by scholars to automatic generation control. Among these, the algorithms based on the Q learning framework are dominant. However, Q learning and its derivative algorithms suffer from the problem of Q value overestimation due to the discrete action space, which will be continuously magnified in high-dimensional space, leading to suboptimal solutions of the algorithm. Meanwhile, the training instability caused by improper gradient boundary setting of the algorithm, resulting in slow convergence. Therefore, this paper proposes a brain-inspired continuous return value distribution algorithm to obtain the optimal coordination of automatic generation control in a multi-area power grid. The proposed algorithm models the continuous return value as a Gaussian distribution through brain-inspired distributed learning and uses the variance of the value function to solve the problem of the Q value being continuously amplified in high-dimensional space. At the same time, a variance-based critical gradient adjustment strategy is introduced to replace fixed boundary clipping to avoid instability caused by improper boundary setting of the algorithm. The effectiveness of the proposed algorithm is verified by simulating a two-area automatic generation control system and a four-area automatic generation control system based on the Central China Power Grid. Compared with other reinforcement learning algorithms, it has better control performance, smaller frequency deviation, and faster convergence speed. Note to Practitioners—This paper proposes a novel model-free reinforcement learning algorithm for automatic generation control in novel power systems. The proposed algorithm overcomes the insufficient control accuracy and instability of existing reinforcement learning algorithms in novel power systems. After pre-training, it can be put into use in real time to maintain the frequency stability of power systems under strong random disturbances.}, archive = {J_TASE}, author = {Lei Xi and Kailun Wei and Wentao Wang and Yu Shi and Haochao Huang}, doi = {10.1109/TASE.2025.3587790}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18458--18469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic generation control based on brain-inspired continuous return distribution algorithm}, volume = {22}, year = {2025}, }
• (2025). Transferring common model parameters from chirp-modulated to steady-state visual evoked potentials for calibration-efficient BCIs. TASE, 22, 18443--18457. (www)

  @article{ , abstract = {While calibration improves the information transfer rate (ITR) of steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs), the calibration process corresponding to all stimuli remains time-consuming and labor-intensive. Transfer learning has been applied to reduce the requirement for calibration data. However, a considerable amount of source domain stimulus data is still required to train the transferable model parameters. To further reduce the calibration time, this study proposes a cross-stimulus transfer learning method that transfers the common impulse response and spatial filter from a single chirp-modulated visual evoked potential (Chirp-VEP) To SSVEPs (CTSS). First, a hypothesis about the common impulse response in Chirp-VEP is made, and the corresponding chirp impulse is constructed. Then, based on the linear superposition theory and the common spatial filter, a least-squares problem is formulated between the Chirp-VEP and the reconstructed Chirp-VEP template to learn the common impulse response and spatial filter. Finally, these common model parameters are transferred to SSVEP-based BCIs for target identification. For performance evaluation, both offline and online experiments were conducted on the custom-developed Chirp-based and SSVEP-based BCI systems. Experimental results showed that Chirp-VEP and SSVEPs exhibited similar impulse responses and spatial filters. In online spelling tasks, the proposed method achieved an information transfer rate (ITR) of $156.39~\pm ~37.11$ bits/min with a calibration time of 4.5 s to achieve comparable performance to state-of-the-art (SOTA) methods (tlCCA: $156.30~\pm ~44.81$ bits/min with 50 s calibration time, CIR: $155.39~\pm ~36.89$ bits/min with 20 s calibration time). These results demonstrate that the proposed method facilitates the practical implementation of SSVEP-based BCIs in real-world applications. The source code and dataset are available at https://github.com/xxxb/CTSS Note to Practitioners—This study is motivated by the lack of efficient calibration methods in multi-command steady-state visual evoked potential (SSVEP)-based brain–computer interfaces (BCIs), which are widely used in home automation and clinical rehabilitation scenarios. The conventional use of SSVEP signals for calibration leads to low bandwidth utilization and prolonged calibration time. In this study, we propose a novel calibration method that transfers common model parameters from a single Chirp-modulated visual evoked potential (Chirp-VEP) signal to multi-command SSVEP-based BCIs. The multi-frequency nature of the Chirp-VEP signal enables the extraction of shared features between Chirp-VEP and SSVEPs, thereby reducing the overall calibration time. The results of the experiments showed that the proposed calibration method achieved comparable classification accuracy to state-of-the-art methods with minimal calibration time, ultimately contributing to the development of user-friendly and efficient EEG-based control systems. Future work will explore the integration of Chirp-VEPs with various visual evoked potential (VEP) paradigms to support the development of diverse BCI applications, enhancing their practicality and accessibility in real-world settings.}, archive = {J_TASE}, author = {Bang Xiong and Bo Wan and Jiayang Huang and Pengfei Yang}, doi = {10.1109/TASE.2025.3587765}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18443--18457}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transferring common model parameters from chirp-modulated to steady-state visual evoked potentials for calibration-efficient BCIs}, volume = {22}, year = {2025}, }
• (2025). Scaled consensus of multiagent systems with unknown input delay and disturbance. TASE, 22, 18431--18442. (www)

  @article{ , abstract = {This paper addresses the scaled consensus of multi-agent systems (MAS) with unknown input delays and disturbances. First, a distributed scaled predictive controller framework with disturbance predictive compensation is established for dealing with both unknown input delays and disturbances. Second, an auxiliary loop for unknown disturbance predictive estimation is developed based on delay input predictive transformation technique. Moreover, a disturbance-compensate based distributed predictive control approach and adaptive switching mechanism, which is for matching the value of estimation delay with the one of real input delay, are developed such that the closed-loop MAS achieve the scaled consensus. A simulation example is provided to demonstrate the effectiveness of the proposed method. Note to Practitioners—This work presents a novel control framework for multi-agent systems (MAS) with unknown input delays and disturbances, integrating distributed adaptive delay prediction and disturbance compensation. By introducing a distributed scaled protocol loop and a dirty derivative filter, the proposed approach ensures practical delay estimation and disturbance mitigation. An adaptive switching mechanism combined with Lyapunov stability analysis guarantees the accuracy of delay estimation and enables the MAS to achieve scaled consensus. These contributions provide a robust and practical solution for MAS control in complex environments with uncertain dynamics.}, archive = {J_TASE}, author = {Yang Liu and Yingchun Wang and Huaguang Zhang and Shujuan Yang}, doi = {10.1109/TASE.2025.3587667}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18431--18442}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scaled consensus of multiagent systems with unknown input delay and disturbance}, volume = {22}, year = {2025}, }
• (2025). Power system short-term voltage stability assessment and margin estimation via neural barrier function. TASE, 22, 18416--18430. (www)

  @article{ , abstract = {Short-term voltage stability (STVS) has emerged as a critical concern in modern power systems due to increasingly complex transient post-fault dynamics. Effectual STVS assessment and margin estimation are essential. Existing data-driven STVS analysis approaches tackle the flexibility issue; however, they often lack a strong mechanism basis. To address these issues, this paper introduces a novel data-mechanism-integrated framework utilizing neural network-based barrier functions (NBFs) for STVS assessment and margin estimation. In particular, we propose a neural barrier certification for STVS analysis that is trained from post-fault response trajectories and constrained by LaSalle’s invariance principle, ensuring theoretical validity without requiring explicit dynamic models. To enhance the applicability of the proposed methodology in large-scale systems with limited observability, a state expansion approach by reconstructing latent system dynamics via system energy is introduced in this paper. Additionally, we raise an NBF-based stability margin estimation approach via a fast binary search method, realizing flexible and timely STVS margin estimation. The proposed methodology preserves the interpretability of mechanism-based methods while maintaining flexibility and computational efficiency. The flexibility and effectiveness of the proposed NBF-based framework are demonstrated via numerous tests on the two-bus system and the IEEE 39-bus system. Note to Practitioners—This paper is motivated by the problem of short-term voltage stability (STVS) analysis in power systems, but it also can be applied to other safety-critical assessment or decision-making tasks in practical systems. STVS is a critical concern in modern power systems since the enlarging scales bring more complex post-fault dynamics to the power system. Existing STVS analysis approaches either rely on accurate system modeling and simulation that limits their flexibility or employ data-driven techniques that weaken their reliability and interpretability. This paper develops a novel data-mechanism-integrated framework utilizing neural network-based barrier functions (NBFs) for STVS assessment and stability margin estimation. In our work, the post-fault power system is modeled as a nonlinear system with a forward-invariant safe set defined in state variable space. We then introduce a neural network to approximate the barrier function of the post-fault system, which is essential for determining the critical safety status of the post-fault system. To enhance the applicability of the proposed methodology in practical power systems with limited observability, we raise a state variable expansion approach based on system energy. We also develop a fast binary search method for estimating the stability margin, i.e., the degree of system stability. Experimental results demonstrate the effectiveness of the proposed framework. This work provides a fast, reliable, and flexible STVS analysis approach for practical power systems. In future research, we will address the NBF-based stability analysis and safe decision-making for emergency control in power systems.}, archive = {J_TASE}, author = {Congbo Bi and Di Liu and Ximing Zhang and Siting Zhu and Shiyang Li and Chao Fu and Chao Lu}, doi = {10.1109/TASE.2025.3587536}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18416--18430}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Power system short-term voltage stability assessment and margin estimation via neural barrier function}, volume = {22}, year = {2025}, }
• (2025). AffordStruct: Weakly supervised affordance grounding based on spatial interaction and knowledge-aware. TASE, 22, 18405--18415. (www)

  @article{ , abstract = {Affordance grounding refers to identifying the interactive regions of an object that enable it to perform a specific function, which is essential for effective robot interaction with complex environments. Current methods face challenges of incomplete localization and unclear distinction of affordance regions. To address these challenges, we propose a novel framework called AffordStruct, which draws on human experiential knowledge to explore affordance structures from the perspectives of spatial configuration and semantic relationships. Specifically, we first compute multi-order spatial interactions from the spatial configuration perspective to enhance long-range pixel dependencies, thereby reinforcing the network’s sensitivity to affordance-relevant regions and suppressing irrelevant areas. Besides, we design a pyramid hierarchical chain-of-thought prompting to guide the large language model in reasoning about affordance structural attributes. An attention-based fusion strategy is then employed to perform multimodal fusion with egocentric images that contain only the target objects, enabling the model to better perceive affordance structures at the semantic level. Finally, based on the local knowledge transfer mechanism, the affordance knowledge extracted from the exocentric view containing human-object interactions is transferred to the multimodal egocentric view. Extensive experiments on public datasets and robotic grasping tasks demonstrate that our method outperforms existing state-of-the-art affordance grounding models. Note to Practitioners—This paper was motivated by the problem of incomplete and unclear distinction of affordance areas in affordance grounding. In affordance grounding, affordance knowledge is typically learned by observing human-object interactions, while traditional methods often overlook affordance structural information. This paper proposes a novel method that first explores affordance spatial interactions to preserve affordance structures at the spatial configuration level. Then, a prompting strategy is designed to guide the large language model in reasoning about affordance-related semantic structures. These semantic insights are fused with visual information through multimodal integration to enhance affordance perception from a semantic relationship perspective. Experimental results demonstrate the superiority of our method. Furthermore, we deployed the method on a robot to evaluate its practical applicability. The results show that our approach enables the robotic arm to accurately grasp the functional regions of objects.}, archive = {J_TASE}, author = {Shiyu Wang and Shanyi Zhang and Fengtao Sun and Wenbai Chen and Peiliang Wu}, doi = {10.1109/TASE.2025.3586983}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18405--18415}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AffordStruct: Weakly supervised affordance grounding based on spatial interaction and knowledge-aware}, volume = {22}, year = {2025}, }
• (2025). Adaptive visual servo of soft robot with interaction estimation and compensation. TASE, 22, 18395--18404. (www)

  @article{ , abstract = {Soft robots face significant control challenges due to the coupling effects of strain and external forces, particularly with external interactions that change their original model. This paper proposes a vision-based controller integrated with an adaptive algorithm to estimate and compensate for external contact disturbances in such environments. Specifically, the adaptive law can online estimate unknown camera parameters, eliminating the need for costly and environment-specific calibration procedures. In parallel, a contact disturbance estimation strategy is introduced to model and compensate for real-time interaction effects without the foreknowledge of the constitutive equation of the soft mechanism. The adaptive algorithm is seamlessly integrated into the adaptive image-based visual servo (IBVS) controller, allowing simultaneous calibration and contact compensation during task execution. We validated the algorithm on a tendon-driven, octo-articular soft robotic manipulator prototype. The experimental results demonstrated the algorithm’s ability to position the end-effector even under external interactions, with the adaptive parameters converging, thereby validating the effectiveness of the online estimation in assessing the impacts of interactions. Note to Practitioners—The motivation of this article is to address the control issue of soft robots under contacts. The inherent softness and compliance of soft robots endow them with excellent safety interaction performance, making them highly applicable in fields such as medicine, education, and human-robot collaboration. However, in constrained and unknown environments, the kinematic model of soft robots can undergo unpredictable changes due to contact interferences. This poses a significant challenge to the control of soft robots, severely impacting the feasibility of their extensive application. To tackle this problem, this article proposes an adaptive visual servo controller that compensates for contact interferences in real-time, enabling accurate position control of the soft robot’s end-effector during external interactions.}, archive = {J_TASE}, author = {Xiangjun Kang and Fan Xu and Lijun Han and Hesheng Wang}, doi = {10.1109/TASE.2025.3584392}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18395--18404}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive visual servo of soft robot with interaction estimation and compensation}, volume = {22}, year = {2025}, }
• (2025). Planning and control for wheel-leg hybrid locomotion in wheeled biped robots for obstacle traversal. TASE, 22, 18382--18394. (www)

  @article{ , abstract = {This paper presents a wheel-leg hybrid gait planning and whole-body motion generation method for wheeled biped robots (WBR). This method utilizes momentum to evaluate the robot’s balance, and employs direct collocation to optimized generation the wheel-leg hybrid gait in real-time based on the environment elevation information. The trajectory optimization (TO) is decoupled into the forward and lateral motion to reduce the solution time for each subproblem, thereby enhancing the efficiency of the overall planning, enabling online trajectory generation at 5 Hz. An obstacle search algorithm is proposed to identify the obstacles along the robot’s reference path and construct terrain and safety constraints, which are incorporated into the optimization problem through exponential control barrier function. This facilitates the online generation of the robot’s contact sequence, ensuring that wheels movement remain within a safe region. Finally, a balance and trajectory tracking controller based on nonlinear model predictive control (NMPC) is proposed to generate the CoM’s spatial motion and contact trajectory at a higher update frequency (200Hz), which are then mapped into the joint space using inverse kinematic, enabling the robot to track the reference gait trajectory while maintaining balance. Experimental validation on a hydraulically driven WBR demonstrates that the method enables periodic hybrid gait generation and real-time obstacle search and traversal. Note to Practitioners—The wheeled-legged robot combines the speed and efficiency of wheeled robots with the terrain traversal capabilities of legged robots, making it an ideal solution for mobile robots operating in challenging terrains. However, the wheeled biped robots, as a subclass of wheeled-legged robot, possesses under-actuated and inherently unstable characteristics, current motion control research for such robots mainly focuses on rolling control. When obstacles are present along the robot’s motion path, it is necessary to re-plan the trajectory to avoid collision, which will reduce the robot’s operational efficiency in real-world environments. Therefore, the motivation of this paper is to develop a wheel-leg hybrid gait planning and control method for WBRs, enabling the robot to identify obstacles along the motion path based on terrain information and generate hybrid gaits on-line to traverse the obstacles. This approach holds significant practical implications for enhancing the robot’s terrain adaptability and operational efficiency.}, archive = {J_TASE}, author = {Haoyang Yu and Xu Li and Shiqi Guan and Zhenguo Tao and Haibo Feng and Songyuan Zhang and Yili Fu}, doi = {10.1109/TASE.2025.3587848}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18382--18394}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Planning and control for wheel-leg hybrid locomotion in wheeled biped robots for obstacle traversal}, volume = {22}, year = {2025}, }
• (2025). H∞ asynchronous secure control of piecewise homogeneous markov jump systems: A switching-like probabilistic event-triggered approach. TASE, 22, 18370--18381. (www)

  @article{ , abstract = {This paper deals with the event-triggered $H_{\infty }$ asynchronous secure control problem for piecewise homogeneous Markov jump systems (PHMJSs) with denial-of-service (DoS) attacks and mode-dependent deception attacks. With the aid of acknowledgement character technique and probabilistic threshold division method, a mode-dependent switching-like probabilistic event-triggered mechanism is established to save communication resources while remedying the adverse effects of both DoS attack-induced packet losses and stochastic fluctuation of the triggering threshold on PHMJSs. Sufficient criteria are developed to ensure the stochastic stability and $H_{\infty }$ performance of PHMJSs with hybrid cyber-attacks. It analytically builds the relationship between the maximum number of consecutive packet losses induced by DoS attacks and the triggering parameter. A co-design scheme is provided to calculate the event-triggering weight matrices and controller gains. Ultimately, an application-oriented example of the F-404 aircraft engine system is offered to demonstrate the validity of the proposed approach. Note to Practitioners—Given the powerful ability of Markov jump systems to characterize abrupt changes of system parameters or structures, they play a crucial role in real-world applications such as mobile robotics, power systems, and aerospace. Note that the system mode transition probability is typically time-varying owing to environmental factors like weather. Besides, in networked control systems, open network communication environments and limited network bandwidth constraints pose two unavoidable key challenges to data transmission: security and energy-saving issues. In response to the above challenges, this paper investigates the switching-like probabilistic event-triggered $H_{\infty }$ asynchronous secure control problem for piecewise homogeneous Markov jump systems with hybrid cyber-attacks. An F-404 aircraft engine system is adopted to check the proposed results. This work provides the effective method for practitioners interested in event-triggered asynchronous secure controller design under the complex realistic networked environments.}, archive = {J_TASE}, author = {Wende Luo and Haiyang Chen and Guangdeng Zong and Ying Guo}, doi = {10.1109/TASE.2025.3587441}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18370--18381}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {H∞ asynchronous secure control of piecewise homogeneous markov jump systems: A switching-like probabilistic event-triggered approach}, volume = {22}, year = {2025}, }
• (2025). Network-level control of intermittent bus lanes using multi-agent deep reinforcement learning with multi-feature state prediction. TASE, 22, 18360--18369. (www)

  @article{ , abstract = {The current method of setting up bus lanes is based on historical or current road traffic conditions for control, but these methods lack a mechanism for predicting traffic conditions. This paper constructs an LSTM multi-feature state prediction model to acquire observed states in the environment, utilizes neural networks to predict future traffic conditions, and considers variables affecting actual traffic conditions. Both are used as joint states to train each agent, helping them better understand the environment and take optimal actions. Additionally, to effectively address the differences in characteristics between social vehicles and buses, a multi-type intelligent agent for signal lights and bus lanes is constructed for training. The DPQ (Deep Prediction Q-Network) deep reinforcement learning algorithm is proposed to optimize signal lights and bus lane settings. Furthermore, using mean field theory, the MFPQ (Mean Field Prediction Q-Network) algorithm models the problem as interactions between a single intelligent agent and neighboring intelligent agents’ average effects, reducing the complexity of interaction between agents. Experimental results show that both the DPQ algorithm and the MFPQ algorithm improve evaluation metrics such as travel time. The LSTM multi-feature state prediction model promotes optimal decision-making by predicting future states for intelligent agents and significantly accelerates the convergence speed of the algorithm, enhancing learning performance. Note to Practitioners—Efficient management of intermittent bus lanes is a significant challenge in urban traffic systems, especially under dynamic and unpredictable conditions. This paper introduces a novel approach combining predictive modeling and multi-agent reinforcement learning to achieve adaptive control of bus lanes and traffic signals. By leveraging an LSTM-based multi-feature state prediction model, intelligent agents can forecast future traffic conditions and make proactive decisions, resulting in significant improvements in traffic flow efficiency. The proposed DPQ and MFPQ algorithms effectively optimize bus lane utilization and signal settings. Simulation results demonstrate notable reductions in travel times and enhanced traffic coordination, highlighting the practical value of this approach. This research provides a robust foundation for intelligent bus lane management and offers strong potential for real-world applications in intelligent transportation systems.}, archive = {J_TASE}, author = {Zundong Zhang and Xun Yang and Yuke Liu and Yutong Song}, doi = {10.1109/TASE.2025.3586674}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18360--18369}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Network-level control of intermittent bus lanes using multi-agent deep reinforcement learning with multi-feature state prediction}, volume = {22}, year = {2025}, }
• (2025). Distributed multiple shape formation in homogeneous robot swarms. TASE, 22, 18341--18359. (www)

  @article{ , abstract = {In this paper, we study the problem of simultaneous formation of multiple shapes in homogeneous swarms, which is rarely explored. Compared to single shape formation, distributed multiple shape formation faces the following challenges: how to divide the swarm into several sub-swarms for multiple shapes and how to satisfy several constraints between shapes, such as avoiding overlap, maintaining shape order, and preserving swarm connectivity after shape completion. To handle these challenges, we propose the following novel results. First, a distributed game-theoretic clustering algorithm considering the sub-swarm size is proposed to divide the robot swarm into several balanced sub-swarms. This balanced division can ensure that each shape can be achieved by several robots, rather than none. Second, each robot can solve a mixed integer convex quadratic optimization problem to determine multi-shape parameters (shape sizes and positions) to satisfy several necessary constraints. Unlike existing shape optimization methods, this optimization does not require the states of all robots, thus achieving distributed control. Third, we propose a robust strategy for robots against swarm-scale variants via fine-tuning existing clustering results and shape parameters. In particular, the previous work on single shape control can drive each sub-swarm to achieve its corresponding shape. Finally, both simulation and experiment with 11 physical robots verify the effectiveness and robustness of the proposed algorithm. Note to Practitioners—In several practical scenarios, it is necessary for a robot swarm to simultaneously form several shapes, such as multiple region exploration, and multiple object transportation. However, most of the existing methods for shape formation focus on the single connected shape for robot swarms, where there is a path between any two points in the shape. To address this limitation, this paper proposes a multiple shape formation control algorithm where the input to the algorithm is just several binary images denoting shapes and several constraints between shapes. This algorithm includes three main components: 1) a distributed game-theoretic clustering algorithm to divide the swarm into several balanced sub-swarms; 2) an optimization problem to solve multi-shape parameters while considering several necessary constraints for multiple shapes; 3) a mean-shift controller, based on the previous work, to guide the swarm in forming multiple shapes. More details of the proposed algorithms and swarm robotic systems are provided in this paper.}, archive = {J_TASE}, author = {Xing Li and Rui Zhou and Guibin Sun and Yunjie Zhang and Zhaozong Wang}, doi = {10.1109/TASE.2025.3586130}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18341--18359}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed multiple shape formation in homogeneous robot swarms}, volume = {22}, year = {2025}, }
• (2025). Optimizing wastewater treatment control using policy-constrained offline reinforcement learning and lyapunov density model. TASE, 22, 18331--18340. (www)

  @article{ , abstract = {Wastewater treatment plants (WWTPs) constitute a crucial component of contemporary societal infrastructure, playing a critical role in purifying and recycling wastewater from both municipal and industrial sources. The effectiveness and reliability of control strategies in these processes are paramount. This paper presents an offline reinforcement learning (RL) controller enhanced with the Lyapunov density model (LDM) to solve the optimal tracking control problem in wastewater treatment processes. The offline RL algorithm employs off-policy learning to derive the optimal control policy from the offline dataset without risky online trials. By incorporating the density estimation of the offline dataset, the LDM estimates the probability that the system adheres to the distribution of the dataset under the learned policy. The introduction of LDM helps constrain the learned policy, thereby enhancing its reliability when encountering unknown states. Additionally, a weighting scheme, based on the coefficient of variation, is applied to determine the relative weights of the LDM and the critic in the update of the control policy. Simulations conducted on a nonlinear system and the wastewater treatment process demonstrate that the proposed method learns a more effective policy from the offline dataset, leading to reduced implementation costs. Note to Practitioners—This research addresses the optimal tracking control of Wastewater Treatment Plants (WWTPs) using offline reinforcement learning (RL), thereby eliminating the need for risky real-world trials. This method is applicable to any unknown nonlinear system with available historical operational data. Offline RL is achieved by constraining the policy learning process to remain close to the behavior policy derived from the dataset. However, during deployment, especially when encountering out-of-distribution states, the learned agent may still produce dangerous actions. To address this, the Lyapunov Density Model (LDM) is introduced to constrain the learned policy based on the dataset’s density estimation, ensuring it remains in-distribution. The LDM acts as a guardrails, preventing the system from learning from infrequently occurring data. Although developed for WWTPs, this control approach has potential applications in other industrial processes where safety-critical control decisions rely on historical operational data, such as chemical processing plants and power generation facilities.}, archive = {J_TASE}, author = {Ruyue Yang and Ding Wang and Junfei Qiao}, doi = {10.1109/TASE.2025.3584762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18331--18340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimizing wastewater treatment control using policy-constrained offline reinforcement learning and lyapunov density model}, volume = {22}, year = {2025}, }
• (2025). Hierarchical safety-critical control method for DDEV during handling limit: A strategy via safety region reconstruction and extension. TASE, 22, 18314--18330. (www)

  @article{ , abstract = {Benefited from the advantage of four-wheel independent driving, Distributed Driving Electric Vehicle (DDEV) can precisely generate direct yaw moment (DYC) to influence lateral motion. In coordination with Active Front Steering (AFS), this allows for decoupled control under normal driving condition. However, DYC is fundamentally generated through the longitudinal forces of the tires, so its impact on dynamic stability under handling limit cannot be ignored. Particularly, the intervention of DYC changes the vehicle’s safety boundary, a factor seldom addressed in existing research. Extension literatures use the safety boundaries of traditional Centralized Driving Vehicle (CDV) to design stability controller for DDEV, resulting in conservative or aggressive performance. To this end, we propose a hierarchical safety-critical controller for DDEV during handling limit. A feedback control law based on high-order polynomials is constructed, and the closed-loop stability boundary is expanded using the Sum of Square Programming (SOSP) algorithm. Additionally, DYC is included in the stability margin assessment, and a safety envelope boundary suitable for DDEV is reconstructed. This safety boundary is used as a state constraint to design a safety-critical controller based on ODCBF. Simulation and experiments show that the proposed algorithm can widen the stable yaw rate boundary by 15% under high-speed continuous steering condition, enhancing the vehicle’s maneuverability while ensuring safety. Note to Practitioners—DDEV use in-wheel motors as actuators, providing more control freedom and superior handling performance, making it a core research area for many automotive manufacturers. Stability is a crucial factor affecting vehicle safety. Almost all vehicles should be equipped with active safety systems, such as Bosch’s Electronic Stability Program (ESP) and GM’s Electronic Stability Controller (ESC). However, most of these systems are developed for CDV, while DDEV have an additional control input DYC. This means that traditional ESP control strategies may be too conservative for DDEV, preventing the full utilization of the capabilities. Therefore, this paper reconstructs the stability region for DDEV and expands the stability boundary through closed-loop DYC control. This technology fills the gap in the stability margin determination of DDEV and has promising applications in the active safety systems.}, archive = {J_TASE}, author = {Tong Shen and Fanxun Wang and Mingzhuo Zhao and Yanjun Ren and Jinhao Liang and Shuo Bai and Guodong Yin}, doi = {10.1109/TASE.2025.3581675}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18314--18330}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical safety-critical control method for DDEV during handling limit: A strategy via safety region reconstruction and extension}, volume = {22}, year = {2025}, }
• (2025). Tight fusion of odometry, kinematic constraints, and UWB ranging systems for state estimation of integrated aerial platforms. TASE, 22, 18297--18313. (www)

  @article{ , abstract = {Integrated Aerial Platforms (IAPs), consisting of multiple interconnected aircraft, offer a promising framework for aerial manipulation tasks by enhancing localization accuracy and reliability. Unlike aerial swarms, the interconnected nature of IAPs allows for exploiting physical constraints among aircraft to improve positioning and navigation systems. In this paper, we present an advanced decentralized multi-aircraft visual-inertial-range-physical odometry system that considers the position, velocity, and attitude constraints inherent to IAPs. By tightly fusing visual-inertial-range odometry and Ultra-Wideband (UWB) with kinematic constraints, we optimize odometry accuracy through the use of novel constraint-based methods. Our algorithm’s performance is validated on simulated datasets and self-collected datasets, with flight experiments conducted on the IAP, demonstrating a significant improvement with a 28.7% reduction in drift over the baseline on real-world datasets and a 24.5% reduction on simulation datasets. Note to Practitioners—Motivated by the need for accurate localization in Integrated Aerial Platforms (IAPs), this paper introduces a novel approach to multi-aircraft odometry technology through the application of physical constraints. We propose a method that combines the physical constraints of IAPs with ranging data to achieve rapid and efficient unification of coordinate systems and precise estimation of anchor positions. We then develop a theoretical framework to analyze the position, velocity, and attitude constraints induced by the multiple aircraft within the IAP and introduce a decentralized optimization method to enhance odometry accuracy. This framework is referred to as Visual-Inertial-Range-Physical Odometry (VIRPO). Furthermore, using multiple self-collected datasets for IAPs, we conduct a thorough evaluation of the proposed VIRPO algorithm, comparing its performance with baseline methods. Test results show that VIRPO significantly improves localization accuracy. Finally, we implement the IAP platform and integrate the VIRPO algorithm to conduct flight experiments, verifying its effectiveness in real-world flight conditions. This marks the first time the VIRPO algorithm has been run onboard an actual IAP platform. In future research, we plan to explore more complex IAP configurations and additional types of constraints to further enhance the localization performance of IAPs.}, archive = {J_TASE}, author = {Yushu Yu and Yingjun Fan and Ganghua Lai and Chuanbeibei Shi and Fuchun Sun and Xin Meng}, doi = {10.1109/TASE.2025.3586765}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18297--18313}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tight fusion of odometry, kinematic constraints, and UWB ranging systems for state estimation of integrated aerial platforms}, volume = {22}, year = {2025}, }
• (2025). Finite-time sliding mode adaptive control for unknown nonlinear beam system with neural network disturbance observer. TASE, 22, 18281--18296. (www)

  @article{ , abstract = {This paper proposes a finite-time sliding mode adaptive tracking control strategy with a neural network disturbance observer to address problems of inertia uncertainties, unknown external disturbances, singularity, and chattering. First, an accurate mathematical model of the Bernoulli-Euler beam system is constructed using an in-domain constraint strategy for surface-mounted sensors. Next, a novel finite-time disturbance observer based on the Radial Basis Function (RBF) neural network is designed to handle uncertainties, enabling fast response. In addition, a novel integral sliding surface is designed to resolve the singularity issue in terminal sliding mode. Finally, an adaptive terminal sliding mode controller is developed based on the finite-time disturbance observer and integral sliding surface to ensure finite-time convergence of tracking errors; its stability is rigorously analyzed. Simulation and experimental results demonstrate that the proposed sliding mode control strategy achieves superior tracking performance and finite-time stability compared to traditional methods in industrial beam systems. Note to Practitioners—The motivation of this paper is the vibration control problem of beam systems. Existing methods have limitations in improving system accuracy and addressing convergence time issues, primarily due to the difficulty in accurately estimating disturbances and system uncertainties. To address this, a finite-time sliding mode adaptive tracking control strategy based on neural network approximation is proposed. By applying an in-domain control strategy with sensors mounted on the Bernoulli-Euler beam equation, an accurate mathematical model of the beam system is constructed to achieve finite-time adaptive sliding mode control. The system uncertainties and unknown disturbances are treated as a lumped disturbance. Utilizing the universal approximation capability of the RBF neural network, a finite-time neural network disturbance observer is designed for compensation. Additionally, a norm-based adaptive law is designed to avoid redundant weight calculations, thereby improving computational efficiency. To mitigate the chattering effect of sliding mode control, a nonlinear gain function is applied, avoiding the singularity issue of terminal sliding mode control and enhancing the system’s tracking performance. The feasibility of the proposed method is verified through both simulation and physical experiments; however, the system has not yet been tested in actual production environments. In future research, the focus will be on improving the accuracy of the finite-time strategy and integrating data-driven approaches to build models for specific engineering applications.}, archive = {J_TASE}, author = {Tongming Huo and Xiaoli Li and Lianqing Zhu and Kang Wang}, doi = {10.1109/TASE.2025.3586229}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18281--18296}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time sliding mode adaptive control for unknown nonlinear beam system with neural network disturbance observer}, volume = {22}, year = {2025}, }
• (2025). Automated live cell evaluation via a CNN-transformer combined microscopy image enhancement network. TASE, 22, 18269--18280. (www)

  @article{ , abstract = {Automated morphological measurement of cellular and subcellular structures in live cells is important for evaluating cell functions. Due to their small size and transparent appearance, visualizing cellular and subcellular structures often requires high magnification microscopy and fluorescent staining. However, high magnification microscopy gives a limited field of view, and fluorescent staining alters cell viability and/or activity. Therefore, microscopy image enhancement methods have been developed to predict detailed intracellular structures in live cells. Existing image enhancement networks are mostly CNN-based models lacking global information or Transformer-based models lacking local information. For these purposes, a novel CNN-Transformer combined bilateral U-Net (CTBUnet) is proposed to effectively aggregate both local and global information. Experiments on the collected sperm cell enhancement dataset demonstrate the effectiveness of proposed network for both super-resolution and virtual staining prediction. Note to Practitioners—Automated and accurate intracellular morphology measurement is crucial for cell quality analysis. Microscopy image enhancement methods including super-resolution and virtual staining prediction were proposed to enhance or highlight details of subcellular structures without high magnification microscopy or invasive staining. To effectively combine local and global information, a novel CNN-Transformer combined image enhancement network is proposed. Different from traditional CNN-Transformer combined structures that only directionally fuse outputs from CNN and Transformer, the proposed bilateral fusion module bidirectionally fuses and exchanges features from CNN and Transformer. The proposed bilateral fusion module incorporates not only channel-wise fusion but also spatial-wise fusion to effectively aggregate local and global features. Additionally, a region-aware attention gate is proposed to urge the network to only focus on reconstructing cell structures regardless of background. The proposed method outperformed existing networks with a better enhancement effect for subcellular details.}, archive = {J_TASE}, author = {Wenyuan Chen and Haocong Song and Zhuoran Zhang and Changsheng Dai and Guanqiao Shan and Hang Liu and Aojun Jiang and Chen Sun and Wenkun Dou and Changhai Ru and Clifford Librach and Yu Sun}, doi = {10.1109/TASE.2025.3585728}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18269--18280}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated live cell evaluation via a CNN-transformer combined microscopy image enhancement network}, volume = {22}, year = {2025}, }
• (2025). Improving floating offshore wind farm flow control with scalable model-based deep reinforcement learning. TASE, 22, 18255--18268. (www)

  @article{ , abstract = {This paper proposes a model-based deep reinforcement learning (DRL) framework to maximize the total power output and minimize the fatigue load of a floating offshore wind farm subject to wake effect. Recognizing the extensive interactions required for the DRL training, we first develop an open-source physics-based model that describes the time-averaged dynamics of the floating wind farm. This model is designed with sufficient fidelity to support the wind farm control and high computational efficiency to facilitate DRL training. Subsequently, a model-based DRL approach is proposed, featuring simultaneous learning of system dynamics and optimal control policies. This dual learning process enhances the scalability of the DRL agent, making the framework suitable for large-scale floating wind farms. Finally, the effectiveness of the proposed scheme is validated by case studies with a dynamic floating wind farm simulator FAST.Farm. Note to Practitioners—This paper was motivated by the problem of improving the energy production and reducing fatigue load for floating offshore wind farms affected by wake effects, but it also applies to other types of wind farms. Traditional approaches to wind farm optimization often struggle with the computational complexity of modeling turbine interactions and wake dynamics. To address this, this paper proposes a new approach for optimal wind farm flow control using model-based DRL, which simultaneously learns system dynamics and optimizes control strategies. This integrated approach enhances the training stability and scalability of the DRL agent, making it suitable for large-scale wind farms. Simulation results on a dynamic wind farm simulator FAST.Farm suggest that this approach is feasible but it has not yet been incorporated into a real wind farm energy management system. Future work will explore its integration with real-time monitoring systems.}, archive = {J_TASE}, author = {Mingyang Mei and Peng Kou and Yilin Xu and Zhihao Zhang and Runze Tian and Deliang Liang}, doi = {10.1109/TASE.2025.3585016}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18255--18268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Improving floating offshore wind farm flow control with scalable model-based deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). High-definition light and shadow reconstruction of blast furnace burden surface images based on a spatiotemporal multi-domain guided generative model. TASE, 22, 18243--18254. (www)

  @article{ , abstract = {The blast furnace (BF) is the most energy-consuming and emission-intensive equipment in the steel industry. Burden surface images provide reliable and concrete information that is crucial for real-time monitoring, understanding burden distribution, and ensuring efficient and green operations. However, the high-temperature, enclosed, dusty, and low-light environment within the BF results in images captured by high-temperature industrial endoscopes that suffer from unevenly illumination, abnormal light and shadow, and blurriness, limiting their application. This paper proposes a method for high-definition reconstruction of BF surface images based on a spatiotemporal multi-domain guided generative model (STMGG). This method constructs a deep image prior (DIP) network with an encoder-decoder mechanism to decompose the original burden surface image into illumination and reflectance maps across multiple domains. For the illumination maps, a multi-frequency-attention fusion block (MFAFB) extracts extreme bright and dark features from the original image, calculates the appropriate fusion weights for the illumination maps of each domain, and synthesizes an optimal distributed total illumination map to accurately restore light and shadow. For the reflectance maps, global temporal and multi-spatial domain local structural texture information guidance based on wavelet transform is introduced, along with a structural texture entropy loss, to ensure the generated reflectance maps focus on both global contours and local details, thus reducing blur. The final high-brightness, high-definition reconstruction of the BF burden surface is achieved by merging the total illumination and reflectance maps. Experimental results demonstrate that STMGG significantly enhances the quality of BF burden surface images and exhibits strong adaptability in general scene images, highlighting its broad applicability and significant practical value. Note to Practitioners—The motivation for this paper arises from the challenges of acquiring clear BF burden surface images using high-temperature endoscopes due to uneven lighting, abnormal light and shadow, and blurriness. These unclear images hinder accurate extraction of burden surface morphology, challenging precise BF control. A STMGG method leverages depth image priors and an encoder-decoder mechanism to reconstruct burden surface images with global temporal information and multi-frequency domain local detail information. BF operators can convert original images into high-definition, high-brightness images using STMGG, aiding in monitoring burden distribution and BF operations, thereby improving smelting process control. STMGG method has been tested in industrial settings, with future work focusing on optimizing performance for reliable and efficient use in harsh environments.}, archive = {J_TASE}, author = {Jianjun He and Xiao Ji and Zhipeng Chen and Ling Shen and Weihua Gui}, doi = {10.1109/TASE.2025.3584202}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18243--18254}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-definition light and shadow reconstruction of blast furnace burden surface images based on a spatiotemporal multi-domain guided generative model}, volume = {22}, year = {2025}, }
• (2025). Synth4Seg—Learning defect data synthesis for defect segmentation using bi-level optimization. TASE, 22, 18231--18242. (www)

  @article{ , abstract = {Defect segmentation is crucial for quality control in advanced manufacturing, yet data scarcity poses challenges for state-of-the-art supervised deep learning. Synthetic defect data generation is a popular approach for mitigating data challenges. However, many current methods simply generate defects following a fixed set of rules, which may not directly relate to downstream task performance. This can lead to suboptimal performance and may even hinder the downstream task. To solve this problem, we leverage a novel bi-level optimization-based synthetic defect data generation framework. We use an online synthetic defect generation module grounded in the commonly-used Cut&Paste framework, and adopt an efficient gradient-based optimization algorithm to solve the bi-level optimization problem. We achieve simultaneous training of the defect segmentation network, and learn various parameters of the data synthesis module by maximizing the validation performance of the trained defect segmentation network. Our experimental results on benchmark datasets under limited data settings show that the proposed bi-level optimization method can be used for learning the most effective locations for pasting synthetic defects thereby improving the segmentation performance by up to 21.3% when compared to pasting defects at random locations. We also demonstrate up to 2.7% performance gain by learning the importance weights for different augmentation-specific defect data sources when compared to giving equal importance to all the data sources. Note to Practitioners—Data limitations have long been a significant challenge for defect detection in the manufacturing industry. Consequently, the generation of synthetic defect samples is a crucial task. In our research, we developed an easy-to-implement, end-to-end, online differentiable Cut&Paste module that can be integrated into any image segmentation-based defect detection framework. This module can generate synthetic defect samples in real-time during the training of deep neural networks. Additionally, to ensure the quality of the generated synthetic defect samples, we introduced a bi-level optimization method that allows for controllable defect generation. This method is guided by validation set performance to enhance task-specific outcomes. Currently, our proposed method can not only learn optimal defect generation strategies but also identify critical areas on product surfaces for generating synthetic defects.}, archive = {J_TASE}, author = {Shancong Mou and Raviteja Vemulapalli and Shiyu Li and Yuxuan Liu and C. Thomas and Meng Cao and Haoping Bai and Oncel Tuzel and Ping Huang and Jiulong Shan and Jianjun Shi}, doi = {10.1109/TASE.2025.3583713}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18231--18242}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synth4Seg—Learning defect data synthesis for defect segmentation using bi-level optimization}, volume = {22}, year = {2025}, }
• (2025). Diffusion-based surrogate modeling and multi-fidelity calibration. TASE, 22, 18217--18230. (www)

  @article{ , abstract = {Physics simulations have become fundamental tools to study myriad engineering systems. As physics simulations often involve simplifications, their outputs should be calibrated using real-world data. In this paper, we present a diffusion-based surrogate (DBS) that calibrates multi-fidelity physics simulations with diffusion generative processes. DBS categorizes multi-fidelity physics simulations into inexpensive and expensive simulations, depending on the computational costs. The inexpensive simulations, which can be obtained with low latency, directly inject contextual information into diffusion models. Furthermore, when results from expensive simulations are available, DBS refines the quality of generated samples via a guided diffusion process. This design circumvents the need for large amounts of expensive physics simulations to train denoising diffusion models, thus lending flexibility to practitioners. DBS builds on Bayesian probabilistic models and is equipped with a theoretical guarantee that provides upper bounds on the Wasserstein distance between the sample and underlying true distribution. The probabilistic nature of DBS also provides a convenient approach for uncertainty quantification in prediction. Our models excel in cases where physics simulations are imperfect and sometimes inaccessible. We use a numerical simulation in fluid dynamics and a case study in laser-based metal powder deposition additive manufacturing to demonstrate how DBS calibrates multi-fidelity physics simulations with observations to obtain surrogates with superior predictive performance. Note to Practitioners—In engineering applications, physics-based simulators are often employed to model complex systems. While these simulations encode our understanding of the underlying physics, they are frequently oversimplified or miscalibrated, leading to biased outputs. A natural approach to mitigating this bias is to calibrate simulation outputs using real-world data. Traditionally, Gaussian processes have been used for this purpose. In this paper, we introduce an alternative calibration framework called Diffusion-based Surrogates (DBS). DBS leverages the flexibility of diffusion generative models to calibrate high-dimensional physics simulations. We introduce two designs to explicitly or implicitly incorporate physics simulations into the generative process. Our approach effectively integrates information from multi-fidelity physics models and excels in large-scale, high-dimensional calibration tasks. Notably, DBS operates without requiring additional domain knowledge beyond simulation outputs. Further, DBS is shown to effectively quantify the uncertainty in the predictions.}, archive = {J_TASE}, author = {Naichen Shi and Hao Yan and Shenghan Guo and Raed Al Kontar}, doi = {10.1109/TASE.2025.3582171}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18217--18230}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Diffusion-based surrogate modeling and multi-fidelity calibration}, volume = {22}, year = {2025}, }
• (2025). Iterative durability design of products via degradation-informed bayesian optimization. TASE, 22, 18201--18216. (www)

  @article{ , abstract = {Green manufacturing has become a pressing issue in recent years, driven by the acknowledgment of its long-term social and economic benefits, increasing demand for environmentally friendly products, and expanding regulations. One of the key approaches to attaining sustainability in green manufacturing is to design long-lasting products. Nevertheless, the traditional approaches have faced several unique challenges in designing a product with an extended lifetime, such as costly and time-consuming procedures, as well as the noisy, sparse, insufficient or incomplete data. This paper proposes a novel framework to tackle these issues from a data-driven perspective. Specifically, the proposed method employs Bayesian optimization with Monte-Carlo acquisition functions to take into account both product design factors and degradation signals and incorporate the inherent uncertainty in the modeling of underlying degradation processes and prediction of product lifetimes. A series of simulation studies are presented to assess the performance of the proposed method. A case study on the Lithium-ion battery dataset is further conducted, which demonstrates the advantages of the proposed method over existing benchmark approaches. Note to Practitioners—This research paper provides a novel data-driven approach to finding an optimal product design with a prolonged lifetime. The conventional methods for the product lifetime extension are often time-consuming and expensive, while the associated data is in general noisy or incomplete. The proposed framework addresses these challenges by encoding the general degradation path model into the Bayesian optimization. In particular, the proposed method involves an iterative process that consists of the following three main steps: 1) collecting degradation signals, product design factors, and lifetime (optional) of existing products; 2) constructing a general degradation model and incorporating these results into the Bayesian optimization; and 3) identifying the next design factors to evaluate. The proposed method is especially useful in cases where the exact lifetimes of many existing products are unknown, i.e., products are still operating or have been replaced prior to their failure, and we collect signals that (indirectly) reflect the degradation status of a product.}, archive = {J_TASE}, author = {Minhee Kim}, doi = {10.1109/TASE.2024.3356950}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18201--18216}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Iterative durability design of products via degradation-informed bayesian optimization}, volume = {22}, year = {2025}, }
• (2025). An integrated decision-making method of flexible process plan and cutting parameter considering dynamic machining resources. TASE, 22, 18184--18200. (www)

  @article{ , abstract = {The integration of flexible process planning and cutting parameter optimization is of great significance to reduce energy consumption and shorten production time. Flexible manufacturing system brings great uncertainty to the flexible process planning and cutting parameter optimization. Most studies are conducted in a static manufacturing environment and lack of adaptive capacity to the uncertainty of the machining resources. To this end, an integrated decision-making method of flexible process plan and cutting parameter is proposed to improve energy efficiency. Specifically, the improved AND/OR network graph is employed to describe various types of process flexibility. Secondly, the coupling characteristics between energy consumption and machining resources, cutting parameters, and operation sequences are analyzed. Then, a Markov Decision Process is utilized to simulate the dynamic generation process of flexible process plans and cutting parameters, and the integrated decision-making method considering dynamic machining resources is designed with actor-critic framework. Finally, extensive comparative experiments are carried out to verify the validity of the proposed method. Experimental results indicate that: 1) the proposed method can determine flexible process plans and cutting parameters to adapt to the dynamics of machining resources. 2) The integrated optimization method reduces $E_{total} $ and $T_{p}$ by 3.59% and 3.45% compared to the two-stage optimization method. Note to Practitioners—Decision making of flexible process plans and cutting parameters relies on machining resources in the machining process. Dynamic changes in machining resources make the energy-aware decision of the flexible process plans and cutting parameters a challenging problem. To the best of our knowledge, this paper develops an integrated method of flexible process planning and cutting parameter optimization considering dynamic machining resources that can adapt to the change of machining resources. It may assist decision-makers to provide more practical flexible process plans and cutting parameters based on dynamic machining resources in the machining process.}, archive = {J_TASE}, author = {Xikun Zhao and Congbo Li and Ying Tang and Yan Lv}, doi = {10.1109/TASE.2023.3315546}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18184--18200}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An integrated decision-making method of flexible process plan and cutting parameter considering dynamic machining resources}, volume = {22}, year = {2025}, }
• (2025). Stress-aware optimal placement of actuators for high precision quality management in composite aircraft assembly. TASE, 22, 18174--18183. (www)

  @article{ , abstract = {Better quality management in manufacturing systems usually means preventing defects, reducing carbon emissions, and making systems greener. Modeling stress-induced processes is challenging and extremely critical in the quality management of advanced manufacturing systems. While residual stresses may be beneficial in some situations, in composite aircraft assembly, high residual stresses and extreme deformations are crucial and must be accounted for to prevent future catastrophic failures. Currently, conventional approaches to the optimal placement of actuators on composite structures are non-optimal, require ten actuators heuristically, and do not consider residual stresses. To overcome these limitations, we propose a Stress-Aware Optimal Actuator Placement framework. We provide theoretical investigations that demonstrate the convergence to global optimum, computational complexity, and mean prediction error of the proposed optimization algorithm. The stress-aware optimal actuator placement framework is able to achieve significant reductions of at least 39.3% in mean root mean squared deviations (RMSD) and 52% in maximum forces (MF), and only requires eight actuators on average while satisfying the safety threshold of residual stresses. Note to Practitioners—In aerospace manufacturing, about 80% of defects are associated with the assembly process. Defects may result in large errors and waste, high energy costs, low product quality, or even endanger human lives. The actuator placement usually significantly impacts the final quality of the assembled airplanes. Existing actuator placement strategies are not sufficient for composite aircraft assembly due to the complex nonlinear properties, ultra-high precision requirement, and residual stress requirement. The proposed method can improve the dimensional quality by optimizing actuator placement, as well as lower the residual stress and ensure the safety of products. Although our optimization framework was applied to the placement of actuators on composite fuselages, it could also be applied to the engineering design of other actuating systems in which both dimensional quality and stresses are sensitive. The proposed approach can reduce carbon emissions by preventing defects and improving quality management, ultimately aiming at zero-defect green manufacturing.}, archive = {J_TASE}, author = {Areej AlBahar and Inyoung Kim and Oliver Tim Lutz and Xing Wang and Xiaowei Yue}, doi = {10.1109/TASE.2023.3347357}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18174--18183}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stress-aware optimal placement of actuators for high precision quality management in composite aircraft assembly}, volume = {22}, year = {2025}, }
• (2025). Energy efficient multi-robot task allocation constrained by time window and precedence. TASE, 22, 18162--18173. (www)

  @article{ , abstract = {To meet the demands in terms of energy-efficient and fast production and delivery of goods, robotic fleets began to populate warehouses and industrial environments. To maximize the profitability of the operations, multi-robot systems are required to coordinate agents and avoid downtime efficiently. In this paper, agent coordination is formulated as a multi-robot task allocation (MRTA) problem with time and precedence constraints. The method capitalizes on a graph method to build a measure graph reflecting the sparsity of tasks and a precedence graph, which includes the task constraints, to group the tasks into batches. A batch solver is provided to obtain the final solutions to the MRTA. In this way, the sustainability and environmental impact of logistics operations can be improved by reducing the number of robots needed to complete tasks and also by assigning tasks closest to the robot location, reducing the amount of time and the total energy required for the robots to complete the job. Extensive experiments on both uniformly distributed and sparse data sets prove the effectiveness of the proposed algorithm compared to state-of-the-art algorithms such as MIP and TePSSI. Note to Practitioners—This paper was motivated by the problem of minimizing the energy consumption of multi-robot systems in the execution of complex tasks, which requires, in the most general case, the motion of the robot to a target location and further on-site operations. This scenario is particularly relevant in smart, automated warehouses, where mobile robots are repeatedly demanded to store or dispatch goods in a structured environment, where operation duration and future requests are known a priori. The paper formulates this problem by means of a batched multi-robot task allocation (BMRTA) optimization, which can include time windows and precedence constraints jointly. First, the task constraints are encoded into two graphs and then combined to group subtasks together in batches. Then, each batch is solved separately, minimizing the overall energy required to achieve the tasks in the batch. Although the optimality of the solution is ensured only locally, i.e., within the same batch, the task clustering improves the computational efficiency with respect to global approaches, especially in large-sized problems. Experimental results demonstrated that when comparing BMRTA with literature approaches such as MIP and TePSSI, not only the energy consumption but also the total travel distance can be minimized while the total duration of the tasks remains comparable.}, archive = {J_TASE}, author = {Lixuan Zhang and Jianzhuang Zhao and Edoardo Lamon and Yabin Wang and Xiaopeng Hong}, doi = {10.1109/TASE.2023.3312214}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18162--18173}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy efficient multi-robot task allocation constrained by time window and precedence}, volume = {22}, year = {2025}, }
• (2025). Layout optimization for photovoltaic panels in solar power plants via a MINLP approach. TASE, 22, 18148--18161. (www)

  @article{ , abstract = {Photovoltaic (PV) technology is one of the most popular means of renewable generation, whose applications range from commercial and residential buildings to industrial facilities and grid infrastructures. The problem of determining a suitable layout for the PV arrays, on a given deployment region, is generally non-trivial and has a crucial importance in the planning phase of solar plants design and development. In this paper, we provide a mixed integer non-linear programming formulation of the PV arrays’ layout problem. First, we define the astronomical and geometrical models, considering crucial factors such as self-shadowing and irradiance variability, depending on the geographical position of the solar plant and yearly time window. Subsequently, we formalize the mathematical optimization problem, whose constraints’ set is characterized by non-convexities. In order to propose a computationally tractable approach, we provide a tight parametrized convex relaxation. The resulting optimization resolution procedure is tested numerically, using realistic data, and benchmarked against the traditional global resolution approach, showing that the proposed methodology yields near-optimal solutions in lower computational time. Note to Practitioners—The paper is motivated by the need for efficient algorithmic procedures which can yield near-optimal solutions to the PV arrays layout problem. Due to the strong non-convexity of even simple instances, the existing methods heavily rely on global or stochastic solvers, which are computationally demanding, both in terms of resources and run-time. Our approach acts as a baseline, from which practitioners can derive more elaborate instances, by suitably modifying both the objective function and/or the constraints. In fact, we focus on the minimum set of necessary geometrical (e.g., arrays position model), astronomical (e.g., irradiance variation), and operational (e.g., power requirements) constraints which make the overall problem hard. The Appendices provide a guideline for suitably choosing the optimization parameters. All data and simulation code are available on a public repository at: https://github.com/nicomignoni/pvlayout.git.}, archive = {J_TASE}, author = {Nicola Mignoni and Raffaele Carli and Mariagrazia Dotoli}, doi = {10.1109/TASE.2023.3322786}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18148--18161}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Layout optimization for photovoltaic panels in solar power plants via a MINLP approach}, volume = {22}, year = {2025}, }
• (2025). Optimal location of electrical vehicle charging stations with both self- and valet-charging service. TASE, 22, 18133--18147. (www)

  @article{ , abstract = {The inconvenience of charging is one of the major concern for potential electric vehicle (EV) users. In addition to building more charging facilities, electric vehicle charging assistance service has emerged for making EV charging more convenient to customers. In this paper, we consider an optimal EV charging station location problem with two types of customers. One is ordinary self-charging customers whereas the other is customers using a new service mode called valet-charging. We formulate the problem via bi-level location optimization model, where the lower level problem is a game model that characterizes customers’ station choice behaviors. To solve the hard nonlinear mixed-integer optimization problem, we design an adaptive large neighbourhood search (ALNS) algorithm for the upper level problem and a construct-improve heuristic for the lower level problem. We conduct numerical experiments to justify the efficiency of our solution method. We also conduct a need-inspired case study to derive practical insights which will help EV charging assistant service providers make strategic decisions. Note to Practitioners—The convenience of charging service is one major concern for EVs. In China, NIO Inc., NETA AUTO, and FAW-Volkswagen have started to provide valet-charging service. Charging station location problem becomes complicated while taking this service into account. We believe our work develops an effective tool for charging station planners to analyze station locations as well as the impact of valet charging services.}, archive = {J_TASE}, author = {Tianyi Zhao and Na Li and Nan Kong and Xiaoqing Xie}, doi = {10.1109/TASE.2023.3321412}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18133--18147}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal location of electrical vehicle charging stations with both self- and valet-charging service}, volume = {22}, year = {2025}, }
• (2025). Robust design of a green hydrogen supply chain under spatiotemporal imbalance. TASE, 22, 18118--18132. (www)

  @article{ , abstract = {To reduce substantial renewable energy spillage and satisfy the increase in green hydrogen demand in China, conversion of surplus renewable energy into hydrogen via electrolyzers has attracted increasing attention in recent years. Seasonal storages and interregional hydrogen supply chains are commonly employed to eliminate the inherent spatiotemporal imbalance between renewable energy supply and hydrogen demand. Moreover, the efficiency and reliability of the hydrogen supply chain must be ensured. This paper addresses a robust design and planning problem for an integrated hydrogen supply chain in which the uncertainty of hydrogen demand is considered. We propose a two-stage robust optimization model with a budget-based uncertainty set, and the corresponding robust approximation model is derived. We develop a Benders decomposition based algorithm to solve the resulting model. Extensive numerical experiments reveal that the proposed algorithm outperforms CPLEX. A case study using real data is presented to demonstrate the applicability of the proposed model. Finally, conclusions are drawn, and several future research directions are outlined. Note to Practitioners—This paper was motivated by the problem of renewable energy spillage in green manufacturing systems. The intermittent nature of renewable energy leads to serious spillage and mismatch of sources and demand. This paper proposes a novel method by converting surplus renewable energy into hydrogen and designing a green hydrogen supply chain to deal with the inherent spatiotemporal imbalance between renewable energy supply and hydrogen demand. In this paper, we characterize a more reliable way to handle to uncertainty in demand side. Practitioners can deploy the proposed model to the facility location and capacity problems in risk averse systems. We then design a mathematical algorithm incorporated with accelerating approaches significantly outperforms CPLEX. Practitioners can apply our algorithm to cope with facility location and capacity problems with inventory decisions in multiple levels. Real data experiments demonstrate the applicability of the proposed model and several managerial are explored. In future research, we will address the design of multi-uncertainties in green manufacturing systems.}, archive = {J_TASE}, author = {Jiaqi Xu and Qiaofeng Li and Jiarong Li and Zhi-Hai Zhang and Li Zheng}, doi = {10.1109/TASE.2024.3381449}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18118--18132}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust design of a green hydrogen supply chain under spatiotemporal imbalance}, volume = {22}, year = {2025}, }
• (2025). Modeling and analysis of productivity and energy in serial production lines with setups. TASE, 22, 18102--18117. (www)

  @article{ , abstract = {Reducing energy consumption plays a critical role to implement the carbon peak and carbon neutral strategies in manufacturing processes with energy-intensive setups and operations. In this paper, a serial production line model with setups is introduced, where both productivity and energy consumption performances are considered. First, exact solutions of both measures are derived for two-machine lines. Next, using the results in two-machine lines as a building block, an aggregation and interpolation based approximation method is presented to evaluate system production rate and energy usage for longer lines. Using this method, system properties and sensitivities are investigated. Extensions to more complex scenarios with multiple downtimes are also discussed. Note to Practitioners—Achieving the carbon peak and carbon neutral goals has been the national strategy in response to climate change. With such a strategy, reducing energy consumption in manufacturing processes is of critical importance. Many manufacturing systems include energy-intensive setups and operations, which complicates the study of such systems and proposes new challenges in modeling and analysis of system performance. To solve this problem, a system-theoretic method is presented in this paper. At first, a two-machine line model is introduced and an exact solution can be derived to calculate the line production rate and energy consumption. Then, for longer lines, the evaluation method in two-machine lines is utilized as a building block to update system performance repeatedly. Specifically, by aggregating the setup states into down states, the model is simplified. Then, using backward and forward aggregations, an iterative method is introduced to overcome the curse of dimensionality. In addition, through interpolation based on coefficient of variation (CV), the estimates of system performance can be obtained. Moreover, by examining the performances, system-theoretic properties, such as monotonicity and reversibility, as well as sensitivity are investigated. Furthermore, it is shown that the proposed model can be extended to analyze more complex cases with multiple sequential downtimes. Such a model provides an effective method to design, analyze, and improve productivity and energy efficiency in production systems.}, archive = {J_TASE}, author = {Heng Dong and Jingshan Li}, doi = {10.1109/TASE.2023.3298781}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18102--18117}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling and analysis of productivity and energy in serial production lines with setups}, volume = {22}, year = {2025}, }
• (2025). Problem formulation and solution methodology of energy consumption optimization for two-machine geometric serial lines. TASE, 22, 18088--18101. (www)

  @article{ , abstract = {Manufacturing systems consume a tremendous amount of energy and contribute about a quarter of greenhouse gas emissions. To achieve the sustainable production, it is vital to reduce the total energy consumption and improve the energy efficiency of manufacturing systems, especially energy-intensive manufacturing systems. In this paper, the energy consumption optimization problem for a two-machine geometric line is investigated. Specifically, it is formulated as a nonlinear programming which minimizes the energy consumption of the system while maintaining a required production rate. For this nonlinear programming with complex constraints, two optimality equations are explored and their mathematical properties are analyzed. Based on these properties, an effective and computationally efficient algorithm is developed to solve the optimal solution of the energy consumption optimization problem. In addition, the sensitivity of the optimal solution with respect to system parameters is analyzed. Finally, several extensions of the problem with an alternative objective and more practical considerations, are addressed as well. Note to Practitioners—For energy-intensive manufacturing systems, reducing the energy consumption and improving the energy efficiency are of both economic and ecological significance. In the literature, almost all existing related researches assume that machines obey the Bernoulli reliability model, which is not applicable for modeling some production lines. In this paper, we extend the problem formulation and solution methodology for the two-machine line with Bernoulli reliability model to that with geometric model, which has wider applications in practical systems. However, the extension is non-trivial since it is much more complicated to solve the energy consumption optimization problem for geometric lines and more insights on the results are gained. Although the line studied is short, as a step stone, this research will be extended to long geometric lines and production lines with more practical, e.g., exponential and non-exponential, reliability models.}, archive = {J_TASE}, author = {Chao-Bo Yan and Lingchen Liu}, doi = {10.1109/TASE.2023.3313565}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18088--18101}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Problem formulation and solution methodology of energy consumption optimization for two-machine geometric serial lines}, volume = {22}, year = {2025}, }
• (2025). Fast schedule recovery method for semiconductor packaging lines with machine failures and constrained time windows. TASE, 22, 18076--18087. (www)

  @article{ , abstract = {In semiconductor packaging line, a master schedule is normally created every shift to optimize production for a three-week horizon. However, semiconductor packaging lines are susceptible to machine failures, causing the master schedule to be sub-optimal or even infeasible. Specifically, machine failures can result in due dates not to be met and time window constraints to be violated. In this study, we classify machine failures in semiconductor packaging lines into two categories: short and long machine failures, which can be identified when the failure happens. To handle short machine failures, extra time is added into the processing time of each lot to make the master schedule robust. When a long machine failure occurs, a mixed integer programming model is formulated to adjust the master schedule. The master schedule is taken as a warm start, and a short period schedule is obtained using CPLEX for the semiconductor packaging line to follow immediately. In this way, the semiconductor packaging line can quickly respond to long machine failure without replacing the whole master schedule or giving the master scheduler enough time to remake a new master schedule. Thus, the negative impact of machine failure is mitigated. Using the data from shop floor, a simulation model is developed with SimPy to simulate a real-world semiconductor packaging line and evaluate the proposed method. The experiment results show that the proposed method can achieve fast response to machine failures in semiconductor packaging lines. Note to Practitioners—The semiconductor packaging line operates as a highly intricate production system. Its inherent flexibility allows for the simultaneous processing of diverse products from various orders. However, this flexibility necessitates adherence to a master schedule, often created without accounting for potential disruptions, such as machine failures. When not promptly addressed with efficient production control measures, machine failures can lead to production losses and product quality concerns. Moisture absorption and die surface oxidation are common quality issues that arise in the semiconductor packaging line, primarily caused by extended wait times in buffers. To equip production engineers with an effective solution for real production settings, this study introduces a practical tool to swiftly respond to machine failures. By implementing this tool, production teams can mitigate the impact of disruptions and ensure smoother operations in the semiconductor packaging process.}, archive = {J_TASE}, author = {Feifan Wang and Yutong Su and Feng Ju and Balakrishnan Ananthanarayanan and Husam Dauod and Nital S. Patel}, doi = {10.1109/TASE.2024.3357610}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18076--18087}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast schedule recovery method for semiconductor packaging lines with machine failures and constrained time windows}, volume = {22}, year = {2025}, }
• (2025). Parallel machine scheduling with peak energy consumption limits. TASE, 22, 18063--18075. (www)

  @article{ , abstract = {In recent years, the industrial sectors, particularly in the manufacturing of steel, chemicals, automotive parts, and semiconductors, have been focusing on reducing energy usage in order to improve energy efficiency and maintain competitiveness in the global market. This study addresses an energy-conscious scheduling problem of identical parallel machines by considering the peak energy consumption limits. We provide an optimal schedule that minimizes the total energy consumption by developing novel integer programming (IP) models by reducing the number of decision variables and constraints from the previous model and a branch and bound (B&B) algorithm with three dominance properties and three lower bounds. The proposed B&B algorithm performs better than the existing IP and improved IP models. We then introduce a modified simulated annealing (MSA) algorithm, which accounts for the energy consumption rates for generating neighboring solutions, to solve large-sized instances. Note to Practitioners—In this work, we address parallel machine scheduling by taking into account the peak energy consumption limits imposed by various manufacturing companies. We present new mathematical programming models for the problem, demonstrating that these models provide optimal solutions faster than the previous one. Additionally, we enhance these models by developing a branch and bound (B&B) algorithm that yields optimal solutions even faster. Finally, we present a heuristic algorithm to handle large-sized instances.}, archive = {J_TASE}, author = {Sung-Ho Min and Sang-Wook Lee and Hyun-Jung Kim}, doi = {10.1109/TASE.2024.3366949}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18063--18075}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Parallel machine scheduling with peak energy consumption limits}, volume = {22}, year = {2025}, }
• (2025). Design of green power clouds for intelligent virtual power plants. TASE, 22, 18051--18062. (www)

  @article{ , abstract = {Traditional virtual power plants (VPPs) combine power from distributed energy resources (DER) to supply energy to users. However, they fall short of net-zero goals because of neglecting carbon footprints during power aggregation. This paper proposes a novel intelligent virtual power plant framework (iVPPF) to address this gap. iVPPF comprises a central iVPP (iVPP $_{\mathrm {C}}$ ) and several regional iVPPs (iVPP $_{\mathrm {n}}$ ), n = E, S, M, and N. These iVPPn are geographically distributed systems for intelligently managing iVPPs in four regions: east, south, middle, and north, respectively, while the iVPPC is responsible for dispatching power across iVPPn. We built iVPPC and iVPPn on individual green power clouds, which can provide abundant computing resources and realize intelligence through AI technologies for iVPPF. We also design universal computing devices called cyber-physical agents (CPAs) to collect essential data on manufacturing, carbon footprint, and energy usage for iVPPn. iVPPn can intelligently control DERs based on the collected data. Also, iVPPF can empower enterprises to participate in power balancing services offered by Taipower, thereby enhancing the flexibility of the overall power grid. Furthermore, we integrate iVPPF with the I4.2-GiM framework, offering intelligent carbon and energy management capabilities to achieve the net-zero goal. The testing results show that iVPPF can significantly reduce energy usage (up to 25.6%) and carbon emissions (up to 509 kg) through power dispatch. Thus, the proposed iVPPF promises to contribute economic benefits for businesses and the pursuit of net-zero emissions. Note to Practitioners—This paper proposes an intelligent virtual power plant framework $({i} \text { VPPF})$ consisting of a central coordinator $({i}\text {VPP}_{\text {C}})$ and distributed regional managers ( ${i}\text {VPP}_{\text {n}}$ for East, South, Middle, and North). Leveraging green power clouds, both ${i} \text { VPP}_{\text {C}}$ and ${i}\text {VPP}_{\text {n}}$ harness AI for intelligent management and power dispatch across regions. We detail the system architecture and showcase practical applications, including scenarios like dispatching and aggregating for demand response, using the IEEE 13-node test feeder. Additionally, we explore the design of green power clouds and cyber-physical agents (CPAs).}, archive = {J_TASE}, author = {Ting-Chia Ou and Hao Tieng and Tsung-Han Tsai and Yu-Yong Li and Min-Hsiung Hung and Fan-Tien Cheng}, doi = {10.1109/TASE.2024.3406412}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18051--18062}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design of green power clouds for intelligent virtual power plants}, volume = {22}, year = {2025}, }
• (2025). I4.2-GiM: A novel green intelligent manufacturing framework for net zero. TASE, 22, 18030--18050. (www)

  @article{ , abstract = {Industry 4.0 is accelerating, with manufacturing enterprises embracing digital transformation and intelligent manufacturing (iM) to enhance competitiveness. Manufacturing enterprises must also improve their environmental performance to meet the goal of net zero by 2050 and to avoid border carbon tax imposed by large economies. Thus, green intelligent manufacturing (GiM), i.e., conducting iM while pursuing to maximize energy conservation and carbon reduction, has become the key development trend and a fundamental challenge for the modern manufacturing industry. To address the challenges of green intelligent manufacturing (GiM), this paper proposes a novel framework called Industry 4.2 for GiM (I4.2-GiM). This framework builds on the Intelligent Factory Automation ( $i$ FA) platform, which the authors developed to achieve zero-defect manufacturing (i.e., Industry 4.1). I4.2-GiM uses various IoT devices, called Cyber-Physical Agents (CPAs), to collect and integrate large amounts of data. It also includes two interrelated systems, an intelligent carbon emission management system ( $i$ CMS) and an intelligent energy management system ( $i$ EMS), simultaneously tackling carbon reduction and energy saving. Existing factory EMSs typically save less than 10% of energy, but I4.2-GiM has been shown to conserve 10.9% of energy and reduce carbon emissions by 12.55% while conducting iM in daily production. I4.2-GiM is a promising new framework that can help manufacturing enterprises approach net zero intelligently. Note to Practitioners—This paper proposes a new green intelligent manufacturing (GiM) framework called I4.2-GiM (Industry 4.2 for GiM). I4.2-GiM builds on the Intelligent Factory Automation ( $i$ FA) platform, which the authors developed to achieve zero-defect manufacturing (i.e., Industry 4.1). I4.2-GiM also uses various IoT devices called Cyber-Physical Agents (CPAs) to collect data from multiple sources. It also includes two interrelated systems, an intelligent carbon emission management system (iCMS) and an intelligent energy management system (iEMS), to address carbon reduction and energy saving simultaneously. Moreover, this paper provides a systematic implementation procedure and several practical examples to help practitioners adopt the designs and niches of I4.2-GiM and build their desired GiM systems for reaching the goal of net zero.}, archive = {J_TASE}, author = {Hao Tieng and Ting-Chia Ou and Tsung-Han Tsai and Yu-Yong Li and Min-Hsiung Hung and Fan-Tien Cheng}, doi = {10.1109/TASE.2023.3340149}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18030--18050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {I4.2-GiM: A novel green intelligent manufacturing framework for net zero}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Automation science and engineering for green manufacturing and logistics systems. TASE, 22, 18027--18029. (www)

  @article{ , archive = {J_TASE}, author = {Feng Ju and Lei Zhao and Lars Moench and Zhihai Zhang and Cindy Chang}, doi = {10.1109/TASE.2025.3581088}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18027--18029}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Automation science and engineering for green manufacturing and logistics systems}, volume = {22}, year = {2025}, }
• (2025). Global universal finite-time stabilizing control for feedforward nonlinear systems with non-parametric uncertainty: A non-nested strategy. TASE, 22, 18017--18026. (www)

  @article{ , abstract = {The global stabilization control issue concerning a class of general upper-triangular nonlinear systems subject to non-parametric uncertainties is dealt with in this paper by constructing a nonsmooth dynamic nonrecursive state-feedback controller. The major challenge in handling such systems is that they cannot be stabilized through the conventional feedback linearization approaches. The investigated strategy does not involve a nested process, commonly seen in existing methods, but rather constructs a homogeneous stabilizer directly via a low-gain design, which greatly facilitates the design of the controller and the analysis of global stability. Meanwhile, by developing a dual-layer adaptive low-gain online updating law, the system with unknown homogeneous nonlinearity growth conditions under consideration can be flexibly stabilized. A numerical simulation and an application to the nonlinear liquid level control resonant circuit system are implemented in this paper to demonstrate the efficacy of the built framework. Note to Practitioners—Upper-triangular model construction is commonly found in practical systems, such as the nonlinear liquid level control resonant circuit system and the planar vertical take-off and landing (PVTOL) system. In these applications, there are inevitably uncertainties within model construction, e.g., internal parameter perturbations and external disturbances. They pose challenges to achieving the control goals of rapid and precise stabilization or regulation for the system. The common strategies for handling upper-triangular systems are recursive nested control and integral feedforward techniques, etc. However, for practical engineers, these strategies are hard to operate due to the complex design process and numerous design parameters to be tuned. Fortunately, nonrecursive strategies have been frequently mentioned in recent years for more convenient engineering implementation. These strategies typically feature simple controller forms, fewer control parameters and are easy to tune, which is of great significance for rapidly and precisely controlling practical systems with various uncertainties. Based on the dynamic low-gain domination strategy, a nonsmooth nonrecursive finite-time controller has been proposed in this paper under the condition that the nonlinearities growth rate is unknown. Finally, through a complex numerical simulation and an application to the nonlinear liquid level control resonant circuit system with parametric uncertainties, the effectiveness of the proposed nonrecursive algorithm has been verified.}, archive = {J_TASE}, author = {Xin Dong and Xiang Zhang and Hongtian Chen and Zehua Jia and Chuanlin Zhang and Weidong Zhang}, doi = {10.1109/TASE.2025.3586766}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18017--18026}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Global universal finite-time stabilizing control for feedforward nonlinear systems with non-parametric uncertainty: A non-nested strategy}, volume = {22}, year = {2025}, }
• (2025). An air-ground unmanned swarm collaborative area search strategy based on the learning wolf pack algorithm. TASE, 22, 18000--18016. (www)

  @article{ , abstract = {Collaborative search by air-ground unmanned swarm, as a pivotal and efficient approach for intelligence gathering and disaster relief, highlights the critical role of search path planning in enhancing overall performance. Addressing the inefficiency resulting from insufficient collaboration between air and ground unmanned platforms in current research, this paper delves into the fundamental characteristics and challenges of collaborative search by air-ground unmanned swarm. This paper clarifies the objectives and constraints of path planning and introduces a method for collaborative search path planning based on the Learning Wolf Pack Algorithm (LWPA). This method initially constructs an optimization model that comprehensively considers area coverage, target detection probability, and search uncertainty. It integrates Distributed Model Predictive Control (DMPC) with the Distributed Constraint Optimization Problem (DCOP) framework, forming an architecture for real-time search path planning. To overcome the limitation of existing DCOP solution methods, which tend to get stuck in local optimal solutions, the LWPA employs a Q-learning mechanism for hierarchical learning and dynamically adjusts parameters to balance local refinement and global exploration. Experimental results demonstrate that this method offers significant advantages in improving search efficiency, coverage, and target detection rates, with an average area coverage of 99.28% and uncertainty as low as 0.86%. These results fully validate its effectiveness and superiority in search tasks in complex urban environments. Furthermore, tests on dynamic adaptability and scalability further verify the potential and value of this method in practical applications. Note to Practitioners—This study is stimulated by the complexity of executing area search missions by air-ground unmanned swarms in unknown environments and the pressing need for efficient path planning algorithms. In practical applications such as disaster response, environmental monitoring, and search and rescue, air-ground unmanned swarms can significantly broaden the search scope and enhance efficiency. However, the complex and unpredictable nature of unknown environments makes it challenging for existing path planning methods to balance comprehensiveness and real-time performance, often leading to local optima and reducing search efficiency. To address this challenge, the present work integrates DMPC and DCOP to construct an overall search objective function within a finite time horizon, and designs the LWPA to solve it. Experimental validation demonstrates that this method is both effective and practical.}, archive = {J_TASE}, author = {Qiang Peng and Husheng Wu and Renjun Zhan and Yinan Guo and Jingyi Geng and Feng Wang and Wenxing Fu}, doi = {10.1109/TASE.2025.3586073}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {18000--18016}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An air-ground unmanned swarm collaborative area search strategy based on the learning wolf pack algorithm}, volume = {22}, year = {2025}, }
• (2025). Adaptive sliding mode security control for rotary inverted pendulum against randomly occurring false data injection attacks. TASE, 22, 17988--17999. (www)

  @article{ , abstract = {This paper investigates the adaptive sliding mode security control (ASMSC) problem of rotary inverted pendulum against randomly occurring false data injection attacks (ROFDIAs). To accomplish the control objectives for both swing-up and stabilization stages, this study proposes an adaptive backstepping nonlinear control method and an ASMSC method, respectively. These approaches distinctly differ from conventional strategies such as energy-based swing-up control methods, linear quadratic regulators, or adaptive sliding mode control techniques. Besides, note that the existing control methods are proposed in the normal network environment, once the network is malicious attacks, these methods will be invalid. Based on this, a novel ASMSC method is proposed of rotating inverted pendulum against ROFDIAs. Meanwhile, the validity of the proposed method is proved by two comparative experiments on the hardware-in-the-loop simulation platform. Note to Practitioners—This paper addresses the interesting problem of guaranteeing stable operation of an inverted pendulum system under malicious attacks. Through the deep revelation of the swing-up and stabilization stages, applying ASMSC to mitigate ROFDIAs, attacks occurring at any time can be detected and compensated for in real time, so as to realize the normal operation of the whole system. We believe this work can advance the development of systems in fields such as aerospace and robotics, where inverted pendulums are used to verify algorithm feasibility. Experimental results demonstrate that the proposed scheme outperforms existing methods.}, archive = {J_TASE}, author = {Qiang Zhang and Dakuo He and Hailong Liu and Xingling Shao}, doi = {10.1109/TASE.2025.3583888}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17988--17999}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive sliding mode security control for rotary inverted pendulum against randomly occurring false data injection attacks}, volume = {22}, year = {2025}, }
• (2025). Model and collision avoidance for motion planning of rigid-soft robot with continuous expression and input mapping. TASE, 22, 17978--17987. (www)

  @article{ , abstract = {This paper presents a motion planning approach for rigid-soft hybrid robots that incorporates continuous expression and input mapping, aiming to achieve both tracking and collision avoidance. The conventional piecewise constant curvature model utilized in soft robots suffers from issues of discontinuity and singularity. These inherent drawbacks pose significant challenges when attempting to address the nonlinear problems associated with motion planning. To overcome these limitations, we establish a relationship between the length of drive cable and the coordination of robot terminal. Through the application of Taylor transformation, we effectively eliminate the control variation present in the denominator of this expression. Then, the nonlinear model predictive control (NMPC) with this model expression and input mapping is proposed for motion planning. By directly constructing the current input and output of the motion planning based on the combination of previous data and substituting the linearized part of the model with this data, we mitigate the impact of model inaccuracies on the solution process. Both the obstacles and robot are modeled as polyhedra, and the collision problem is reformulated in the form of continuous nonlinear inequalities. These expressions are then incorporated into the NMPC optimization function as collision avoidance constraints. This approach enables the attainment of a larger feasible movement area and more precise tracking results during the motion planning process. Finally, experimental and simulation results conducted on a rigid-soft hybrid robot demonstrate the feasibility and superiority of the proposed method, validating its effectiveness in practical applications. Note to Practitioners—This paper was developed for the motion planning of rigid-soft robots in tracking and collision-avoidance tasks. In industrial robotic arms used in manufacturing processes, it can enhance motion planning to ensure seamless part handling while avoiding collisions with surrounding equipment. For example, in automotive assembly lines, these robots need to precisely pick and place components without hitting fixtures or other robots. In service robots for indoor environments, like those used in hospitals for patient care or in warehouses for inventory management, the proposed method can be used to plan paths that avoid obstacles such as furniture, people, or stacked goods. The ability to convert collision-avoidance constraints into nonlinear continuous inequalities, treating the robot and obstacles as polyhedra, provides a larger motion range, which is crucial in these complex indoor settings.In future work, we aim to improve the algorithm efficiency and real-time performance, which will further expand the applicability of this approach in time-critical robotic applications.}, archive = {J_TASE}, author = {Shaoying He and Bihui Jin and Tong Chen and Yunwen Xu and Dewei Li and Xu Zhang and Tao Zou}, doi = {10.1109/TASE.2025.3582945}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17978--17987}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model and collision avoidance for motion planning of rigid-soft robot with continuous expression and input mapping}, volume = {22}, year = {2025}, }
• (2025). Full channel multi-chaotic-state encryption strategy and error-feedback-boundary output constraint method for multiagent systems. TASE, 22, 17967--17977. (www)

  @article{ , abstract = {This paper investigate the input constraint and signal encryption for distributed control. Firstly, for the algorithm operation session, an additional information protection mechanism is designed, and the designed encryption algorithm can ensure that both the encrypted signal and the key signal have chaotic randomness. The mask signal is composed of multiple chaotic states, which has higher randomness. The key signal is also encrypted, effectively improving the security of the encryption algorithm. Even when the channel of key is not secure, information security can still be guaranteed. Moreover, information decryption only requires an integral algorithm and does not require an embedded decryption module that satisfies chaotic synchronization conditions, effectively reducing the complexity of decryption. Secondly, the proposed output constraint function can dynamically adjust the constraint boundaries based on control performance, which avoids the problem that the boundary is too strict to affect the control performance. At the same time, the boundary initial value is related to the output initial value, which effectively improves the convenience of algorithm migration and avoids the need to consider designing different parameters in different application environments. Based on the above two main works, the designed control algorithm can meet the considered control objectives using Lyapunov theory, and the simulation results also verify the effectiveness of proposed methods. Note to Practitioners—In the actual operation of multiagent systems, the presence of sensitive information necessitates ensuring the confidentiality of data transmission. Some chaotic encryption algorithms typically need to transmit key signals, which requires that the transmission of keys be absolutely secure. However, establishing this level of security in practice is challenging. Therefore, it is important to enhance the security of the key to ensure that the designed encryption algorithm maintains sufficient confidentiality even when the key channel is insecure. Additionally, when the output of the system is constrained, it will inevitably have an impact on the control accuracy. The output constraint function designed in this paper can dynamically adjust the constraint boundary based on control performance, thereby avoiding the problem of overly strict boundaries that could impair control performance.}, archive = {J_TASE}, author = {Zeyi Liu and Huaguang Zhang and Jiayue Sun and Lei Wan}, doi = {10.1109/TASE.2025.3581426}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17967--17977}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Full channel multi-chaotic-state encryption strategy and error-feedback-boundary output constraint method for multiagent systems}, volume = {22}, year = {2025}, }
• (2025). Recursive higher order non-singular terminal sliding mode control with prescribed convergence time: Application to PMSM servo systems. TASE, 22, 17955--17966. (www)

  @article{ , abstract = {In this study, we investigate the prescribed-time non-singular terminal sliding mode control (TSMC) for high-order nonlinear systems. Initially, a novel time-varying, high-order terminal sliding mode manifold is recursively constructed. Based on this newly proposed manifold, a prescribed-time non-singular TSM controller is subsequently developed. The proposed algorithm effectively eliminates the singularity problem and offers a simpler method for specifying the convergence time, independent of initial system conditions and other design parameters. To mitigate the chattering issue of the controller, we introduce a variable-gain prescribed-time disturbance observer. Leveraging this observer, we develop a prescribed-time non-singular TSMC strategy. A rigorous Lyapunov analysis demonstrates the prescribed-time stability of both the proposed observer and controller. The uniform boundedness of the state signals and control input is also established using an inductive method. Finally, experimental validation is provided, confirming the effectiveness of the proposed control algorithm. Note to Practitioners—This work addresses practical challenges in controlling high-order nonlinear systems, such as motor servo systems, robotic manipulators, and aerospace actuators, where fast convergence and robustness to disturbances are critical. Traditional TSMC often suffers from singularity issues, chattering, and convergence times dependent on initial conditions, complicating their deployment in real-world scenarios. To overcome these problems, this paper proposes a prescribed-time higher order non-singular TSMC, which ensures user-defined convergence time independent of system initial states and parameters. Experimental verification on a PMSM servo system shows that this method is effective in practical scenarios requiring fast and precise convergence.}, archive = {J_TASE}, author = {Shang Shi and Xinjian Guo and Shengyuan Xu and Huifang Min and Liaoxuan Dai}, doi = {10.1109/TASE.2025.3586612}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17955--17966}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Recursive higher order non-singular terminal sliding mode control with prescribed convergence time: Application to PMSM servo systems}, volume = {22}, year = {2025}, }
• (2025). Enhancing negotiation policies via spatio-temporal directed graphs for autonomous interaction. TASE, 22, 17941--17954. (www)

  @article{ , abstract = {Mixed traffic environments, comprising autonomous vehicles (AVs) and human-driven vehicles (HDVs), present substantial challenges for developing negotiation policies. These policies are essential for enabling AVs to make adaptive decisions and achieve harmonious interactions with HDVs in dynamic and complex scenarios. Despite their potential in autonomous driving, reinforcement learning-based decision-making schemes remain insufficient in addressing interaction-critical situations. To overcome this limitation, this study proposes a graph-enhanced negotiation-aware policy optimization (GNPO) framework, which embeds interaction awareness into all key components, including representation, understanding, and response. To support interaction representation, the spatio-temporal directed graph (STDG) captures the dynamic and asymmetric nature of interactions. It integrates multiple directed graph topologies, each encoding distinct social priors, to collectively depict realistic interactive behaviors. Additionally, a multimodal feature extractor fuses environmental perception, interaction cues, and task-related information, enabling comprehensive state understanding. Building on this representation, GNPO incorporates an interaction-critical reward function and an entropy-regularized adaptive policy optimization scheme, both designed to promote harmonious and context-aware behaviors in dense traffic scenarios. The proposed framework is validated in both interaction-intensive simulated environments and real-data-based digital twin scenarios, with its effectiveness quantitatively demonstrated by substantial improvements in success rate, safety, and efficiency. Qualitative evaluations further show that GNPO is capable of generating human-like cooperative and competitive behaviors. Note to Practitioners—Practitioners aiming to deploy decision-making algorithms in mixed traffic should pay special attention to modeling asymmetric and dynamic interactions. It is important to ensure real-time processing capability when applying graph-based methods such as GNPO. High-quality traffic data, reliable interaction cues, and efficient graph computation frameworks are key to effective implementation. To facilitate deployment in diverse environments, robustness is improved by including rare interaction cases and adversarial behaviors in training.}, archive = {J_TASE}, author = {Fan Guo and Chuyu Ma and Dezong Zhao and Guochen Liu and Kang Song}, doi = {10.1109/TASE.2025.3586434}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17941--17954}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing negotiation policies via spatio-temporal directed graphs for autonomous interaction}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered output feedback control under round-robin protocol for networked switched systems. TASE, 22, 17931--17940. (www)

  @article{ , abstract = {This paper studies the dynamic event-triggered output feedback control problem for networked switched systems (NSSs) under communications consisting of network congestion and data collisions. A dynamic event-triggered mechanism (ETM) is constructed in the sensor-controller network channel, which reduces unnecessary communications and saves network resources. To prevent data collisions in the controller-actuator network channel, the Round-Robin protocol (RRP) is used for communication scheduling, where a periodic function is used to select nodes gaining access to the network. Under the dynamic ETM and RRP, a dynamic output feedback controller is designed, and sufficient conditions are derived to ensure the mean square exponential stability (MSES) of the underlying NSSs using the mode-dependent average dwell time (MDADT) approach and multiple Lyapunov function techniques. Switching-mode-dependent controllers are further derived to achieve the desired stability. Finally, an application oriented example of F-18 aircraft model is provided to verify the usefulness of the proposed theoretical results. Note to Practitioners—Networked switched systems (NSSs) are commonly applied in areas like industrial automation, smart grids, and transportation systems. NSSs enable interconnected subsystems to collaborate with each other and achieve performance objectives, which also face significant challenges due to communication delays, network congestion, and data collisions. This paper provides a combined framework that integrates a dynamic event-triggered mechanism (ETM) and the Round-Robin protocol (RRP) to address these challenges. A dynamic output feedback controller is developed to ensure the mean square exponential stability (MSES) of NSSs through the mode-dependent average dwell time (MDADT) approach and multiple Lyapunov function techniques. A controller design algorithm is further derived which is verified by an application-orientated example. The developed results provide a useful method for networked control under limited communication resources.}, archive = {J_TASE}, author = {Huishan Du and Haiyang Chen and Guangdeng Zong and Wencheng Wang}, doi = {10.1109/TASE.2025.3585973}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17931--17940}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered output feedback control under round-robin protocol for networked switched systems}, volume = {22}, year = {2025}, }
• (2025). Collision avoidance of moving 3-D objects in dynamic environments. TASE, 22, 17914--17930. (www)

  @article{ , abstract = {In this paper, we consider the problem of reactive collision avoidance for objects that have elongated and/or non-convex shapes. The literature largely models objects using spheres, but this can either make the avoidance maneuvers over conservative (when a single large sphere is used to approximate each object) or lead to increased computational complexity (when multiple smaller spheres are used to approximate each object). We therefore model such objects using quadric surfaces that may include suitable combinations of ellipsoids, paraboloids and one-sheeted/two-sheeted hyperboloids. We present a method to construct the 3D collision cones for such object shapes using tools from projective geometry, as well as provide an analytical characterization of the $3\sigma $ bounds of the collision cone that account for noisy measurements. These are then used to devise suitable avoidance strategies. Computational analysis and simulation results are presented to demonstrate the efficacy of the methodology. Note to Practitioners—This paper is motivated by the problem of autonomous collision avoidance with moving obstacles having elongated or irregular shapes. Common approaches typically model obstacles as spheres. Using one large sphere for each object makes the avoidance too conservative, while using many smaller spheres makes it harder to calculate the avoidance commands quickly. Instead, we represent these objects with shapes that provide a tighter approximation to their actual forms. We outline a computationally efficient method to map out potential collision zones for these objects that account for their shapes, as well as velocities, and also adjust for sensor measurement errors. This approach is used to determine avoidance strategies.}, archive = {J_TASE}, author = {Kashish Dhal and Abhishek Kashyap and Animesh Chakravarthy}, doi = {10.1109/TASE.2025.3585450}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17914--17930}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Collision avoidance of moving 3-D objects in dynamic environments}, volume = {22}, year = {2025}, }
• (2025). Observer-based fuzzy adaptive dynamic surface force control for pneumatic polishing system end-effector with uncertain contact environment model. TASE, 22, 17898--17913. (www)

  @article{ , abstract = {This study presents an observer-based adaptive fuzzy dynamic surface force control strategy for pneumatic polishing system with uncertain contact environment model and full-state constraints. First, a fuzzy logic system is employed to approximate the uncertain nonlinear dynamics of the pneumatic polishing system end-effector, and a fuzzy state observer is designed to estimate the unmeasured states. Further, to enhance the controller’s response speed, a variable separation method employing fuzzy basis functions and the dynamic surface control technique are adopted to address algebraic loop issues and prevent computational complexity explosion. A logarithmic-type barrier Lyapunov function is embedded into each step of the controller design to ensure that all system states satisfy the predefined constraints. Finally, an adaptive fuzzy dynamic surface force controller with full-state constraints is developed to ensure fast and accurate control of the polishing force in an uncertain contact environment while maintaining all states within predefined constraints. The effectiveness and applicability of the proposed control algorithm are experimentally verified using a pneumatic polishing system. Note to Practitioners—The goal of this study was to investigate an adaptive fuzzy dynamic surface force control strategy for pneumatic polishing system with uncertain contact environment model and full-state constraints. Current pneumatic polishing force control schemes generally have low accuracy owing to dynamic time-varying contact characteristics that cannot be accurately modeled; therefore, this study proposes an adaptive fuzzy dynamic surface force control method that approximates model uncertainty and nonlinear functions using a fuzzy logic system. In addition, to improve the slow response speed of the polishing force, which easily leads to over- or under-polishing, the operational complexity of the controller is reduced by introducing a variable separation method based on a fuzzy basis function and dynamic surface control. This approach avoids problems such as component damage or poor machining consistency due to untimely adjustment of the polishing force. The stability and convergence of the proposed algorithm are theoretically analyzed, and its superiority and applicability are verified in polishing force control experiments.}, archive = {J_TASE}, author = {Zhiguo Yang and Wenbo Zhao and Jiange Kou and Yushan Ma and Zhibo Sun and Yixuan Wang and Yan Shi}, doi = {10.1109/TASE.2025.3585136}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17898--17913}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based fuzzy adaptive dynamic surface force control for pneumatic polishing system end-effector with uncertain contact environment model}, volume = {22}, year = {2025}, }
• (2025). Observer-based fuzzy event-triggered LFC for wind power systems under multiple cyberattacks via an NN approximation approach. TASE, 22, 17886--17897. (www)

  @article{ , abstract = {This paper presents an observer-based fuzzy event-triggered load frequency control (LFC) strategy and a neural network (NN) approximation approach for wind power systems (WPSs) facing network resource constraints and multiple cyberattacks. First, a Takagi-Sugeno (T-S) fuzzy model is developed to capture the nonlinear dynamics of the wind turbine. A novel multiple cyberattack model is then proposed, relaxing stringent assumptions about denial-of-service (DoS) and false data injection (FDI) attacks prevalent in prior studies. An observer-based dynamic event-triggered controller is subsequently designed to reduce network burden in WPSs while mitigating DoS effects. This controller incorporates an NN-based online weight adjustment algorithm to approximate and counteract the false data injected by FDI attacks. Lyapunov analysis demonstrates that the augmented error system achieves asymptotic stability with ${\mathcal {H}}_{\infty }$ performance in the absence of DoS attacks and maintains uniformly ultimate boundedness under multiple cyberattacks. Finally, a numerical example validates the effectiveness of the proposed control strategy and NN approximation approach. Note to Practitioners—Integrating WPSs with communication networks enables remote control but exposes them to cyberattack threats and resource constraints. This paper proposes a low-bandwidth, high-reliability LFC strategy to ensure frequency stability of WPSs under these conditions. Unlike existing methods, our dynamic event-triggered LFC and NN approach not only reduces communication demands but also mitigates performance degradation from cyberattacks. Additionally, the proposed cyberattack model relaxes strict boundary assumptions, enhancing the adaptability of LFC strategy to real-world attack scenarios.}, archive = {J_TASE}, author = {Dong Xu and Yajuan Liu and Ju H. Park}, doi = {10.1109/TASE.2025.3584950}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17886--17897}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based fuzzy event-triggered LFC for wind power systems under multiple cyberattacks via an NN approximation approach}, volume = {22}, year = {2025}, }
• (2025). Adaptive finite-time trajectory tracking control for wheeled mobile robots based on event-triggered and time-varying functions. TASE, 22, 17875--17885. (www)

  @article{ , abstract = {Nonholonomic wheeled mobile robots (WMRs) are typical multiinput, multioutput, coupled, and underactuated nonlinear systems, and their motion control, especially trajectory tracking, is highly challenging. However, due to the complexity of the control, it is difficult to ensure that the system converges within a finite time, let alone quickly complete various tracking tasks without consuming excessive computational resources. Therefore, this paper proposes a novel improved finite-time control scheme. Firstly, the incorporation of a smooth time-varying function enables the system to achieve a better and smoother tracking performance. Secondly, finite-time control allows the system to converge rapidly within a finite time, and its stability is proven through the Lyapunov function. Moreover, the adaptive control method makes the controller more flexible and applicable to various situations. Additionally, due to the introduction of the time-varying function, the controller’s input coupling complicates the solution of the event-triggered mechanism (ETM). To address this issue, this paper employs an extended Barbalat’s lemma to decouple the system and derive an appropriate ETM. This mechanism significantly reduces the update frequency of control signals, thereby conserving system resources. It is also proven that the designed controller can avoid Zeno behavior. Finally, the effectiveness of the proposed control scheme is verified through simulation studies. Note to Practitioners—This work focuses on developing an efficient trajectory tracking control method for nonholonomic WMRs. The goal is to tackle key challenges, including uneven trajectory tracking, slow convergence, and the difficulty in rapidly adapting to diverse tracking tasks. To this end, we propose an event-triggered adaptive finite-time trajectory tracking controller. The introduced event-triggered mechanism can effectively address practical scenarios with limited computational and transmission resources, thereby ensuring efficient resource utilization. Furthermore, the adaptive parameter adjustment mechanism in this study automatically optimizes parameters based on specific task requirements and automatically saturates once satisfactory performance is achieved, avoiding potential system stability issues caused by excessive adjustments. This design not only enhances control flexibility and robustness, but also ensures long-term system stability. Looking ahead, the control algorithm proposed in this research has great potential for further exploration and application. Subsequent studies will further consider complex factors such as external disturbances and vehicle dynamics, with the goal of validating and optimizing the algorithm’s performance in broader and more complex scenarios. With appropriate adjustments and optimizations, this algorithm is expected to provide robust technical support for trajectory tracking tasks in the field of autonomous driving, thus promoting the practical application of autonomous driving technology.}, archive = {J_TASE}, author = {Gaohan Li and Yuchao Guo and Qiang Chi and Siqi Wang and Yuan Fan}, doi = {10.1109/TASE.2025.3584937}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17875--17885}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive finite-time trajectory tracking control for wheeled mobile robots based on event-triggered and time-varying functions}, volume = {22}, year = {2025}, }
• (2025). Enhanced control of nonlinear cyber-physical systems: A higher-order relaxation matrix-based convex optimization algorithm. TASE, 22, 17863--17874. (www)

  @article{ , abstract = {This paper investigates the static output feedback (SOF) control problem of T-S fuzzy systems based on higher-order Lyapunov functions (HO-LFs). Firstly, to alleviate the conservatism of controller design, an advanced high-order time-varying relaxation matrix (HO-TVRM) approach is proposed. The HO-TVRM approach can completely adapt to the homogeneous polynomial analysis framework and effectively utilize the potential characteristics of fuzzy control to broaden the feasible domain of the reference problem. Secondly, to deal with the non-convex constraints caused by SOF control, a more flexible switching sequence convex optimization (SSCO) algorithm is developed, which adopts a switching-type optimization variable structure and combines variable-weight-based switching mechanism to alternately linearize non-convex terms during the iteration process. As a result, a convex optimization problem that approximates the original problem is obtained. Finally, simulations are presented to testify the effectiveness of the relaxed SOF control strategy. Note to Practitioners—As is well known, SOF control is a tricky bilinear matrix inequality (BMI) problem that involves the product of relaxation variables and controller gains. Existing methods typically introduce additional equality constraints and combine matrix decomposition techniques to decouple controllers, relaxation variables, and Lyapunov variables. In the decoupling process, treating the controller as an indirect optimization parameter inevitably introduces approximation errors that degrade computational precision. Therefore, this paper proposes a new SSCO algorithm that does not rely on any auxiliary constraints and treats the controller as a direct optimization parameter, which improves computational accuracy. On the other hand, this paper focuses on the SOF control problem from the perspective of fuzzy membership functions, which provides insights into designing new switching mechanisms using the potential features of fuzzy control. At the same time, according to the homogeneous matrix polynomial technique, a class of high-order matrices that fully matching the homogeneous polynomial analysis framework is obtained by integrating membership functions into the controllers, relaxation matrices, and Lyapunov variables, which provides the possibility of incorporating the switching mechanism to establish the relaxed performance constraints.}, archive = {J_TASE}, author = {Jingjing Gao and Xiangpeng Xie and Jianwei Xia and Vladimir Stojanovic}, doi = {10.1109/TASE.2025.3584831}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17863--17874}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhanced control of nonlinear cyber-physical systems: A higher-order relaxation matrix-based convex optimization algorithm}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant optimized control of switched complex networks via an adaptive dynamic programming approach. TASE, 22, 17852--17862. (www)

  @article{ , abstract = {The paper addresses the fault-tolerant optimized control of switched complex networks with unknown state and actuator fault. A proportional-integral intermediate observer is constructed to estimate unknown elements by relaxing known boundary requirement. An optimized controller is proposed to achieve fault-tolerant synchronization via adaptive dynamic programming scheme. A critic neural network is employed to solve the value of the Hamilton–Jacobi–Bellman equation. Sufficient conditions are established to ensure the uniformly ultimately bounded synchronization performance. Finally, an example is simulated to deliver the effectiveness of the proposed approach. Note to Practitioners—In practical engineering, actuator faults may result in performance degradation and even cause huge disaster under unknown states. Therefore, it is meaningful to explore a fault-tolerant optimized synchronization of multi-agent systems with unknown state and actuator fault. To treat this issue, this paper is devoted to handling two key difficulties. Namely, one is to tackle incompatibility arising from the simultaneous switching of coupling strength, outer coupling matrix, and inner coupling matrix and the other is to implement control performance optimization against unknown system states and actuator faults. As a result, a proportional-integral intermediate observer and adaptive dynamic programming incorporating a critic neural network are explored to achieve fault-tolerant optimized synchronization. The effectiveness of the proposed control policy is validated through an example.}, archive = {J_TASE}, author = {Mouquan Shen and Chen Wang and Qing-Guo Wang and Guangdeng Zong and Tingwen Huang}, doi = {10.1109/TASE.2025.3582898}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17852--17862}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant optimized control of switched complex networks via an adaptive dynamic programming approach}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based prescribed-time optimal formation control for multiagent systems. TASE, 22, 17841--17851. (www)

  @article{ , abstract = {This paper addresses the problem of prescribed-time optimal formation control for nonlinear multi-agent systems (MAS) via reinforcement learning (RL). First, a new finite horizon performance cost function is constructed, which incorporates specified time and terminal constraints, thereby it is useful for embedding the prescribed convergence time performance into optimal control framework. Based on the cost function, RL-based optimal formation controller is subsequently devised under actor-critic structure, by incorporating terminal constraint errors into the critic update law, it ensures that the optimal value function can be approximated while satisfying terminal constraints. Then, the prescribed-time optimal formation controller is designed by integrating an adjustment function with the devised optimal formation controller, with which, the formation tracking errors are ensured to converge to the origin within specified time. The proposed approach provides some advantages such as system model information is not required, optimal performance is achieved with satisfied terminal constraints, and asymptotic rather than bounded prescribed-time formation control is achieved. Finally, the effectiveness of the proposed strategy is validated through simulation examples. Note to Practitioners—Prescribed-time formation control of multi-agent systems is the theoretical foundation of numerous industrial processes, thereby motivating the research work in this paper. The existing formation control methods predominantly focus on achieving formation objective with satisfied stead-state performance, but usually neglect the equivalent important transient performance and optimal performance, that restricts the role of formation control method in engineering applications. This paper proposes a reinforcement learning based approach that enables multiagent systems to simultaneously achieve formation within user-defined time constraints and optimal performance. Moreover, the actor-critic reinforcement learning framework eliminates system’s reliance on precise dynamical model, demonstrating enhanced adaptability for industrial applications. It should be noted that current research has not considered the effective utilization of communication resources. Future work will incorporate event-triggered control approach to improve communication efficiency while maintaining formation control objective.}, archive = {J_TASE}, author = {Guanglei Zhao and Zhizhong Li and Changchun Hua and Weili Ding}, doi = {10.1109/TASE.2025.3586022}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17841--17851}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based prescribed-time optimal formation control for multiagent systems}, volume = {22}, year = {2025}, }
• (2025). Dual memristor-coupled hopfield neural network with any multi-scroll amplitude control and its application for medical image classification. TASE, 22, 17828--17840. (www)

  @article{ , abstract = {In practical applications, effectively regulating the amplitude of chaotic signals and maintaining the chaotic nature of the system are extremely critical to ensure system stability and prevent failures. However, traditional amplitude control methods usually change the bifurcation threshold or attractor geometry, impairing the integrity of chaos and increasing the risk of system instability, thus struggling to achieve effective control over complex chaotic signals. Given the rapid advancement in brain-inspired intelligence technology, it has become imperative to investigate new control techniques based on memristors to overcome the limitations of conventional approaches. To address these challenges, in this paper, a novel dual memristor-coupled Hopfield Neural Network (DMCHNN) is established, where one memristor represents external electromagnetic radiation and the other mimics synaptic connections. Two independent amplitude controllers are devised for signal rescaling, being capable of adjusting signal amplitudes in various modes, such as single-scroll, double-scroll, multi-double-scroll and coexisting homogeneous multi-scroll attractors induced by initial offset boosting. Simulations indicate that the parameter operating range of the amplitude controllers can reach up to ${10}^{5}$ or beyond. Furthermore, the performance of the amplitude controllers is additionally verified through the implementation based on the CH32 microcontroller. Rescaled chaotic signals are evaluated to determine their robust effectiveness in the deployment of pseudo-random number generators (PRNG). Eventually, the multi-scroll chaotic data with different amplitudes generated from DMCHNN is fed into the optimization algorithms for neural network optimization, which is utilized for medical image classification. Note to Practitioners—This work is motivated by utilizing the unique advantages of memristors to achieve amplitude control of complex chaotic signals in neural networks and enhance their modulation range to meet the requirements of practical engineering applications. Nevertheless, the majority of the existing control approaches are aimed at the amplitude control of single- and double-scroll within a limited range, and have less engagement in the amplitude control of relatively complex multi-scroll and coexisting multi-scroll chaotic signals. To overcome this challenge, this paper concurrently employs the synaptic plasticity and electromagnetic radiation effects of the memristors to construct a novel dual memristor-coupled Hopfield neural network (DMCHNN). The DMCHNN can not only achieve the modulation of single- and double-scroll chaotic signals via memristors, but also regulate the amplitudes of multi-scroll as well as the initial offset enhancement-induced homogeneous multi-scroll attractors. More significantly, these amplitude controllers have parameter control ranges of ${10}^{5}$ or larger, that is, realizing ultra-large-scale amplitude control of diverse chaotic signals. This presents a new control technique for the regulation of chaotic signals. Finally, the multi-scroll chaotic data with different amplitudes generated by DMCHNN is combined with the neural network optimization algorithm, which is successfully used for medical image classification and demonstrates outstanding performance, providing novel ideas and solutions for the field of image processing.}, archive = {J_TASE}, author = {Sen Zhang and Dazhe He and Yongxin Li and Daorong Lu and Chunbiao Li}, doi = {10.1109/TASE.2025.3585935}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17828--17840}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual memristor-coupled hopfield neural network with any multi-scroll amplitude control and its application for medical image classification}, volume = {22}, year = {2025}, }
• (2025). A coarse-to-fine robotic fabric alignment system integrating visual servoing and admittance control. TASE, 22, 17814--17827. (www)

  @article{ , abstract = {Fabric alignment is essential to key production processes such as cutting, sewing, and fusing in garment manufacturing. Traditionally, this task has relied heavily on the dexterity and expertise of skilled human workers. Although automated systems have been introduced, they often lack the flexibility required for complex alignment tasks. In this paper, we present a novel robotic fabric alignment framework that fully automates the process with high precision and adaptability. First, we propose a coarse-to-fine alignment strategy, where an initial imprecise target position is roughly computed based on a basic perception module and eye-to-hand calibration. This is followed by a sliding mode control (SMC)-based visual servoing approach (in an eye-in-hand configuration) to ensure a close-up view of feedback features for the fine alignment process. Additionally, we consider system disturbances estimated by a fuzzy logic system (FLS) and combine it with the controller to further enhance the system’s robustness. Finally, we developed an advanced end-effector equipped with force/torque (F/T) sensors and air-powered needle grippers for gentle fabric manipulation using admittance control. We validate our framework through a series of experiments that demonstrate its effectiveness in fabric alignment tasks. Note to Practitioners—Fabric alignment in garment production is a labor-intensive task that heavily relies on skilled human workers. Existing automated fabric alignment machines are typically limited to specific fabric shapes, patterns, and materials. Furthermore, transitioning between different fabric types requires extensive testing and adjustments, resulting in a lack of adaptability and flexibility. In this work, we propose a robotic fabric alignment system based on a coarse-to-fine alignment strategy. The initial target pose is roughly estimated using a basic perception module and eye-to-hand calibration, followed by fine adjustments through a visual servoing controller in an eye-in-hand configuration. Additionally, to ensure gentle handling, the contact force between the end-effector and the platform surface is optimized using admittance control. The proposed system offers practitioners a comprehensive solution for automating fabric alignment with high precision, while maintaining flexibility and adaptability across various fabric types and initial environmental conditions.}, archive = {J_TASE}, author = {Jiaming Qi and Liang Lu and Lei Yang and Yan Ding and Pai Zheng and David Navarro-Alarcon and Jia Pan and Peng Zhou}, doi = {10.1109/TASE.2025.3585644}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17814--17827}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A coarse-to-fine robotic fabric alignment system integrating visual servoing and admittance control}, volume = {22}, year = {2025}, }
• (2025). User-led modular robot manipulator systems interaction tasks-oriented hierarchical approximate optimal control: A stackelberg-pareto differential game perspective. TASE, 22, 17801--17813. (www)

  @article{ , abstract = {A Stackelberg-Pareto differential game-based approximate optimal interaction control approach is proposed for user-led modular robot manipulator (MRM) systems modeled by joint torque feedback (JTF) technique. The major objective of optimal control with physical human-robot interaction (pHRI) is evolved into approximating Stackelberg-Pareto equilibrium by adopting cooperative differential game in MRM and Stackelberg differential game between the human and robot. Learning from adaptive dynamic programming (ADP), the approximate optimal interaction control strategy with pHRI is developed by critic neural network (NN) for solving the coupled Hamilton-Jacobian (HJ) and HJ-Bellman (HJB) equations. The position tracking error under pHRI task is ultimately uniformly bounded (UUB) by the concept of Lyapunov theorem. Two distinction experiments demonstrate the superiority of proposed control approach. Note to Practitioners—Major challenges of controlling user-led MRM systems include optimizing MRM system performance as well as layering interaction task between the robot and human. Traditional differential game, such as zero-sum game, nonzero-sum game and cooperative game only consider the same layer in the interaction progress that is not suitable for user-led MRM with leader and follower. Besides, the existed hierarchical game deals with one leader and one follower or followers formulated as nonzero-sum game that ignores the fully cooperative relationship among followers. Therefore, this paper proposes Stackelberg-Pareto differential game-based approximate optimal control for user-led MRM to improve system performance with theoretical analysis and experimental verification.}, archive = {J_TASE}, author = {Tianjiao An and Xiaogang Dong and Bo Dong and Ruiqi Cong and Lei Liu and Bing Ma}, doi = {10.1109/TASE.2025.3585484}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17801--17813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {User-led modular robot manipulator systems interaction tasks-oriented hierarchical approximate optimal control: A stackelberg-pareto differential game perspective}, volume = {22}, year = {2025}, }
• (2025). Outlier-removal memory-event-triggered proportional-integral state estimation for wind turbine systems under multimodal deception attacks. TASE, 22, 17790--17800. (www)

  @article{ , abstract = {This paper studies the outlier-removal memory-event-triggered proportional-integral state estimation for variable-speed wind turbine systems subject to multimodal deception attacks. An innovative outlier-removal memory-event-triggered mechanism using historic data and comparison error upper bound is proposed to enhance the utilization of the limited network resource. The first merit of this mechanism is that the average value of some historical data is employed to smooth the measured outputs with stochastic fluctuations, which is able to reduce the unnecessary triggering events caused by the fluctuations. The second one is that the triggering error term’s upper bound is introduced to avoid the false transmissions induced by measurement outliers. A multimodal deception attack model with different occurrence probabilities for different modes is considered, which is more general and practical than the existing single-modal model. In order to estimate the system state, a fuzzy proportional-integral estimator is presented. With the introduction of an extra integral term, the presented estimator has the potential to improve the estimation performance compared to the traditional proportional estimator without the integral information. Then, the closed-loop estimating error system for variable-speed wind turbine is established as a T-S fuzzy system. Resorting to the Lyapunov theory and linear matrix inequality technique, a set of sufficient conditions for co-solving the estimator gains and triggering matrices are produced. Ultimately, the effectiveness of the developed method is verified by some simulation results. Note to Practitioners—This paper investigates the T-S fuzzy-model-based proportional-integral state estimation for variable-speed wind turbine systems, addressing challenges raised from practical constrained network bandwidth, sampling outliers and multimodal deception attacks. The novel outlier-removal memory-event-triggered mechanism improves network utilization, and the fuzzy proportional-integral estimator enhances the state estimation accuracy and ensures the asymptotic stability for wind turbine systems with nonlinearities and multimodal deception attacks. Note that the investigated method can be applied to deal with the above practical issues in real-world systems, which can provide technique support for practitioners in real-world implementation. In addition, based on the estimated state, various state-feedback control methods can be realized to improve the control performance of wind turbine systems.}, archive = {J_TASE}, author = {Shen Yan and Xinyi Yang and Ju H. Park}, doi = {10.1109/TASE.2025.3585410}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17790--17800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Outlier-removal memory-event-triggered proportional-integral state estimation for wind turbine systems under multimodal deception attacks}, volume = {22}, year = {2025}, }
• (2025). Resilient sampled-data control for bipartite synchronization of cooperation-competition neural networks against denial-of-service attacks. TASE, 22, 17778--17789. (www)

  @article{ , abstract = {This paper deals with the bipartite synchronization problem of cooperation-competition neural networks (CCNNs) subject to denial-of-service (DoS) attacks. A resilient sampled-data control strategy is proposed to mitigate the adverse impact of DoS attacks, which takes both the attack signal and the periodic sampling communication protocol into account. The directed signed graph is introduced to characterize the cooperation and competition interactions among nodes. By leveraging coordinate transformation and graph theory techniques, a zero-row-sum Laplacian matrix is constructed to facilitate subsequent analysis. In combination with DoS attacks and control strategies, a tractable error system model is formulated. An interval-dependent function is further introduced, taking into account both attack intervals and data transmission intervals. Based on Lyapunov stability theory, the convex combination approach, and inequality techniques, the bipartite synchronization criteria for CCNNs are obtained. Moreover, the constructed interval-dependent function can improve the maximum allowable attack rate or reduce the minimum allowable coupling strength. The proposed control scheme is demonstrated to be effective and superior through the two numerical examples. Note to Practitioners—In many practical engineering and natural systems, cooperative and competitive behaviors often coexist and evolve dynamically. Directed signed graphs serve as an effective modeling tool for such interactions, in which positive and negative edge weights represent cooperation and competition, respectively. Bipartite synchronization provides a powerful framework for capturing these dynamics, offering a more accurate representation of real-world system behavior. However, networked control systems are increasingly vulnerable to DoS attacks, posing significant challenges to both robustness and control efficiency. To address these issues, this study proposes a resilient sampled-data control scheme that accounts for both periodic sampling protocols and DoS attacks. An interval-dependent function is constructed based on both the attack duration and data transmission intervals, thereby enhancing the tolerance to attacks or reducing the coupling strength. The effectiveness and superiority are validated through numerical examples, providing valuable insights into the secure coordination of multi-agent systems and the design of resilient industrial automation networks.}, archive = {J_TASE}, author = {Xindong Si and Zhen Wang and Xia Huang and Hao Shen}, doi = {10.1109/TASE.2025.3585403}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17778--17789}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient sampled-data control for bipartite synchronization of cooperation-competition neural networks against denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Disturbance-compression extended state observer with noise insensitivity: Application to electro-optical tracking system. TASE, 22, 17761--17777. (www)

  @article{ , abstract = {While augmenting the observer gain can diminish the estimation errors of the extended state observer (ESO), it inevitably leads to an amplification of the observing noise, thus creating the trade-off between high-gain observation and noise sensitivity. To address the trade-off, this paper introduces the disturbance-compression extended state observer (DC-ESO) from a frequency-domain perspective. By employing disturbance compression, the DC-ESO modifies the disturbance characteristics encountered by the observer, effectively diminishing the upper bounds of the disturbance and its derivative. Frequency-domain analysis indicates that the DC-ESO necessitates substantially lower observer gain than the ESO while maintaining the same estimation errors. The principles of disturbance compression and its effects on disturbance estimation and noise amplification are elucidated, revealing that the DC-ESO effectively alleviates the inherent conflict between high-gain observation and susceptibility to noise. The parameter tuning methodology and frequency-domain stability criteria are user-friendly for practicing engineers. The effectiveness of the proposed approach is empirically verified through its application to an electro-optical tracking system. Note to Practitioners—This work tackles the disturbance rejection problem in control systems, which encompasses disturbance including mechanical vibrations, friction, and parametric perturbations. Existing extended state observers (ESOs) rely on high gain (or high bandwidth) for disturbance suppression, which unfortunately results in substantial amplification of driving noise. In this research, we incorporate a disturbance compression channel that divides the compensation of disturbance into two separate stages. This approach effectively pre-compresses the disturbance, thus reducing the observational burden on the observer. We reconstruct the ESO in the frequency domain as a transfer function and present a frequency-domain disturbance compression ESO that is easily grasped by engineers. Accordingly, the proposed method is highly practical and effective according to experiments. Future research may consider expanding the targeted suppression to the specific harmonics disturbance and nonlinear system.}, archive = {J_TASE}, author = {Jiuqiang Deng and Wenchao Xue and Luyao Zhang and Qiliang Bao and Yao Mao}, doi = {10.1109/TASE.2025.3585348}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17761--17777}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Disturbance-compression extended state observer with noise insensitivity: Application to electro-optical tracking system}, volume = {22}, year = {2025}, }
• (2025). Retrosynthesis prediction via search in (Hyper) graph. TASE, 22, 17749--17760. (www)

  @article{ , abstract = {Predicting reactants from a specified core product remains a critical challenge in retrosynthesis prediction. While semi-template-based and graph-edit-based methods have shown promising results in accuracy and interpretability, they struggle to handle complex reactions. In this paper, complex reactions refer to chemical reactions involving the participation of multiple bonds, such as those involving the multiple reaction center or the same leaving group being attached to multiple atoms. To address these limitations, we propose RetroSiG (Retrosynthesis via Search in (Hyper) Graph), a semi-template-based framework that reformulates retrosynthesis as a two-phase search problem: (i) reaction center identification as a search task in the product molecular graph, and (ii) leaving group completion as a search task in the leaving group hypergraph. RetroSiG’s novel search mechanism systematically explores subgraphs by leveraging reinforcement learning to guide node selection at each step. This approach ensures connectivity and efficient decision-making, thereby enabling the effective handling of complex reaction predictions. In addition, RetroSiG incorporates the one-hop constraint, a domain-specific prior inspired by the observation that reaction center molecular subgraphs and leaving group subgraphs are almost always connected. This constraint focuses exploration on first-order neighbors, significantly reducing the search space and improving computational efficiency without compromising accuracy. Furthermore, RetroSiG leverages a hypergraph structure to model implicit dependencies among leaving groups, which enhances robustness for reactions with multiple leaving group configurations. Comprehensive experiments demonstrate RetroSiG’s competitive performance across standard benchmarks, while ablation studies confirm the contributions of its key design components, including the hypergraph and the one-hop constraint. Our results highlight RetroSiG’s scalability and effectiveness in handling diverse and complex retrosynthesis tasks. Note to Practitioners—To practitioners engaged in the field of organic synthesis, particularly those leveraging computational tools to predict retrosynthetic routes, the RetroSiG framework offers a powerful new approach. Motivated by the limitations of existing semi-template-based and graph-edits-based methods, which struggle with complex reactions involving multiple reaction centers or the attachment of the same leaving group to more than one atom, RetroSiG addresses these issues through a novel search mechanism. Our key finding is that by treating reaction center identification and leaving group completion as search problems within a product molecular graph and a leaving group hypergraph, respectively, RetroSiG can predict complex reactions more accurately. This approach not only enhances the interpretability of predictions but also exploits the implicit dependencies between leaving groups, making the method more robust. For practical application, we recommend integrating RetroSiG into existing retrosynthesis planning software to expand the scope of predictable reactions. However, practitioners should be aware that the effectiveness of the one-hop constraint and the leaving group hypergraph may vary depending on the complexity of the reactions being considered. Future work could focus on refining the search strategies and further exploiting the structural properties of hypergraphs to improve performance and generalizability.}, archive = {J_TASE}, author = {Zixun Lan and Binjie Hong and Maochun Xu and Zuo Zeng and Zhenfu Liu and Limin Yu and Fei Ma}, doi = {10.1109/TASE.2025.3585129}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17749--17760}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Retrosynthesis prediction via search in (Hyper) graph}, volume = {22}, year = {2025}, }
• (2025). A hybrid method based on multi-agent reinforcement learning and integer programming for dynamic slab design problems in steel industry. TASE, 22, 17734--17748. (www)

  @article{ , abstract = {This paper investigates a dynamic slab design problem in the steel industry, where order demands arrive dynamically during a given period. Slabs are the raw materials for producing order plates demanded by customers—slabs are first rolled in a rolling mill to create desired mother plates, and then the mother plates are cut into order plates. The dynamic nature of orders, along with practical considerations regarding rolling methods and nonlinear size constraints, distinguish our problem from existing ones. The goal is to determine slab design schemes to fulfill order demands for the period. However, the stochastic nature of dynamic production and the inherent complexity of slab design present significant challenges in the efficient solution. To address these challenges, we formulate a Partially Observable Markov Decision Process (POMDP) and propose a hybrid method (MARLIP) based on multi-agent reinforcement learning (MARL) and integer programming. The MARL component determines the order plate set, the slab type, and the rolling method involved in a single slab design. Then, an integer programming model is constructed to optimize the arrangement of the order plates based on MARL’s decisions. To further refine the MARLIP approach, we introduce dual factors and a soft boundary into the MARL and propose a two-core-based dynamic programming method to solve the integer programming model. Experimental results demonstrate our algorithm’s superior performance compared to several competitive reinforcement learning and mathematical programming methods in the dynamic slab design problem. Note to Practitioners—This paper is motivated by the medium-heavy plate production in the steel industry, where slabs are rolled into mother plates, and order plates required by customers are cut from the mother plates. This study is also applicable to other industries for which larger objects need cutting into smaller ones, such as the glass industry and paper industry. Current research mainly focuses on slab design for known batches of orders but falls short when dealing with dynamically arriving orders. We formulate a Partial Markov Decision Process to link slab design decisions over time, aiming to maximize the total slab utilization of slab designs for the given period. Then, we propose a hybrid method named MARLIP to maximize the total slab utilization of a given period by dynamically determining slab design schemes. At each time step, MARLIP determines a slab for producing order plates, a rolling method for rolling the slab into a mother plate, and the optimal arrangement of order plates on the mother plate. Experimental results indicate that MARLIP increases total profit, reduces total cutting loss, and utilizes fewer slabs to fulfill orders. By making rapid and automated decisions for dynamically arriving orders, MARLIP enables smart and automated production in the steel industry. In the future, we will possibly extend the proposed method to dynamic cutting and packing problems in manufacturing and then promote more practical applications.}, archive = {J_TASE}, author = {Baiyang He and Ying Meng and Lixin Tang}, doi = {10.1109/TASE.2025.3585108}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17734--17748}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hybrid method based on multi-agent reinforcement learning and integer programming for dynamic slab design problems in steel industry}, volume = {22}, year = {2025}, }
• (2025). Adversarial subgraph contrastive learning for predicting grasp stability of robotic hands with multimodal signals. TASE, 22, 17720--17733. (www)

  @article{ , abstract = {Accurate prediction of grasp stability is crucial for reliable and precise operations with multi-fingered robotic hands. Traditional methods tend to oversimplify tactile information and pay equal attention to all regions of the data. This can obscure subtle yet critical variations and introduce noise, increasing the risk of stability assessment errors. To address these challenges, a novel self-supervised method, Adversarial Subgraph Contrastive Learning (ASCL), is proposed. It constructs an instance graph from the spatial distribution and features of perceptual nodes. It employs a bi-level adversarial strategy to enhance latent data representations by maximizing the mutual information between the instance graph and its semantic subgraphs, while minimizing it with its noisy subgraphs. To prevent trivial solutions and continuous relaxation of semantic subgraphs, node confidence and edge connection terms are incorporated to ensure stabilization. From an information-theoretic perspective, ASCL exhibits notable advantages on unlabeled or sparsely labeled data, well outperforming existing methods in empirical tests with robotic hands. Note to Practitioners—Grasp stability prediction is vital for the monitor and control of robotic interactions, particularly in industrial and service applications where precision and reliability are important. This paper introduces a self-supervised graph contrastive learning method that constructs a graph structure from tactile and near-contact data in order to capture intricate spatial distributions and feature details. By employing a bi-level adversarial strategy, the method emphasizes meaningful substructures within data, highlighting critical variations while suppressing redundancy and noise. The experimental results demonstrate that significant improvements in grasp stability prediction for downstream tasks involving objects with varying attributes are achieved, effectively reducing the reliance on manual labeling. Furthermore, the model exhibits generalization capabilities to unseen objects. The work represents an advancement in improving robotic fine manipulation and autonomous decision-making skills for intelligent automated systems.}, archive = {J_TASE}, author = {Junjie Liao and Pengwen Xiong and MengChu Zhou and Peter X. Liu and Aiguo Song}, doi = {10.1109/TASE.2025.3585040}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17720--17733}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adversarial subgraph contrastive learning for predicting grasp stability of robotic hands with multimodal signals}, volume = {22}, year = {2025}, }
• (2025). Sampled-data-based load frequency control for islanded microgrids via a virtual inertia control mechanism. TASE, 22, 17710--17719. (www)

  @article{ , abstract = {This paper presents a robust load frequency control (LFC) scheme for an islanded microgrid (IMG) based on sampled-data control. First, a novel LFC system model for an IMG is proposed by incorporating a virtual inertia control (VIC) mechanism with a sampled-data-based auxiliary controller. Then, through applying Lyapunov theory and introducing an exponential decay rate (EDR) as a performance indicator, a novel stability criterion is established for the LFC of the IMG. Based on the stability criteria, the controller design is presented by utilizing $H_{\infty }$ performance index $\mu $ and EDR as conditions, resulting in a fast and robust LFC scheme. Finally, simulation results and comparison analysis show that the proposed method is superior, and the introduced VIC mechanism with a sampled-data-based auxiliary controller can effectively improve the dynamic performance of the system. Note to Practitioners—The high penetration of renewable energy sources (RESs) reduces the inertia of the IMG, leading to a fast and large frequency deviation after a contingency, and desiring fast and robust LFC. This paper focuses on designing a robust LFC scheme to address issues like these rapid and significant frequency deviations after contingencies. Specifically, we propose a novel LFC system model for IMGs by incorporating a VIC mechanism with a sampled-data-based auxiliary controller into the LFC system to supply virtual inertia power to the system under high RES penetration. To achieve rapid frequency stabilization, we introduce the EDR as a performance index to guide the controller design. To meet various requirements within the LFC design, a comprehensive design methodology is outlined, integrating the sampling period, $H_{\infty }$ performance index $\mu $ , and EDR. Simulation results and comparative analysis validate that the proposed LFC scheme effectively compensates for the adverse effects of low inertia in IMGs and achieves the desired stability.}, archive = {J_TASE}, author = {Wei-Min Wang and Yan-Wu Wang and Hong-Bing Zeng and Xiao-Kang Liu}, doi = {10.1109/TASE.2025.3584985}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17710--17719}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sampled-data-based load frequency control for islanded microgrids via a virtual inertia control mechanism}, volume = {22}, year = {2025}, }
• (2025). Fast and accurate category-level object pose estimation without shape priors for robotic grasp detection. TASE, 22, 17696--17709. (www)

  @article{ , abstract = {Category-level object pose estimation is crucial for enabling robot grasping. Currently, many methods rely on 3D shape priors for pose estimation, but obtaining priors for specific categories often requires a significant amount of time to generate. Although some methods do not rely on priors, However, these methods struggle to achieve a balance between speed and accuracy. Achieving fast and accurate category-level object pose estimation remains a challenging issue. In this paper, we propose an algorithm called FAPose, which aims to simultaneously achieve speed and accuracy in category-level object pose estimation without any shape priors. Firstly, we design an RGB-point cloud feature aggregation method based on transformers to fuse RGB and point cloud features. Secondly, we develop a dual-constraint object pose estimation method to effectively leverage both feature space and geometric space information. This approach constructs pose constraints in both feature space and geometric space to achieve optimal pose prediction. Finally, we validate FAPose through benchmark datasets and real robot grasping experiments. The experimental results demonstrate that our proposed method surpasses most existing state-of-the-art pose estimation methods, achieving superior pose estimation performance. The code for FAPose will be available after the paper is accepted for publication. Note to Practitioners—Robotic grasp detection based on category-level 6D object pose estimation is an important way to realize reliable robotic grasp. How to achieve fast and accurate object pose estimation is important for the real-world application. Currently, there are many category-level object pose estimation methods, which can be divided into two categories: methods with priors and methods without priors. Methods with priors rely on object prior models, but they require a significant amount of time to generate, which is not ideal for practical applications. As for methods without priors, although they do not rely on object priors, they struggle to achieve a balance between real-time performance and accuracy. To solve these problems, in this paper, we try to develop a fast and accurate category-level object pose estimation network and apply it to the robotic grasp tasks in the real world. Experimental results show that our proposed method outperforms most realated methods on benchmark datasets. What’s more, the proposed network consumes less time in the pose estimation process, and achieves a fast and accurate pose estimation in the real world.}, archive = {J_TASE}, author = {Sheng Yu and Jian Yin and Di-Hua Zhai and Yuanqing Xia}, doi = {10.1109/TASE.2025.3584816}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17696--17709}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast and accurate category-level object pose estimation without shape priors for robotic grasp detection}, volume = {22}, year = {2025}, }
• (2025). Ablation study on features in learning-based joint calibration of cable-driven surgical robots. TASE, 22, 17680--17695. (www)

  @article{ , abstract = {Surgical robots equipped with cable-driven mechanisms have flexible, light, and compact arms and tools. However, cable slack, stretch, and gear backlash introduce unavoidable errors from motor positions to joint positions and the end-effector pose. This paper presents a learning-based joint position calibration method for the RAVEN-II surgical robot, employing deep neural networks and gated recurrent units. Compared to fixed offset compensation, the learning-based calibrations reduce the joint position errors by over 62.4% (unloaded) and 54.8% (loaded). Furthermore, removal and inaccurate ablation studies on input features identify that raw joint positions and motor torques are the most important model inputs for calibration accuracy. These studies also reveal that the models are capable of inferring joint positions from the end-effector pose and prioritize the direction of motor torques over their amplitude. When guided appropriately, the models can also compensate for encoder value inconsistencies occurring with robot re-homings. By excluding the unnecessary input features, lightweight models are developed and achieve better performance and efficiency simultaneously, reducing the training time on the CPU to 2.5 minutes. All data and code are open-source at https://github.com/uw-biorobotics/RAVEN-2-Feature-Ablation Note to Practitioners—This paper presents a data-driven neural network/AI based calibration method for accurate joint position estimation on a cable-driven research surgical robot (RAVEN-II). We aim to understand our learning calibration by studying the importance of each input feature using two ablation methods. We selectively eliminate (removal ablation) or add distorting noise to (inaccurate ablation) input features one at a time, and retrain the calibration model. Then, the increase in error suggests the importance of the target input feature. By excluding unnecessary input features and thus reducing the input dimension, the size of the machine learning model can be reduced without losing accuracy. The simpler model can be trained and inferred efficiently without GPU acceleration. All data and code used in this paper, including robot control, data collection, and model training, are available online.}, archive = {J_TASE}, author = {Haonan Peng and Andrew Lewis and Yun-Hsuan Su and Blake Hannaford}, doi = {10.1109/TASE.2025.3584757}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17680--17695}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Ablation study on features in learning-based joint calibration of cable-driven surgical robots}, volume = {22}, year = {2025}, }
• (2025). Sequential image restoration and segmentation for interface detection in primary separation cells. TASE, 22, 17665--17679. (www)

  @article{ , abstract = {The primary separation cell (PSC) plays a key role in bitumen recovery during oil sands extraction. Controlling the froth-middling interface level in PSC is vital for optimal bitumen recovery. Existing sensors for the interface level measurement can be either costly or less reliable. This paper introduces an image restoration algorithm to enhance degraded images of PSC sight glasses followed by an image segmentation technique, which can serve as an alternative for interface level estimation. The restoration algorithm uses a regular state-space model with a skew-t distribution for measurement noise to account for image contamination. States and parameters are estimated using an expectation-maximization (EM) algorithm along with a robust Kalman filter (KF). The restored images are segmented using a Gaussian mixture model (GMM) with Markov random fields (MRF) for interface detection. Experimental results on a lab-scale PSC demonstrate the method’s effectiveness in improving interface level estimation compared to the existing models. Note to Practitioners—Accurate detection and control of the froth-middling interface is crucial for maintaining product quality and maximizing oil recovery in oil sand extraction. This paper presents an innovative approach that combines adaptive image restoration techniques to enhance image quality and a segmentation algorithm for more accurate interface level detection in PSCs. Through rigorous experimentation, our method improves accuracy and certainty in interface detection, offering a valuable solution for enhanced oil recovery processes.}, archive = {J_TASE}, author = {Amir Mohseni and Yousef Salehi and Ranjith Chiplunkar and Biao Huang}, doi = {10.1109/TASE.2025.3584739}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17665--17679}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sequential image restoration and segmentation for interface detection in primary separation cells}, volume = {22}, year = {2025}, }
• (2025). Data-based adaptive event-triggered transfer stabilization for nonlinear networked systems. TASE, 22, 17654--17664. (www)

  @article{ , abstract = {This paper investigates the adaptive event-triggered data-driven control problem for a class of unknown nonlinear discrete networked systems. To address this problem, a stochastic configuration network-based algorithm is developed to construct a candidate mapping set within the modeling-valid domain. Subsequently, an adaptive event-triggered control protocol is proposed, and a closed-loop mapping set is obtained. Then, by leveraging the ideas of transfer stabilization and the S-lemma, a data-driven stability criterion for nonlinear discrete networked systems is derived. The stability criterion solely relies on the data of the controlled system and is independent of both the system model and the data model. Based on this stability criterion, the control gain and triggering matrix of the controlled system can be obtained. Additionally, the effectiveness and practicality of the proposed method are validated through a numerical example and a complex memristive Hopfield neural network circuit. Note to Practitioners—The increasing complexity of systems leads to higher data density, which in turn imposes significant computational demands, particularly when system models are unavailable, making efficient computation a challenging task. Neural network-based modeling techniques provide powerful tools for reconstructing system models. However, the high-dimensional weight matrices introduced during modeling, along with the need to address additional modeling errors, often intensify computational complexity. This paper investigates adaptive event-triggered data-driven control for unknown nonlinear discrete networked systems, presenting an enhanced modeling algorithm based on stochastic configuration neural networks to avoid high-dimensional weight matrices, along with an adaptive event-triggered transfer stabilization strategy that accounts for modeling errors. The proposed approach is validated through its application to a complex memristive Hopfield neural network circuit, demonstrating its efficiency, practicality, and effectiveness in tackling real-world challenges.}, archive = {J_TASE}, author = {Qingyu Shi and Xia Huang and Guozeng Cui and Zhen Wang and Hao Shen}, doi = {10.1109/TASE.2025.3584190}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17654--17664}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-based adaptive event-triggered transfer stabilization for nonlinear networked systems}, volume = {22}, year = {2025}, }
• (2025). Estimation and control of 3D diffusion-advection system by partial differential neural network combined with divide-space sampling strategy. TASE, 22, 17642--17653. (www)

  @article{ , abstract = {In this paper, an optimization framework of partial differential neural network combined with divide-space sampling (PDNN-DSS) is proposed to solve the estimation and control problem of 3D diffusion-advection systems. The PDNN-DSS framework contains a divide-space sampling observer, a model prediction estimator, and a partial differential neural network solver. First, the open-loop state of the diffusion-advection system is obtained by designing the division space observer. The observer provides a deployment scheme for sensors so that mobile robots only need to carry actuators not sensors. Then, the control problem is transformed into an optimization problem by designing a model prediction estimator. Subsequently, a novel partial differential neural network is designed. The neural network is based on partial differential equations containing spatial diffusion features and can solve spatio-temporal optimization problems. Finally, the effectiveness of the proposed method in this paper is demonstrated through simulations from multi-dimensional perspectives of 2D and 3D diffusion-advection systems. Note to Practitioners—Spatio-temporal processes such as marine oil spills and riverwater chemical pollution can be modeled as diffusion-advection systems. This paper utilizes mobile robots and stationary sensors to solve the estimation and control problems of 3D diffusive-advection systems. In the proposed optimization framework, the divide-space sampling observer is used to deploy the stationary sensors, and the partial differential neural network is used to solve the guidance law of the mobile robot. Naturally, if the spatially defined domain of the diffusion-advection system is 2D or 1D, the method of this paper is equally applicable. In addition, if the system does not have an advection term, the design idea of the method in this paper remains the same, but the derivation needs to be simplified.}, archive = {J_TASE}, author = {Xingru Li and Zhijun Zhang}, doi = {10.1109/TASE.2025.3584121}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17642--17653}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Estimation and control of 3D diffusion-advection system by partial differential neural network combined with divide-space sampling strategy}, volume = {22}, year = {2025}, }
• (2025). Global model recovery anti-windup control with prescribed performance for saturated systems. TASE, 22, 17631--17641. (www)

  @article{ , abstract = {In this paper, we investigate the problem of reference tracking for a class of linear systems subject to actuator saturation and propose a global model recovery anti-windup (MRAW) strategy with prescribed performance. We adopt the classic MRAW framework, where one of the outputs of the anti-windup compensator is regarded as the tracking error, reflecting the difference between the unconstrained system and the saturated system. Then, instead of employing the commonly used ${\mathcal {L}}_{2}$ gain, we present the prescribed performance functions (PPFs) to characterize the real-time tracking error such that the transient performance can be captured more specifically. In particular, to avoid singular issues arising from the occurrence of saturation, we modify the existing PPFs with an auxiliary system when saturation occurs. Based on these modified performance functions, we follow a modified performance control approach and design the remaining output that is injected into the anti-windup activation module. Such a design procedure is constructive, making it conducive to its extension to nonlinear systems. Theoretical results establish the boundedness of all signals in the closed-loop system. Simulation results verify the effectiveness of our design strategy. Note to Practitioners—This paper is motivated by the commonly encountered actuator saturation in practical systems. In this paper, under the prescribed performance control (PPC) framework, a novel model-recovery anti-windup design, combined with an auxiliary system, is proposed to handle adverse effects caused by actuator saturation. Its possible application to existing controllers makes such a strategy a popular choice with practitioners. Differently from the existing PPC-based works, most of which pertain to semi-global nature, the strategy presented in this paper eliminates the initial condition dependence restriction and renders global results. Unlike the Linear Matrix Inequality (LMI)-based anti-windup designs, the proposed design can be extended to a class of nonlinear systems, which encompass a fairly general class of practical systems such as aerospace and mechatronic systems.}, archive = {J_TASE}, author = {Wenxin Lai and Yuanlong Li and Zongli Lin}, doi = {10.1109/TASE.2025.3584102}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17631--17641}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Global model recovery anti-windup control with prescribed performance for saturated systems}, volume = {22}, year = {2025}, }
• (2025). Soft contact simulation and manipulation learning of deformable objects with vision-based tactile sensor. TASE, 22, 17618--17630. (www)

  @article{ , abstract = {Deformable object manipulation is a challenging problem due to its complex deformable properties. With the development of artificial intelligence, learning-based methods have shown outstanding performance in robotic manipulation. Previous works have investigated the manipulation of deformable objects via Reinforcement Learning (RL) in simulation. However, they approximate object deformation with particles, using particle states as observations, which are unavailable in reality. To address these issues, we utilize Vision-Based Tactile Sensors (VBTSs) as the end-effector to manipulate and observe the deformable objects. In this work, we develop a new contact simulation environment for deformable objects, including elastic, plastic, and elastoplastic. We utilize RL strategies and expert demonstrations to train agents in the simulation. Finally, we build a real experimental platform to complete the sim-to-real tasks and robustness testing. Our work introduces an innovative strategy that utilizes high-resolution VBTSs for contact simulation and manipulation of deformable objects. The experimental results show superior performances of deformable object manipulation with the proposed method. Note to Practitioners—The manipulation of deformable objects is a classic and challenging contact-rich robotic task. Compared to rigid objects, deformable objects have higher deformation characteristics, which poses a higher challenge to robot manipulation. As a soft optical sensor, the Vision-based tactile sensor can observe the deformation state of a deformable object while manipulating it, providing a new idea for deformable object manipulation. In this work, we developed a soft contact simulator between the deformable object and the vision-based tactile sensor. The simulator can simulate the contact between the sensors and objects with elastic, plastic, and elastoplastic. Upon this simulator, we proposed a benchmark for deformable object manipulation based on vision-based tactile sensors, including transferable observations, tasks, and expert demonstration systems. Finally, we built a corresponding experimental platform to complete the Sim-to-real and robustness experiments of related tasks. The experimental results show that our method is reliable.}, archive = {J_TASE}, author = {Yuhao Sun and Shixin Zhang and Zixi Chen and Zirong Shen and Fuchun Sun and Cesare Stefanini and Di Guo and Shan Luo and Jianwei Zhang and Jianhua Shan and Bin Fang}, doi = {10.1109/TASE.2025.3582341}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17618--17630}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Soft contact simulation and manipulation learning of deformable objects with vision-based tactile sensor}, volume = {22}, year = {2025}, }
• (2025). Discretized data-driven neural dynamics for model-adaptive kinematic control of redundant manipulators. TASE, 22, 17606--17617. (www)

  @article{ , abstract = {This paper proposes discrete-time learning algorithms that utilize a data-driven technology to address the uncertain issues of optimization and structure. The main challenge lies in acquiring accurate optimization indices and Jacobian matrix, which can be addressed through iterative estimations enabled by these algorithms. On this basis, we propose a new model-adaptive kinematic control (MAKC) scheme for redundant manipulators without prior structure knowledge, incorporating the estimated optimization index and Jacobian matrix. To solve this scheme, a discretized data-driven neural dynamics ( $\text {D}^{3}$ ND) controller is proposed based on the 94LVI algorithm, Kalman filter, and discrete-time learning algorithms. Theoretical analysis is provided to demonstrate its convergence. Subsequently, simulations and experiments are carried out on redundant manipulators using manipulability and joint drift as performance criteria. The results substantiate the robustness, practicability, and superiority of the proposed controller when encountering uncertain issues. Note to Practitioners—Traditional optimization schemes for kinematic control of redundant manipulators may encounter structural and optimization uncertainties due to modifications or environmental factors, which reduce their practicality. To address uncertain issues, this paper proposes discrete-time learning algorithms to estimate the Jacobian matrix and optimization index in real time. Based on the estimated information, an MAKC scheme is designed to achieve kinematic optimization for redundant manipulators without prior structure knowledge. Furthermore, a $\text {D}^{3}$ ND controller is developed, integrating the 94LVI algorithm, Kalman filter, and discrete-time learning algorithms for optimization control, noise resistance, and parameter estimation, respectively. The global convergence of the proposed controller is demonstrated through three theorems. The advantages of the $\text {D}^{3}$ ND controller in addressing uncertain issues are evident from simulative and experimental results when compared to existing methods.}, archive = {J_TASE}, author = {Xin Ma and Zhengtai Xie and Mei Liu}, doi = {10.1109/TASE.2025.3582240}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17606--17617}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discretized data-driven neural dynamics for model-adaptive kinematic control of redundant manipulators}, volume = {22}, year = {2025}, }
• (2025). Cooperative quantized event-based fuzzy tracking control of nonlinear autonomous surface vehicles with prescribed performance. TASE, 22, 17594--17605. (www)

  @article{ , abstract = {This paper investigates the cooperative fuzzy tracking control of nonlinear autonomous surface vehicles with input quantization and event-triggered mechanism. The proposed cooperative control scheme consists of two parts: (i) the distributed observer and (ii) the dynamic event-based fuzzy tracking controller. The distributed observer is designed to obtain the nonlinear leader’s trajectory information on a directed communication topology. Under this framework, uncertain nonlinearity within the vehicle model is approximated through fuzzy logic systems, and, according to the state of the distributed observer, the dynamic event-based adaptive fuzzy tracking control law is developed with an input switching quantizer. Furthermore, a prescribed performance method is introduced to ensure the transient performance of tracking errors and obtain zero-tracking errors ultimately, which is proved through Lyapunov stability theory. Finally, the effectiveness of the proposed control strategy is verified by simulation experiments. Note to Practitioners—In marine operations, autonomous surface vehicles face challenges including limited network resources and model nonlinearity. This paper proposes a tracking control strategy to address these issues by integrating switching quantization and dynamic event-triggered mechanism for control input signals, which can reduce communication burden while avoiding Zeno behavior, with prescribed performance introduced to ensure transient performance. For the nonlinear and uncertain aspects of actual autonomous surface vehicle models, fuzzy logic systems are employed to approximate the nonlinear components. Additionally, in tracking control scenarios with nonlinear leader trajectory, a distributed observer is proposed to obtain the nonlinear trajectory information. The proposed method can ensure zero-tracking errors, achieving more precise tracking control under the complex nonlinearity, uncertainty, and limited network resources encountered in real-world scenarios.}, archive = {J_TASE}, author = {Shanling Dong and Zhiyi Lai and Zheng-Guang Wu and Meiqin Liu and Guanrong Chen}, doi = {10.1109/TASE.2025.3581224}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17594--17605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative quantized event-based fuzzy tracking control of nonlinear autonomous surface vehicles with prescribed performance}, volume = {22}, year = {2025}, }
• (2025). Adaptive event-triggered sampled-data fuzzy security control for nonlinear delayed DPSs with DoS attacks and stochastic actuator failures. TASE, 22, 17581--17593. (www)

  @article{ , abstract = {This article addresses adaptive event-triggered sampled-data (SD) fuzzy security control under spatially local averaged measurements (LAMs) for nonlinear delayed distributed parameter systems (DPSs) with denial of service (DoS) attacks and stochastic actuator failures. Firstly, a Takagi-Sugeno (T-S) fuzzy model of delayed partial differential equations (PDEs) is introduced to precisely characterize the dynamic behavior of nonlinear delayed DPS. Secondly, an adaptive event-triggered SD fuzzy security control strategy is designed under DoS attacks and stochastic actuator failures, which can be flexibly modified in accordance with the present sampling and the most recently transmitted signals, and is implemented utilizing a restricted number of sensors and actuators. Subsequently, by establishing a Lyapunov functional, sufficient conditions that guarantee the mean square exponential stability of closed-loop nonlinear delayed DPSs are obtained based on linear matrix inequalities (LMIs). Finally, two examples are provided and the presented controller are compared to demonstrate the applications and advantages of the proposed approach. Note to Practitioners—Cyber-attacks, time-delays, and stochastic actuator failures have always been of great concern in security control because they occur frequently in practical DPSs. Consequently, when the DPSs is affected by these factors, it can lead to degraded performance and even system instability. In particular, existing security control strategies for DPSs are now based almost on time-triggered or regular event-triggered control mechanisms, resulting in high operating costs for security controllers. To overcome the challenge, adaptive event-triggered SD fuzzy security control strategy is proposed in this paper. By utilizing appropriate Lyapunov functional and inequality techniques, the proposed method in this paper novelty translates the problem of adaptive event-triggered SD fuzzy security controller design into a standard LMI feasibility problem. Several numerical examples are given to show the effective of the proposed control method, which contain the temperature regulation for a high-speed aerospace vehicle via the proposed control strategy. With the method proposed in this paper, the adaptive event-triggered SD fuzzy security controller requires only a small number of measurements to enable good performance when subjected to cyber-attacks, time-delays, and stochastic actuator failures, with significant cost savings, which is important for the improvement of stability and performance of nonlinear delayed DPSs.}, archive = {J_TASE}, author = {Zi-Peng Wang and Bo-Ming Chen and Feng-Liang Zhao and Junfei Qiao and Huai-Ning Wu and Tingwen Huang and Guangwei Chen}, doi = {10.1109/TASE.2025.3584104}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17581--17593}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive event-triggered sampled-data fuzzy security control for nonlinear delayed DPSs with DoS attacks and stochastic actuator failures}, volume = {22}, year = {2025}, }
• (2025). Incremental rank continual dictionary learning for multimode process monitoring with continually emerging operational mode. TASE, 22, 17568--17580. (www)

  @article{ , abstract = {Modern industrial processes typically operate under variable conditions due to fluctuations in raw materials and production loads. Consequently, accurate monitoring across multiple operational modes is essential for process optimization. Although some global modeling based methods have been proposed, they struggle to adapt to the continual emergence of new modes. On the other hand, online updating monitoring methods often suffer from “catastrophic forgetting”, i.e., they lose the ability to represent historical modes. To address these challenges, this article proposes a multimode process monitoring method based on Incremental Rank Continual Dictionary Learning (IRCDL). The method integrates the incremental learning framework based on low-rank matrix augmentation with the adaptive weight selector, allowing the dictionary to retain the knowledge of both historical and new modes. Specifically, to address the issue of information redundancy in overcomplete dictionaries, an innovative dictionary construction method based on low-rank matrices weighting is proposed. Using low-rank matrices as the fundamental units for dictionary construction, this method accurately captures the intrinsic features of different modes while significantly reducing redundant parameters. Then, to handle the continual emergence of new modes, an incremental learning framework based on low-rank matrix augmentation is developed. This framework continually adds new low-rank matrices to learn representations of new modes, while keeping the historical low-rank matrices intact to maintain “zero-forgetting” of historical modes, thereby enabling accurate monitoring of multimode processes. Finally, an adaptive weight selector is designed to automatically optimize the weights of low-rank matrices, thereby enhancing the representation of different modes and significantly improving the performance of multimode monitoring. Extensive experiments have demonstrated that the proposed method achieves state-of-the-art continual learning capabilities (i.e., “zero forgetting”) and exceptional process monitoring performance. Note to Practitioners—Motivated by the fact that the real industrial processes typically operate under variable conditions and traditional online updating monitoring methods struggle to adapt to processes where operational modes frequently vary, as they often suffer from catastrophic forgetting when the model is updated to adapt to new modes continually, this article proposes a zero-forgetting continual dictionary learning method for multimode process monitoring with continually emerging operational mode. The proposed method can capture the intrinsic features of different modes by low-rank matrices, allowing dictionary to retain and adaptively fuse the knowledge of historical and new modes, so as to realize accurate monitoring of multimode process. Extensive experiments show that the proposed method has the state-of-the-art continual learning ability and process monitoring performance, it can maintain the monitoring performance for historical modes after updating while it is difficult for other methods. Above all, it is verified as suitable for multimode process monitoring.}, archive = {J_TASE}, author = {Keke Huang and Weiyi Feng and Zixuan Chen and Dehao Wu and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2025.3582733}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17568--17580}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Incremental rank continual dictionary learning for multimode process monitoring with continually emerging operational mode}, volume = {22}, year = {2025}, }
• (2025). Building accurate exchange-correlation functional for density functional theory through data analytics and optimization. TASE, 22, 17552--17567. (www)

  @article{ , abstract = {Density functional theory (DFT) forms the foundation for computing the electronic structure of many-electron systems through the Kohn-Sham (KS) equations, with electron density $\mathbf {\mathit {n}}(\mathbf {r})$ serving as the fundamental variable. This approach allows for the understanding and manipulation of material properties at the atomic and electronic levels. However, the exchange-correlation (XC) functional in DFT is not known in an exact form and is typically approximated using analytical models, which limits the accuracy of DFT calculations. The recent advancements in data analytics techniques, such as machine learning, which excel in pattern recognition, offer new opportunities for more effective approximations of the XC functional. In this study, we introduce a novel approach for XC functional modeling by fusing shared representation learning. This method leverages information sharing across tasks to improve the fused XC (FXC) functional model generalization. To improve the robustness of the FXC functional model, we further develop a multi-objective constrained multitask adjustment mechanism. Additionally, to expedite the determination of the optimal FXC model, we propose a group-theory-based multi-objective optimization (MOO) algorithm, which enables faster exploration of the solution space. Our experimental results demonstrate that the KS equations incorporating the FXC functional substantially improve the accuracy of molecular atomization energies, ionization potentials, and electron density calculations. Moreover, the method exhibits applicability to the computation of isomerization reaction energies in large molecules. These findings validate the effectiveness of our data analytics and optimization-based approach in enhancing the precision of the XC functional approximation. Note to Practitioners—Density functional theory (DFT) provides theoretical support for addressing many challenges in industrial processes. However, the exact form of the exchange-correlation (XC) functional remains unknown, which limits the computational accuracy of DFT. Additionally, modeling the XC functional based on traditional mechanism analysis presents considerable difficulty and barriers. To overcome these limitations, we propose a novel approach for accurate XC functional modeling using data analytics and optimization methodology. The FXC functional developed herein not only improves the solution accuracy of the KS equations but also enhances the generalization and robustness in predicting multiple molecular properties. This method provides a streamlined and efficient pathway for XC functional modeling in DFT. Extensive experimental validation confirms the effectiveness of the FXC functional. Looking ahead, this approach could offer valuable insights for solving real-world industrial problems, including material design and optimization, and device performance analysis in nanotechnology or semiconductor manufacturing.}, archive = {J_TASE}, author = {Junfeng Zhao and Lixin Tang and Jiyin Liu and Jian Wu and Xiangman Song}, doi = {10.1109/TASE.2025.3582140}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17552--17567}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Building accurate exchange-correlation functional for density functional theory through data analytics and optimization}, volume = {22}, year = {2025}, }
• (2025). Enhancing performance in Human–Robot collaboration: A modular architecture for task scheduling and safe trajectory planning. TASE, 22, 17535--17551. (www)

  @article{ , abstract = {The integration of robots into shared workspaces alongside humans is the basis of Human-Robot Collaboration (HRC). This field of research has changed the paradigm of the industrial context, making HRC of pivotal importance for both researchers and the industry. In this context, a suitable task scheduling and trajectory planning strategy are crucial to achieve good performances and create a synergy between the two actors. Indeed, the task scheduling should be able to optimally distribute the tasks between the actors and recover from possible failures, i.e. by rescheduling the tasks. The trajectory planning strategy must comply with the safety standards that impose a reduction of velocity based on human behaviour. To this end, the monitoring system must also be safe-certified; otherwise, safety cannot be guaranteed. This paper proposes a novel architecture that integrates a dynamic task scheduling module with a dynamic trajectory planning module that explicitly considers ISO/TS 15066. For this purpose, the framework exploits a secure and certified monitoring system capable of tracking the human operator even in case of occlusions. The overall platform has been extensively validated both in a real and complex industrial scenario within the context of the ROSSINI EU project, where a dual-arm mobile robot collaborates with a human operator in an automatic machine-tending operation, and in a mock-up scenario. Note to Practitioners—This paper focuses on the implementation of a modular and integrated architecture to handle the challenges associated with effective task scheduling and trajectory planning, with the aim of improving the performances of collaborative robots in real HRC scenarios. The proposed methodology is structured into three distinct layers. Firstly, the safety layer monitors in real-time the human operator and provides the other layers with the maximum admissible robot speed. Secondly, the scheduling layer distributes tasks between humans and robots efficiently, recovering from possible failures and rescheduling the tasks, if necessary. Lastly, the flexible execution layer plans collision-free paths and maximizes the robot speed while explicitly considering ISO/TS 15066 guidelines. Thus, it ensures that the robot behaviour is compliant with the safety standards. Although the overall architecture is fully integrated, each layer is designed to function independently. Practitioners have the flexibility to utilize each layer as a standalone module if desired. Moreover, the architecture has been validated in a real complex industrial scenario, proving its practical applicability, robustness, and effectiveness.}, archive = {J_TASE}, author = {Andrea Pupa and Simone Comari and Mohammad Arrfou and Gildo Andreoni and Alessandro Carapia and Marco Carricato and Cristian Secchi}, doi = {10.1109/TASE.2025.3574627}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17535--17551}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing performance in Human–Robot collaboration: A modular architecture for task scheduling and safe trajectory planning}, volume = {22}, year = {2025}, }
• (2025). Early model-based safety analysis for collaborative robotic systems. TASE, 22, 17523--17534. (www)

  @article{ , abstract = {The current era is marked by an accelerated digitization of manufacturing processes, with robotic systems increasingly integrated into various workflows. Yet, despite significant advancements, it is impractical to fully automate certain tasks due to prohibitive costs and technical constraints. As a result, there’s a growing emphasis on human-robot collaboration (HRC) for intricate operations. In HRC scenarios, humans and robots co-inhabit the same work environment, operating side by side. More than just mere coexistence in the same space, they actively collaborate on shared tasks, thus raising the stakes in terms of safety. The dynamic behavior of robots must be synchronized with the anticipated and unexpected human actions, adding another layer of complexity to the safety considerations. It is essential to conduct comprehensive safety analyses that identify potential risks that pose harm to the human operator. As a proactive measure to foster early-stage safety and risk analysis, we propose the use of goal models. The approach enables the specification of safety threats within the HRC context, thereby facilitating the development of safety tasks and supportive monitoring mechanisms. This approach helps in the refinement and implementation of safety measures, ensuring a secure and productive environment for human-robot collaboration. Note to Practitioners—This paper was motivated by the need for conducting thorough safety analyses in future manufacturing scenarios. This is particularly the case as the use of HRC becomes more common in industry. This is fostered by the ideas of smart manufacturing and the spread of collaborative robots (cobots) in work environments. To support proper safety analyses in early phases of development, we propose a model-based methodology to relate the requirements of the system with the corresponding safety hazards and its countermeasures. The proposed model-based approach places particular emphasis on the identification of hazards not only related to a single system but resulting from collaboration scenarios. Thus, it contributes to an extensive safety analysis of collaborative robotic systems right in the early development stages. Application to a manufacturing case example showed its applicability and usefulness to identify potential safety hazards early in development. Results, however, also indicate that future work is needed for implementing a seamless view concept allowing the derivation of models representing specific situations to be closely investigated. For this, abstraction mechanism can aid in guiding the safety analysis procedure.}, archive = {J_TASE}, author = {Meenakshi Manjunath and Jeshwitha Jesus Raja and Marian Daun}, doi = {10.1109/TASE.2024.3440048}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17523--17534}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Early model-based safety analysis for collaborative robotic systems}, volume = {22}, year = {2025}, }
• (2025). Decoupling-based LPV observer for driver torque intervention estimation in Human–Machine shared driving under uncertain vehicle dynamics. TASE, 22, 17512--17522. (www)

  @article{ , abstract = {This paper proposes a method for simultaneous estimation of both the driver torque and the sideslip angle within the context of human-machine shared driving control for autonomous ground vehicles. To this end, the driver torque is considered as an unknown input (UI) and the sideslip angle is an unmeasured state of the vehicle dynamics system. For simultaneous estimation purpose, a decoupling-based technique is leveraged to design an unknown input observer (UIO). The UIO design goal is to decouple the effect of the unknown driver torque while minimizing the influence of the modeling uncertainties, considered as unknown exogenous disturbances, from the lateral tires forces and the steering system. Linear parameter-varying (LPV) framework is used to deal with the time-varying nature of the vehicle longitudinal speed. Based on Lyapunov stability theory, we derive sufficient conditions, expressed in terms of linear matrix inequality (LMI) constraints, for LPV unknown input observer design. The simultaneous vehicle estimation is reformulated as a convex optimization problem, where the modeling uncertainty influence can be minimized via the $\ell _{\infty} -$ gain performance. Hardware-in-the-loop (HiL) tests are performed with the SHERPA dynamic simulator and a human driver to show the effectiveness of the proposed UIO-based estimation method, especially within the cooperative driving control framework. Note to Practitioners—We present a method to jointly estimate the driver torque and the sideslip angle in the context of human-machine shared driving. To this end, we consider the driver torque as an unknown input and treat the sideslip angle as an unmeasured state of the vehicle dynamics system. The core of our method lies in the application of a decoupling-based technique to design an unknown input observer. The primary objective of this UIO is to effectively decouple the influence of the unknown driver torque while mitigating the impact of modeling uncertainties, considered as unknown exogenous disturbances, on the lateral tire forces and the steering system. Using an LPV framework has allowed the time-varying nature of the vehicle longitudinal velocity to be effectively addressed. Via Lyapunov stability theory, we have established sufficient conditions, expressed in terms of LMI constraints, for the design of the LPV unknown input observer. The proposed simultaneous vehicle estimation method has been reformulated as a convex optimization problem, allowing to minimize the influence of modeling uncertainties. To show the effectiveness of the proposed UIO-based estimation method, we have conducted extensive HiL tests using the SHERPA dynamic simulator with a human driver. The real-time experiments demonstrate the effectiveness of the proposed method, especially with respect to related estimation results in the literature, within the cooperative driving control framework. The proposed LPV estimation method contributes to the advancement of the field of autonomous ground vehicles by providing practitioners with a robust tool for joint estimation of essential variables critical for effective vehicle control and safety in the context of human-machine cooperative driving.}, archive = {J_TASE}, author = {Anh-Tu Nguyen and Thierry-Marie Guerra and Chouki Sentouh and Jean-Christophe Popieul}, doi = {10.1109/TASE.2024.3394240}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17512--17522}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Decoupling-based LPV observer for driver torque intervention estimation in Human–Machine shared driving under uncertain vehicle dynamics}, volume = {22}, year = {2025}, }
• (2025). ETS-based Human–Machine robust shared control design considering the network delays. TASE, 22, 17501--17511. (www)

  @article{ , abstract = {This paper proposes an event-triggered control method for the CAN-network delayed human-machine shared steering system. The uncertain model parameter of vehicle speed is handled by the polytypic technology, and then represented by a new state with fewer vertices. Thus, a driver-vehicle path-tracking model is built. After that, the communication model of the shared control scheme considering the CAN network delays is redefined by the event-triggered system (ETS). Instead of employing the periodic communication from the sensor to the controller, it only occurs when the triggered condition of ETS is satisfied. Such a design can reduce the communication load and improve the usage of network resources. Finally, a robust state-feedback controller with the parallel distribution method is adopted to guarantee vehicle path-tracking accuracy, while reducing driver steering efforts. Through constructing a Lyapunov-Krasovskii function, the asymptotic stability of the system is proved. Some essential conditions are derived to obtain the controller and triggered parameters. The hardware-in-the-loop (HIL) tests by Carsim/ Matlab joint platform are further used to validate the proposed shared assistance control strategy. The results illustrate the effectiveness to ensure the prescribed system performance while using fewer CAN network resources. Note to Practitioners—The driver assistance steering system plays an important role to enhance driving performance. Almost all road vehicles have been equipped with the Electrical Power Steering (EPS) function. Based on the information feedback from the onboard sensors, it can generate an extra steering input to assist the driver in real time. The advanced steering control technology, such as Servolectric of ZF and E•STEERTM of Delphi, can significantly reduce traffic accidents and improve the driving experience. Recently, the Original Equipment Manufacturers (OEMs) bring various advanced driver assistance systems to further guarantee the vehicle safety. This is usually accompanied by the communication load of the in-vehicle network CAN (Controller Area Network). Meanwhile, the x-by-wire technique can also induce the possibility of CAN delays. It would deteriorate the control performance. Hence, this paper introduces the event-triggered system to develop the steering assistance controller. The CAN network delays are also integrated into the human-machine shared steering model. The communication only occurs when the triggered condition of ETS is satisfied. A robust control method is further presented to ensure the system prescribed performance. This technology aims to be a paradigm to design active safety control units. Note that some poor driver states, such as fatigued driving and distracted driving have not been considered in this study for the design of the assistance controller. Future research will focus on developing a shared controller to accommodate time-varying driver characteristics.}, archive = {J_TASE}, author = {Jinhao Liang and Yanbo Lu and Faan Wang and Jiwei Feng and Dawei Pi and Guodong Yin and Yanjun Li}, doi = {10.1109/TASE.2024.3383094}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17501--17511}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ETS-based Human–Machine robust shared control design considering the network delays}, volume = {22}, year = {2025}, }
• (2025). Marden-based homotopic enclosed safe motion corridor generation for UAV navigation in complex environments. TASE, 22, 17486--17500. (www)

  @article{ , abstract = {This paper proposes a novel hierarchical methodology to planning safe UAV trajectories in complex environments. We start by improving a canonical hybrid A* in relation to high memory requirements, performance degradation, and the low efficiency customarily observed in the initial global trajectory suggested by the planner. Then, the Marden theorem is applied - for the first time in local path planning - to generate continuous, non-intersecting, enclosed, and safe flight corridors, termed homotopic enclosed safe motion corridors (HESMCs) hereafter. This is efficiently realized through a series of unique ellipsoids along the initial route. Meanwhile, the optimized motion trajectory along the corridors is built by considering two waypoints and prescribed performance functions. The resolved path is safe and complete, with a comprehensive Lyapunov stability analysis included to ensure accurate and efficient trajectory tracking. The simulation and physical tests demonstrate the superiority of our proposed planner over existing state-of-the-art methods, with consistent and significant improvements in processing time and guaranteed completeness. Note to Practitioners—The authors perceived the contribution of the manuscript of particular relevance to users of UAVs seeking advanced safety in their guidance and navigational solutions, offering a blend of theoretical innovation and practical applicability. The work introduces a distinct hierarchical motion planner specifically designed to enhance safety and reliability in UAV navigation. Key to this is the development of an improved hybrid A* algorithm for global planning, which effectively tackles practical issues such as high memory consumption and performance degradation. A significant theoretical contribution is the application of the Marden theorem in local optimization. This facilitates the generation of homotopic enclosed motion corridors using unique safe boundary ellipsoids, thus reducing navigation complexity and the risk of failure during task execution. Additionally, the proposed scheme emphasizes the generation of motion trajectories considering position errors and prescribed performance functions, supplemented by a thorough Lyapunov stability analysis. Looking ahead, we aim to extend the proposed scheme in the context of UAV swarms for more efficient navigation in complex environments.}, archive = {J_TASE}, author = {Chen Li and Xuelei Qi and Bao Chen and Shoudong Huang and Jaime Valls Miro and Hailong Huang and Wei Ni and Hongjun Ma}, doi = {10.1109/TASE.2024.3488692}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17486--17500}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Marden-based homotopic enclosed safe motion corridor generation for UAV navigation in complex environments}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based dynamic coverage control of multi-rotor UAVs with safety priority. TASE, 22, 17474--17485. (www)

  @article{ , abstract = {This paper considers the dynamic coverage control of multi-rotor Unmanned Aerial Vehicles (UAVs) with the limited sensory range, which aims to collect sensor information from all points of interest in the given task area until the desired prescribed level is reached. However, the unknown environments are usually unavoidable for coverage task, where the presence of various obstacles and communication interferences affect the flight safety and communication stability of UAVs. Therefore, collision avoidance and connectivity maintenance are considered as the two safety issues in this paper, in which connectivity maintenance ensures the communication environment for UAVs to collaboratively accomplish task, and collision avoidance is used for UAVs to avoid obstacles and neighbors. In order to realize dynamic coverage control with safety constraints based on local environment information, this paper proposes the reinforcement learning-based algorithm with shield, where the shield designed by discrete-time Control Barrier Function (CBF) not only ensures the safety of the UAVs in the learning and control phases, but also maximizes the coverage performance of UAVs. In addition, each UAV only relies on local information to generate safe actions for advancing the coverage process during the execution phase. Finally, the effectiveness of the algorithm is verified by numerical simulations and physical experiment. Note to Practitioners—A typical application scenario of dynamic coverage control is search and rescue (SAR), in which UAVs equipped with multiple sensors focus on monitoring areas where trapped people may be present, e.g., anomalous areas detected by infrared sensors due to human body temperature. Since SAR always occurs in unknown environments, it is crucial to ensure the safety of UAVs during missions, of which the safety issues considered in this paper include collision avoidance and connectivity maintenance. To perform the SAR mission safely, we construct the CBF-based shield, which minimizes corrections the exploration actions of the UAVs and ensures the safety of the cluster during the mission. In addition, UAVs are difficult to obtain global environmental information in unknown environments and only rely on their sensors to collect local information. Therefore, the reinforcement learning algorithm with shield proposed in this paper adopts the centralized training and decentralized execution strategy, where the UAVs only need local observation information to plan their next actions. Physical experiments were also conducted to validate the feasibility of implementing the proposed algorithm using real UAVs.}, archive = {J_TASE}, author = {Zhuangzhuang Ma and Junjie You and Yunlin Zhang and Yuhua Cheng and Jinliang Shao}, doi = {10.1109/TASE.2024.3420094}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17474--17485}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based dynamic coverage control of multi-rotor UAVs with safety priority}, volume = {22}, year = {2025}, }
• (2025). Joint communication and safe 3D path optimization for multi-UAV assisted mobile internet of vehicles on an uneven terrain. TASE, 22, 17464--17473. (www)

  @article{ , abstract = {The paper studies an Internet of Vehicles system consisting of several UAVs sending downlink information to ground vehicles to move an uneven terrain. In such environments, the transmission from the UAV to the ground vehicles may be blocked by tall buildings or mountains. Our goal is to construct multi-UAV navigation laws, transmission schedules, and transmit power levels of the UAVs that maximize the total downlink throughput of the system while minimizing the total transmit power and the propulsion energy of the multi-UAV team subject to a communication interference constraint for the UAVs. An important feature of the investigation is that a safety constraint that guarantees to avoid collisions between UAVs with a required safety margin should always be satisfied. A real-time law for joint communication and multi-UAV navigation is proposed. A mathematically rigorous analysis of the developed algorithm and proof of its optimality are given. Illustrative examples and simulations demonstrate the effectiveness of the approach. The main purpose of the proposed framework is to achieve joint communication and safe 3D path optimization, which means that the objective is to jointly construct 3D paths of the UAVs, transmission schedules, and transmit power levels of the UAVs. An advantage of this approach is that it allows to simultaneously maximize the total downlink throughput of the system and minimizes the total transmit power and the propulsion energy of all the UAVs subject to safety constraints that guarantee the UAVs avoid collisions between the UAVs as well as the UAVs and the uneven terrain. Note to Practitioners—Unmanned aerial vehicles (UAVs) have been regarded as a promising platform to serve as aerial access points, especially in complex environments with tall buildings and uneven terrains where conventional ground access points might usually provide unqualified service to users. In this paper, we have investigated an Internet of Vehicles (IoV) system that involves multiple unmanned aerial vehicles (UAVs) transmitting downlink information to ground vehicles navigating through challenging terrains. Our primary objective is to develop effective multi-UAV navigation laws, transmission schedules, and transmission power allocations that optimize the total downlink throughput of the system. At the same time, we aim to minimize the overall transmit power and propulsion energy of the UAV team, while adhering to communication interference constraints. It is worth noting that ensuring safety is of utmost importance in our investigation. We have incorporated a safety constraint that guarantees the prevention of collisions between UAVs, maintaining a required safety margin at all times. To tackle these challenges, we propose a real-time law that combines dynamic programming and model predictive control techniques for joint transmission and multi-UAV trajectory planning. The developed algorithm has undergone a rigorous mathematical analysis, and we have provided proof of its optimality.}, archive = {J_TASE}, author = {Andrey V. Savkin and Chao Huang}, doi = {10.1109/TASE.2024.3373007}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17464--17473}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Joint communication and safe 3D path optimization for multi-UAV assisted mobile internet of vehicles on an uneven terrain}, volume = {22}, year = {2025}, }
• (2025). Trajectory prediction for autonomous driving based on structural informer method. TASE, 22, 17452--17463. (www)

  @article{ , abstract = {Accurate and efficient prediction of the future trajectories of surrounding vehicles is of utmost importance in motion planning for autonomous driving. The ability to predict longer-term trajectories provides valuable information for effective motion planning. Numerous studies have contributed to the prediction of long-term vehicle trajectories. However, it is important to note that longer-term predictions can potentially lead to a trade-off between accuracy and computational complexity. In this work, we propose a structural Informer method, which can achieve accurate and efficient long-term trajectory prediction of the target vehicle. Specifically, the proposed method considers not only the temporal and spatial features of the interaction vehicle trajectory, but also the impact of vehicle state changes on the trajectory. To reduce computational redundancy and complexity while improving memory usage and prediction accuracy, the ProbSparse self-attention mechanisms and attention distillation operations are employed. The method is validated and evaluated using the NGSIM dataset, and the results demonstrate that the proposed structural Informer achieves satisfactory accuracy and time cost in long-term prediction of the TV compared with state-of-the art methods. Note to Practitioners—The motivation of this research is to address the impact of future trajectories of surrounding vehicles on the motion planning of autonomous vehicles. The method proposed applies advanced deep learning methods, and its strongpoint is that the proposed network can achieve higher efficiency and accuracy of trajectory prediction compared with state-of-the art methods. The specific implementation method is to use structured embedding methods and networks to extract more valuable features of the target vehicle, such as spatiotemporal features and vehicle state features. The novel attention mechanism is designed to solve the problem of exponential growth in computational complexity of traditional attention mechanisms in long-term prediction. The advancement of the method proposed confirmed by verification on naturalistic driving dataset.}, archive = {J_TASE}, author = {Chongpu Chen and Xinbo Chen and Chong Guo and Peng Hang}, doi = {10.1109/TASE.2023.3342978}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17452--17463}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory prediction for autonomous driving based on structural informer method}, volume = {22}, year = {2025}, }
• (2025). Mobile-DeepRFB: A lightweight terrain classifier for automatic mars rover navigation. TASE, 22, 17442--17451. (www)

  @article{ , abstract = {It requires terrain classification for unmanned Mars Rover to identify the safe areas. The current deep learning-based semantic segmentation and object recognition suffer from a large number of parameters and long training time. In this paper, a lightweight segmentation framework called Mobile-DeepRFB is proposed for the Martian terrain classification. It improves from the DeepLabV3+ by taking the MobileNetV3 as the backbone module to decrease the parameters and the Receptive Field Block (RFB) module to strengthen the feature extraction capability as well as to enlarge the receptive field. Experimental results on the NASA Mars terrain dataset AI4MARS show that the presented method reduces 94% about the parameter number and improves the mean pixel accuracy by 2% compared to the existing ResNet101 and Xception backbone networks. The deployment of this framework on a low-computing power embedded platform (NVIDIA Jetson Xavier) demonstrates its great potential to apply to Mars rovers. Note to Practitioners—This paper was motivated by the problem of terrain classification of planetary rovers. Existing methods are typically based on semantic segmentation technology to recognize various terrains while suffering from the drawback of a large number of parameters. We propose a lightweight segmentation framework to address this issue. In particular, the lightweight backbone network is applied to significantly reduce the number of parameters. The receptive field module is substantially improved to enhance the feature extraction capability. Eventually, we deploy the framework on a low-computing platform. Experimental tests show that the framework can significantly reduce the number of network parameters and it can be used for planetary rovers with limited computational resources.}, archive = {J_TASE}, author = {Lihang Feng and Sui Wang and Dong Wang and Pengwen Xiong and Jinjin Xie and Yong Hu and Miaomiao Zhang and Edmond Q. Wu and Aiguo Song}, doi = {10.1109/TASE.2023.3340190}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17442--17451}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mobile-DeepRFB: A lightweight terrain classifier for automatic mars rover navigation}, volume = {22}, year = {2025}, }
• (2025). MQKIN: Manufacturing quality knowledge-driven interpretable fault diagnosis network for robotic grinding equipment. TASE, 22, 17430--17441. (www)

  @article{ , abstract = {Despite of the fast development of deep learning networks, its inexplicability poses a low credibility challenge for fault diagnosis methods based on them. This article proposes an interpretable fault diagnosis network for robotic grinding equipment driven by the knowledge of grinding process. The network consists of a vibration imaging module, a grinding quality quantification module, an auxiliary learning module, and a fault diagnosis module. In addition, the developed synchronous algorithm is embedded in the network to update the weights and coefficient matrix combinations, which can reconstruct the vibration images from the wavelet domain to be recognized by convolution kernels more easily. Besides, through auxiliary learning, the grinding knowledge can be learned by the network to endow the results with interpretability. Finally, the comparative experiment shows that the proposed method has a maximum accuracy of 4.25%, 7.25%, 2.25%, 3.25%, and 1.25% higher than the five popular state-of-the-art fault diagnosis algorithms, respectively; the ablation experiment shows that the proposed algorithm has a maximum accuracy of 6.25%, 12.5%, and 16.25% higher than each sub-algorithm, respectively; by analyzing the learned weights of the network, it can be concluded that the proposed network has successfully learned the potential features of grinding knowledge with satisfactory performance.Note to Practitioners—In the autonomous robotic grinding system, the reliability of robotic grinding equipment is a prerequisite for ensuring high-quality grinding of thin-walled parts for large equipment such as aircraft, high-speed rail, and ships. In robotic grinding manufacturing, the information of grinding quality is the most direct feedback of valuable information from the manufacturing system. Based on the causal relationship between the fault status of grinding equipment and the grinding quality of workpiece, this article develops an interpretable fault diagnosis network incorporated with the grinding knowledge, which treats grinding quality information as physical knowledge and provides credibility to the fault diagnosis results. Therefore, this article improves the interpretability of the fault diagnosis model in practical applications through the credibility of data and grinding knowledge, which improves the applicable potential of fault diagnosis methods based on deep learning in industry.}, archive = {J_TASE}, author = {Wenbin He and Jianxu Mao and Yaonan Wang and Zhe Li and Xudong Wang and Shaoyuan Wang}, doi = {10.1109/TASE.2024.3418972}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17430--17441}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MQKIN: Manufacturing quality knowledge-driven interpretable fault diagnosis network for robotic grinding equipment}, volume = {22}, year = {2025}, }
• (2025). Knowledge acquisition method of urban rail transit safety event case base for intelligent emergency response. TASE, 22, 17421--17429. (www)

  @article{ , abstract = {Emergency-related knowledge from urban rail transit (URT) safety event case base is crucial for realizing intelligent emergency response in URT. However, safety event case records are non-standard and unstructured, leading to challenges in data mining, information induction, and knowledge organization. To obtain emergency-related knowledge effectively from URT safety event cases, this paper proposes a knowledge acquisition model incorporating label semantics and provides a knowledge acquisition degree evaluation method. In this study, the knowledge acquisition task is formulated as a machine reading comprehension (MRC) task. This formulation can take full advantage of the rich semantic information of labels, which can compensate for the drawbacks of the traditional sequence labeling model. The experiments conducted on the URT safety event case base demonstrate the effectiveness of the proposed methods. The knowledge acquisition model performs well on different labels and achieves a performance boost over baseline models. Note to Practitioners—Urban rail transit (URT) safety events are caused by various factors, such as natural disasters like earthquakes and floods, equipment failures, and terrorist attacks. et al. These events may disrupt the normal operation of URT, pose a great risk to personal safety, and result in property losses. URT safety event cases encompass essential emergency-related experience and knowledge, which serve as valuable references for managing new emergencies and mitigating potential losses. However, the record format of URT safety event cases is non-standard, with the majority stored as fragmented textual data, posing challenges in effectively extracting and utilizing the valuable knowledge they contain. Considering the data characteristics of URT safety event cases, three different methods are developed for knowledge acquisition. Among these methods, a knowledge acquisition model is proposed to extract emergency-related knowledge from URT safety event cases efficiently and accurately.}, archive = {J_TASE}, author = {Guangyu Zhu and Jiaxin Fan and Xinglin Huang and Nuo Zhang and Bo Wu and Yun Wei}, doi = {10.1109/TASE.2023.3337135}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17421--17429}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Knowledge acquisition method of urban rail transit safety event case base for intelligent emergency response}, volume = {22}, year = {2025}, }
• (2025). Sensor-fusion-based event-triggered following control for nonlinear autonomous vehicles under sensor attacks. TASE, 22, 17411--17420. (www)

  @article{ , abstract = {The situation of interest is where a vehicle is equipped with multiple sensors to measure the distance to the leading vehicle but does not need to obtain data from speed and acceleration sensors. The distance measurements are susceptible to asynchronous sampling and noise, nearly half of which may be manipulated by malicious attackers. In this situation, the event-triggered vehicle-following control problem of nonlinear autonomous vehicles with unknown parameters is studied. First, a secure event-triggered mechanism that can resist manipulation is devised to alleviate the burden of data transmission and processing caused by multiple sensors. Then, a novel adaptive sensor fusion algorithm is developed to estimate the actual distance. Subsequently, an improved adaptive observer is designed based on the event-triggered estimated distance to estimate continuous-time distance, velocity, acceleration, and system parameters. Finally, the following controller is designed using the estimated states and parameters with the help of Levant differentiators. The effectiveness of the proposed control scheme is validated through simulation studies. Note to Practitioners—This work aims to develop a secure following control method for nonlinear automated vehicles with unknown states and parameters, which can effectively handle sparse sensor problems caused by attacks, faults, saturation, etc. To address practical issues such as sampling intervals and limited computing and transmission resources, we propose a discrete sampling–event-triggered transmission–continuous estimation and control framework for continuous-time systems. Despite the presence of measurement interferences and potential corruption, the designed event-triggered mechanism and sensor fusion algorithm can be used to reduce data transmission and estimate the actual output, respectively. The designed adaptive observer can estimate continuous-time system states and parameters whether the output is obtained in a continuous, short-interval discrete or suitable event-triggered manner. This capability facilitates control design and real-time monitoring of vehicle states. Additionally, the presented backstepping control design and analysis method utilizing Levant differentiators can be applied in situations where the controlled system’s states possess at least first-order differentiability.}, archive = {J_TASE}, author = {Guangdeng Chen and Qi Zhou and Hongru Ren and Hongyi Li}, doi = {10.1109/TASE.2023.3337073}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17411--17420}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sensor-fusion-based event-triggered following control for nonlinear autonomous vehicles under sensor attacks}, volume = {22}, year = {2025}, }
• (2025). Bayesian inference of collision avoidance intent during ship encounters. TASE, 22, 17397--17410. (www)

  @article{ , abstract = {Ship collision accidents frequently cause casualties and significant property losses. These collisions mainly occur by incorrectly interpreting the intents of other ships’ navigators. However, inferring the avoidance intents of other ships is challenging, because the uncertain motions and the long-lasting dynamic interactions during encounters usually obscure the true intents. To address this, we propose a probabilistic graph model to infer the hidden avoidance intent of ships. Specifically, the dynamic encounter evolution is expressed in a model represented by a factor graph, where the intent beliefs are accumulated and propagated over time through message passing in the graph. In the inference procedure, we develop a hybrid Bayesian inference approach, integrating a data-driven component derived from empirical priors fitted to historical data, and a model-driven component capturing the ship’s control process. We derive 9 types of intents from naturalistic ship encounters and evaluate the model on a validation split. The quantitative metrics demonstrate that, on average, the proposed procedure can accurately infer the intents 14.04 seconds in advance and outperforms the baseline in macro-averaged recall rate (0.2919) and F1-score (0.2843). Note to Practitioners—This work aims to enhance safety in ship navigation by introducing an intelligent navigation support system to prevent collisions. A significant cause of ship collision accidents is the misinterpretation of other navigators’ intents. Estimating the intents of marine traffic participants can significantly reduce collision rates. However, real-time intent inference in ship encounters poses challenges due to complex dynamics and uncertain motions. We propose a novel probabilistic graphical model that allows us to infer avoidance intents, evaluated on a naturalistic Automatic Identification System (AIS) dataset, demonstrating effectiveness in real-world scenarios. The practical implications of our work are substantial for the maritime industry. Ship captains and intelligent automated vessels can use our system to enhance situational awareness, enabling informed collision avoidance decisions. Our work aims to improve the safety and efficiency of ship navigation, reducing collision risks and ensuring the well-being of crew members and valuable cargo.}, archive = {J_TASE}, author = {Chengfeng Jia and Jie Ma and Bert De Vries and Wouter M. Kouw}, doi = {10.1109/TASE.2024.3383819}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17397--17410}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bayesian inference of collision avoidance intent during ship encounters}, volume = {22}, year = {2025}, }
• (2025). Safer conflict-based search: Risk-constrained optimal pathfinding for multiple connected and automated vehicles. TASE, 22, 17382--17396. (www)

  @article{ , abstract = {Coordinating connected and automated vehicles (CAVs) poses challenges in achieving safe and optimal pathfinding simultaneously. In this paper, we propose safer conflict-based search (Safer-CBS), an approach based on conflict-based search that generates risk-constrained optimal paths for multiple CAVs. Safer-CBS addresses these challenges by incorporating risk assessment and optimization techniques. By defining the conflict as the risk between two vehicles exceeding a predetermined threshold, the resultant conflict-free paths can provide safety guarantees. An optimization over risk constraints is devised to enhance the efficiency of the tree search. Additional improvements are also incorporated to expedite the search process. Theoretical analysis demonstrates the optimality of Safer-CBS in finding risk-constrained paths. Furthermore, Safer-CBS can seamlessly integrate trajectory planning to generate risk-constrained trajectories for CAVs. Numerical calculations on grid maps validate the effectiveness of the proposed optimization and additional improvements, resulting in increased success rates (by over 43% for 10 vehicles) and reduced runtime (by over 59%) when compared to the method without the optimization and improvements for Safer-CBS. Simulations in both unstructured and structured environments confirm that Safer-CBS, combined with trajectory planning, ensures that trajectory risks strictly remain below predetermined thresholds, while achieving a superior balance between safety and efficiency compared to the benchmarks. Note to Practitioners—Applying multi-CAV technologies in areas such as warehousing, transportation, and operations typically requires the planning of cooperative paths or trajectories for the CAVs. In this planning process, safety and efficiency issues are centrally considered and balanced. Cooperative performance can be ensured as long as each individual vehicle adheres to the central instructions. This study develops a method for risk-constrained cooperative planning, aiming to generate optimal paths and suboptimal trajectories for multiple CAVs. The output trajectories consist of path points with fixed time intervals, which the vehicle controllers can track accordingly. By maintaining the risk level among the vehicles below predetermined thresholds, overall safety can be achieved. Meanwhile, the planning process is optimized, thus improving overall efficiency.}, archive = {J_TASE}, author = {Yicong Liu and Heye Huang and Qing Xu and Shaobing Xu and Jianqiang Wang}, doi = {10.1109/TASE.2024.3383421}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17382--17396}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safer conflict-based search: Risk-constrained optimal pathfinding for multiple connected and automated vehicles}, volume = {22}, year = {2025}, }
• (2025). Contrastive preference-guided active learning approach based on ranking correlation for real-time safety assessment. TASE, 22, 17370--17381. (www)

  @article{ , abstract = {Real-time safety assessment is a critical process for identifying and analyzing potential safety hazards in industrial applications. Active learning has been widely recognized as an effective technique for addressing such issues by utilizing small labeled samples to achieve high evaluation performance. However, in real-world scenarios, the performance of query strategies can be significantly impacted by the data environment. Therefore, it is crucial to co-design a strategy using different criteria to ensure reliable and stable safety assessment results for system operation. In this paper, we propose a contrastive preference-guided active learning approach to tackle chunk-level real-time safety assessment tasks in non-stationary environments. Firstly, we construct negative ranking lists and random lists. Then, we introduce Jeffrey divergence to measure pairwise ranking correlation. By leveraging contrastive preference relationships, we can effectively obtain the value of samples in data chunks with preference aggregation procedures. To verify the effectiveness of the proposed method, we conduct numerous experiments using realistic data from the JiaoLong deep-sea manned submersible. The results demonstrate that our approach outperforms most existing advanced methods in terms of performance stability and accuracy.Note to Practitioners—In situations where a significant number of samples must be processed concurrently to promptly identify and mitigate potential hazards, it is vital to consider the responsibilities of real-time safety assessment (RTSA). The proposed approach offers a resolution to the difficulty of acquiring annotations for all samples due to the continual stream of data. Multiple query criteria can be seamlessly integrated, which enhances the stability and superiority of learning performance simultaneously. The approach’s scalability, flexibility, and effectiveness render it a valuable instrument for guaranteeing a prompt response to emergent procedural threats and safety. It can accommodate any number and type of advanced single query criteria, making it easy to adapt to various scenarios. The proposed CPRC approach signifies a significant breakthrough in the field of real-time safety assessment, with the potential to enhance safety outcomes in industrial settings. Multiple experiments involving the realistic JiaoLong deep-sea manned submersible were conducted, and the outcomes demonstrate the benefits of this approach for practical RTSA applications.}, archive = {J_TASE}, author = {Zeyi Liu and Xiao He}, doi = {10.1109/TASE.2024.3401470}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17370--17381}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Contrastive preference-guided active learning approach based on ranking correlation for real-time safety assessment}, volume = {22}, year = {2025}, }
• (2025). When is it likely to fail? performance monitor for black-box trajectory prediction model. TASE, 22, 17355--17369. (www)

  @article{ , abstract = {Accurate trajectory prediction is vital for various applications, including autonomous vehicles. However, the complexity and limited transparency of many prediction algorithms often result in black-box models, making it challenging to understand their limitations and anticipate potential failures. This further raises potential risks for systems based on these prediction models. This study introduces the performance monitor for black-box trajectory prediction model (PMBP) to address this challenge. The PMBP estimates the performance of black-box trajectory prediction models online, enabling informed decision-making. The study explores various methods’ applicability to the PMBP, including anomaly detection, machine learning, deep learning, and ensemble, with specific monitors designed for each method to provide online output representing prediction performance. Comprehensive experiments validate the PMBP’s effectiveness, comparing different monitoring methods. Results show that the PMBP effectively achieves promising monitoring performance, particularly excelling in deep learning-based monitoring. It achieves improvement scores of 0.81 and 0.79 for average prediction error and final prediction error monitoring, respectively, outperforming previous white-box and gray-box methods. Furthermore, the PMBP’s applicability is validated on different datasets and prediction models, while ablation studies confirm the effectiveness of the proposed mechanism. Hybrid prediction and autonomous driving planning experiments further show the PMBP’s value from an application perspective. Project page: https://swb19.github.io/PMBP/. Note to Practitioners—This research presents PMBP, a valuable tool for practitioners in the automation industry. The PMBP enables online monitoring of black-box trajectory prediction models, enhancing system reliability and facilitating informed decision-making. The practical application of PMBP lies in improving safety and reliability in critical domains, especially in the context of autonomous vehicles. Black-box trajectory prediction models commonly used in these domains may exhibit unexpected deficiencies, potentially leading to risks. By monitoring the prediction performance online, systems can proactively identify potential insufficiencies and make informed decisions to ensure safer and more reliable operations. The PMBP offers practitioners different monitoring solutions based on various approaches, addressing their specific needs effectively. While the PMBP has shown promising outcomes, further exploration and testing are necessary to fully harness and apply its monitoring results in automated systems. Practitioners are encouraged to adopt the PMBP as an essential monitoring mechanism to enhance the reliability of their trajectory prediction models and achieve safer and more efficient automation in their domains.}, archive = {J_TASE}, author = {Wenbo Shao and Boqi Li and Wenhao Yu and Jiahui Xu and Hong Wang}, doi = {10.1109/TASE.2024.3356070}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17355--17369}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {When is it likely to fail? performance monitor for black-box trajectory prediction model}, volume = {22}, year = {2025}, }
• (2025). Resilient MPC under severe attacks on both forward and feedback communication channels. TASE, 22, 17341--17354. (www)

  @article{ , abstract = {This paper proposes a resilient model predictive control (MPC) strategy for constrained cyber-physical systems (CPSs) subject to disturbances and cyber attacks. The feedback sensor-controller (S-C) channel suffers from replay attack and the forward controller-actuator (C-A) channel suffers from false data injection (FDI) attack, and the defender has no prior information about the intruder. Considering that the abnormal behavior of intruder cannot be predicted, an expected one-step controllable set, and a series of minimally conservative constraints are developed to build attack detector. Two controllers are designed jointly based on infinite-horizon MPC to mitigate the negative effects caused by attack. Compared with the existing resilient control strategies, the attack model considered in this paper is less conservative, the resilient control structure is simpler, and it can avoid continuous channel refreshing caused by close-range attack. Robust constraint satisfaction, recursive feasibility and uniformly ultimate boundedness (UUB) are ensured for any admissible attack scenario and disturbance realization. Finally, simulations on a supply chain model show the efficacy of the algorithm. Note to Practitioners—With the wide application of wireless networks, vulnerabilities in the communication process can be easily exploited by intruders to launch malicious attacks. Resilient control can provide acceptable robustness and improve system safety. Additionally, many practical systems are constrained and disturbed, and MPC is one of the most effective methods for dealing with control problems in such systems. Its rolling optimization characteristics make it robust to disturbances. On the one hand, both the C-A channel and the S-C channel of the networked control system are vulnerable to attack. On the other hand, in a complex environment, the defender may not have prior information about the attacker, such as attack probability. Therefore, it is not practical to assume that only a single channel is attacked or that the algorithm relies on the attacker’s prior information. In this paper, a resilient MPC control structure is proposed to counter two types of attacks: replay attack and false data injection attack. One of its characteristics is that the attack model’s conservativeness is relatively low, and the structure is relatively simple, making the algorithm more applicable to actual needs. Another feature of the proposed approach is that it can be adapted directly to different types of attacks without modifying the overall resilient MPC architecture.}, archive = {J_TASE}, author = {Huan Yang and Li Dai and Huahui Xie and Yang Shi and Yuanqing Xia}, doi = {10.1109/TASE.2023.3323806}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17341--17354}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient MPC under severe attacks on both forward and feedback communication channels}, volume = {22}, year = {2025}, }
• (2025). Adaptive neural network constrained fault tolerant control for nonlinear systems with actuator failures and saturation. TASE, 22, 17333--17340. (www)

  @article{ , abstract = {This article studies a neural network (NN)-based adaptive fault-tolerant control (FTC) scheme to solve the actuator failure problem of nonlinear systems with input saturation and time-varying state constraints. The introduction of asymmetric Barrier-Lyapunov function (BLF) makes controller design more difficult due to the occurrence of actuator saturation and failure. Therefore, the method proposed in this article proposes a saturation fault-tolerant controller with constraint compensation information under the backstepping control design framework to solve the state constraint asymmetry problem. By design-improved asymmetric time to change the BLF, the design of the state constraint controller will become more realistic and constraints will be weakened. Actuator failure in this article considers deviation failure and loss of effectiveness. Based on the nature of the index function, the improved BLF can make the boundaries of state constraints smaller and smaller, and the boundaries of constraints can change with the expected trajectory when satisfying the input saturation. Therefore, this article solves the FTC problem of non -linear strict feedback systems, overcoming the impact of non -symmetrical state constraints and input saturation on system performance. Simulation verified the feasibility of this control method. Note to Practitioners—This work implements asymmetrically constrained fault-tolerant control under actuator sensor failure for a nonlinear system with input saturation. Input saturation and state-constrained control asymmetries caused by actuator failures are still rarely studied. This article proposes an improved BLF to solve the asymmetry problem under actuator failure. A fault-tolerant control method is designed considering multivariable conditions and control input constraints, which covers information on actuator faults and amplitude saturation. Based on the properties of the exponential function, the improved BLF can make the bounds of the state constraints smaller and smaller, and the bounds of the constraints can change with the desired trajectory while satisfying input saturation. The proposed method has a general structure and has the potential to be implemented in real systems.}, archive = {J_TASE}, author = {Lei Ma and Zhanshan Wang and Huiting Zhang and Qiufu Wang}, doi = {10.1109/TASE.2024.3381780}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17333--17340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neural network constrained fault tolerant control for nonlinear systems with actuator failures and saturation}, volume = {22}, year = {2025}, }
• (2025). A PPB-SIADP optimal fault-tolerant attitude control scheme for tailless flying wing aircraft with actuator faults and saturation. TASE, 22, 17320--17332. (www)

  @article{ , abstract = {This paper addresses the problem of optimal fault-tolerant attitude control with prescribed performance for a tailless flying wing aircraft under actuator faults and saturation. Firstly, by using the time-scale separation principle, the considered state variables of the aircraft are divided into two groups, the angular rate loop and the attitude loop, with different response speeds. Secondly, after prescribed performance bound (PPB) transformation, a novel error tracking nominal system is obtained, based on which an incremental neural network observer is designed to approximate the actuator faults. Thirdly, a functional performance index in terms of the actuator saturation is designed for the nominal system and the corresponding Hamilton-Jacobi-Bellman (HJB) equation is derived. A new single-network incremental adaptive dynamic programming (SIADP) algorithm is proposed to solve the optimal control problem. Moreover, the uniformly ultimately bounded stability of the closed-loop incremental system is proved by using the Lyapunov theory. Finally, simulations are given to illustrate the effectiveness of the proposed control approach. Note to Practitioners—The aim of this paper is to develop a new control scheme to realize active fault-tolerant control for tailless flying wing aircraft with actuator faults and saturation. The absence of vertical and horizontal tail fins results in a short longitudinal control moment arm, insufficient longitudinal control efficiency, deep coupling between longitudinal and lateral dynamics and poor stability of the tailless flying wing aircraft. Fault tolerance ability is necessary in flight control design, and the optimal and transient performance with control surface saturation should be considered in control practice. To solve these problems, a PPB-SIADP scheme combined with incremental adaptive observer approach is proposed, the tracking error can converge in an optimal way with prescribed performance. The proposed approach in this paper can be used as a valuable scheme for the tailless flying wing aircraft control law design in engineering practice.}, archive = {J_TASE}, author = {Shaojie Zhang and Maomao Zhao and Qingkai Meng and Bin Jiang and Kun Ji and Zhimu Yang}, doi = {10.1109/TASE.2024.3363838}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17320--17332}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A PPB-SIADP optimal fault-tolerant attitude control scheme for tailless flying wing aircraft with actuator faults and saturation}, volume = {22}, year = {2025}, }
• (2025). Secondary multi-bus voltage and frequency fault-tolerant regulation control via hierarchical mechanism. TASE, 22, 17310--17319. (www)

  @article{ , abstract = {A multi-bus voltage regulation strategy based on distributed fault-tolerant control is proposed in a microgird (MG). The strategy addresses the challenges posed by the presence of actuator partial loss of effectiveness (PLOE) and deviation faults in multiple distributed generators (DGs) communicating through a directed network. The concept of containment control is introduced in the multi-bus voltage regulation strategy. In this paper, the bus voltages are limited to a reasonable range while still allowing voltage differences between buses for efficient power flow. In addition, a hierarchical mechanism is proposed to divide the voltage fault-tolerant containment control problem into a cooperative control problem for virtual systems and a decentralized control problem for real systems. Notably, the mechanism avoids relying on global information about the directed communication network and specific quantitative details about the magnitude, frequency, and type of actuator failures. In addition, we devise a consensus-based control strategy to regulate the frequency and efficiently share active power among DGs. The strategy is thoroughly evaluated through comprehensive simulation experiments, demonstrating its effectiveness and performance in voltage regulation and fault management scenarios. Note to Practitioners—This paper aims to propose a fault-tolerant secondary control strategy for bus voltages and frequency. First, considering that the power flow between buses is based on voltage differences, the problem of regulating the bus voltages is transformed into a containment control problem for linear multi-agent system (MAS) containing unknown nonlinear dynamics. Next, we design a fault-tolerant secondary control strategy with hierarchical control. At the network layer, a fault-free virtual system with adjacency information is created and approximated with a neural network for nonlinear dynamics. In the physical layer, a decentralized adaptive tracking controller is designed to synchronize the virtual system with the existing system. This layering process decouples time-varying PLOE fault parameters and adaptive control gains under an asymmetric network, thus enabling fault-tolerant control of actuator PLOE faults and deviation faults in a fully distributed framework. In addition, the boundaries of the fault parameters can be unknown. A feasible strategy is provided for industrial applications.}, archive = {J_TASE}, author = {Mei-Na Zhai and Jiayue Sun}, doi = {10.1109/TASE.2024.3357773}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17310--17319}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secondary multi-bus voltage and frequency fault-tolerant regulation control via hierarchical mechanism}, volume = {22}, year = {2025}, }
• (2025). MAR and UAR: Solutions for depth profile reconstruction of MFL inspection with high degree of freedoms. TASE, 22, 17299--17309. (www)

  @article{ , abstract = {Magnetic Flux Leakage(MFL) inspection is one of the most efficient non-destructive testing method used to detect anormalies of ferromagnetic materials which is widely used in many industry scenarios. Sizing of the defect from inspection signal is the key problem which is also a classic inverse problem. Reconstruction of corrosion liked defect with high degree of freedom(DOF) is a challenging problem in this area. A multi-agent based reconstruction algorithm(MAR) is proposed in this paper. Instead of designing an algorithm to build the entire defect, the defect region is divided into several segments. Each segment is reconstructed by its corresponding agent. The dimension of reward function is reduced number-of-agent times from the DOF of defect which makes the critic network efficient. Base on MAR, an uni-agent base reconstruction algorithm(UAR) is also given as another solution which requires less computing resources. How to divide the defect and corresponding signal to achieve better performance is also studied. Algorithms proposed in this paper solves the problem of inverse problem of MFL inspection with high DOF. The effectiveness of the algorithms proposed in this paper are validated with simulated data and practical inspection data. The results show that the algorithms proposed in this paper has good reconstruction accuracy with robustness when dealing with defects with high DOF. Note to Practitioners—MFL inspection is the widely used pipeline in-line inspection method. Currently, defects detected are usually quantified as simple cube defects. Since most of the practical defects on industrial pipelines are caused by corrosion which have complex shape, how to quantify defects with more details is not only a research topic but also practical demand. In this paper, MAR and UAR are proposed to reconstruct the depth with 99 degree of freedoms. In addition, the algorithms proposed can be extended to reconstruct the depth with much more higher degree of freedoms by just adopting more agents. The impact of wings data on reconstruction results is also discussed in this paper. The algorithms proposed in this paper focus on 2-D dimension which considers only the length and depth reconstruction problem. In further research, we will try to give solutions to 3-D dimension problem which considers length, width and depth in the same time.}, archive = {J_TASE}, author = {Zhenning Wu and Jinhai Liu and Lixing Wang}, doi = {10.1109/TASE.2023.3326496}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17299--17309}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MAR and UAR: Solutions for depth profile reconstruction of MFL inspection with high degree of freedoms}, volume = {22}, year = {2025}, }
• (2025). Aperiodic sampling artificial-actual h∞ optimal control for interconnected constrained systems. TASE, 22, 17288--17298. (www)

  @article{ , abstract = {This paper analyzes decentralized/centralized event/ self-triggered schemes respectively according to a large-scale interconnected system with state limitation and uncertain interference, and devises a set of parallel controllers for the control input and the worst disturbance. Firstly, the system with asymmetric state constraints is transformed into an auxiliary system by means of coordinate transformation. And then, the physical and artificial systems are integrated to establish a parallel control structure to improve the performance of the system. Next, an aperiodic sampling robust $H_{\infty }$ optimization method to save communication resources is proposed through zero-sum game (ZSG) theory, and the control pair sequence is obtained by solving the event-triggered Hamilton-Jacobi-Isaacs equation (ET-HJIE) via heuristic dynamic programming (HDP). Based on decentralized ET control (DETC), centralized ETC (CETC), decentralized self-triggered control (DSTC) and centralized STC (CSTC) are further proposed. In addition, the stability of the system under these four modes is proved, and the Zeno phenomenon is discussed. In the end, a simulation model is used to verify the control sequences within four trigger schemes, and satisfactory results are obtained. Note to Practitioners—Inspired by the large scale, interrelated terms, unavoidable disturbances, and nonlinear constraints of practical engineering systems, an aperiodic sampling HDP algorithm based on artificial-actual interaction is presented to handle with $H_{\infty }$ optimization of nonlinear interconnected systems with state constraints. The four kinds of aperiodic sampling control methods including DETC, CETC, DSTC and CSTC are proposed in turn, which can greatly reduce data transmission while ensuring the control performance. Moreover, the advantages and characteristics of the four trigger mechanisms are summarized based on simulation, which has certain reference significance for physical systems with different control requirements.}, archive = {J_TASE}, author = {Ruizhuo Song and Lu Liu and Bo Hu}, doi = {10.1109/TASE.2023.3324643}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17288--17298}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Aperiodic sampling artificial-actual h∞ optimal control for interconnected constrained systems}, volume = {22}, year = {2025}, }
• (2025). Fault diagnosis for energy transportation-oriented intelligent automation system via mixed neural networks. TASE, 22, 17278--17287. (www)

  @article{ , abstract = {A leak fault is a safety risk to interrupt the operation of the transportation process for intelligent automation system. To judge the system failure, a fault diagnosis method based on mixed neural networks is proposed in this paper. First, the multi-attribute extraction module is proposed to provide the local and global data changes caused by the leak event for the following neural networks, which reduces the noise influence from the complex environment. Second, a feature-sharing neural network is proposed to capture the multi-scale features of data changes and further implement leak fault detection and location through the sequence information and the corresponding sequential dependencies. Third, the task-dependency loss function is proposed to update the network parameters of the joint learning of different network outputs in the whole training process. Based on the proposed network structure, the proposed method could achieve two different targets of fault detection and location simultaneously for long-time series data. Finally, different case studies of the collected acoustic data are studied, and the analysis results show the effectiveness of the proposed method for leak fault detection. Note to Practitioners—A leak event is considered a sudden system fault in the transportation-oriented intelligent automation system. In order to reduce and minimize the system damage influence, leak fault diagnosis is the key point to identify operation risks based on the analysis result of the collected data changes. Thus, a mixed neural networks-based method is proposed to achieve the detection and location of leak events in this paper. With the combination of different neural network structures, multi-scale data changes could enhance the analysis capability of the proposed method for different fault diagnosis targets in complex operation scenarios. The case results show that the proposed method is an implementation way to ensure system safety and is better than other detection methods through the comparisons of different evaluation metrics.}, archive = {J_TASE}, author = {Chengze Ren and Xuguang Hu and Qiuye Sun and Jigui Zhang}, doi = {10.1109/TASE.2024.3383852}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17278--17287}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault diagnosis for energy transportation-oriented intelligent automation system via mixed neural networks}, volume = {22}, year = {2025}, }
• (2025). Controller design for leader-follower systems with hidden markovian jamming attack. TASE, 22, 17266--17277. (www)

  @article{ , abstract = {Transportation systems often use agent-to-agent communication technologies to enhance control performance by transmitting information across wireless networks to keep a reasonable inter-distance between agents. The agent-to-agent data transmission is influenced by the fading channel and the limited communication bandwidth. Denial of Service attack model is used to describe the attack’s influence on control system. Thus, the communication model is built as Bernoulli distribution, Markovian distribution and so on. But in practical situations, it is more complicated than we imagine. To further push the theoretical results to practical situations, we investigate the jamming attack with non-stationary stochastic process in the leader-follower systems and introduce hidden Markovian distribution model (HMM) to describe this special jamming attack. Furthermore, the controller design method is presented to achieve stochastic stability of the leader-follower system. The structure of the control algorithm is a special MPC method, which is divided into a state feedback part and a modification part. The state feedback part is a typical LMIs controller to guarantee the system’s stability, while the modification part is used to improve the influence of the model mismatch and uncertainty. The highlights of the results are the following two points. Firstly, the more complex and precise transportation system model is founded by the hidden Markovian distribution. Secondly, the controller design method is also presented for the multi-agent systems with this HMM transportation system model. The existing results always find the worst case scenario, and design the controller in this scenario. Usually, the controller is very conservative and even unstable. The results in this paper can change the controller in different scenarios to adapt to the corresponding jamming stochastic distribution model, which is the novelty of this paper. We use two examples to illustrate the proposed results. The first one is used to show the ineffectiveness of the results by the simple stochastic distribution model method and the effectiveness of the proposed results in this paper. The second one is used to show that the method can be used in a 3-vehicle-agent system. Note to Practitioners—The paper’s results have some potential applications in logistics industry, automated factory, and flexible automated manufacture systems. In these scenarios, the feed flow mobile robots are widely used to exchange a mass of manufacturing materials. On one hand, usually, single vehicle control theory does not have the group intelligence ability. Its actions only consider the safety and efficiency itself, which may not be the best actions for the whole system. In this paper, we only consider that the vehicles form a platooning, which is the most time of working conditions for mobile robots. On the other hand, usually, the network data exchange environment in manufacturing industry is complex, which is main reason to cause instability of the networked platooning system and also the bottleneck limiting the network data into the closed loop of the control system. In this paper, the new transmission interference model (HMM) is built, whose parameters can be confirmed by the algorithm and the observation data. Based on this new model, the corresponding unique controller is also designed accordingly. The challenge of this method is to find a method that can suppress large disturbances to the maximum extent and maintain system stability. But if the big interference lasts long enough, it means that the system lasts long enough in the open-loop condition, and the system eventually becomes unstable, which is the limitation of our research in theory.}, archive = {J_TASE}, author = {Zhicheng Li and Ming Li and Yankun Wang and Yang Wang}, doi = {10.1109/TASE.2024.3375339}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17266--17277}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Controller design for leader-follower systems with hidden markovian jamming attack}, volume = {22}, year = {2025}, }
• (2025). Safety-critical automated surface vessels MIMO control with adaptive control barrier functions under model uncertainties. TASE, 22, 17253--17265. (www)

  @article{ , abstract = {Ensuring the side-by-side configuration of the automated surface vessels under the effect of the uncertainties of the environment attracted continued efforts on this challenging issue. This paper investigates the learning-based safety-enhanced adaptive side-by-side control algorithm that suitably defines and uses the modified adaptive control barrier functions to tackle the constrained control satisfied under the uncertainty environment by ensuring the control barriers on state-variables constraints. In addition to the inherent adaptivity of the adaptive control barrier functions, the finite-time auxiliary system is enabled to modify the adaptive control barrier functions, which considers the unknown part of the nominal model to ensure the satisfaction of system constraints, especially under uncertain situations. For the formulated quadratic programs with adaptive control barrier functions and control Lyapunov functions, the operator splitting quadratic program is employed in problem-solving, which effectively enlarges the robustness of problem-solving, making it particularly efficient and reliable for real-time applications. The effectiveness of the proposed method is demonstrated via comparative simulations under uncertain cases to show the superior adaptive ability under the model uncertainties, which can be employed in marine industry applications, e.g., carbon emissions estimation modeling and optimization. Note to Practitioners—This paper was motivated by enlarging the adaptive ability of optimization-based safe operation for safety-critical automated surface vessels control under the side-by-side configuration under the effect of the uncertainty of the environment. The proposed learning-based safety-enhanced adaptive side-by-side control algorithm promotes adaptivity and control performance with the usage of the finite-time auxiliary system to the refined adaptive control barrier functions in the optimization problem formulation. The problem-solving is accelerated with the usage of the operator-splitting quadratic program. These implements effectively enlarge the robustness of the proposed method, making it particularly efficient and reliable for industrial real-time applications. The proposed method employed a complicated system design with the outcome of a simple algorithm implemented with promoted adaptivity and robustness, which can be employed in industrial applications in future research to promote current control system performance.}, archive = {J_TASE}, author = {Yuxiang Zhang and Shuzhi Sam Ge and Xiaoling Liang and Bernard Voon Ee How and Hong Chen}, doi = {10.1109/TASE.2024.3355564}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17253--17265}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safety-critical automated surface vessels MIMO control with adaptive control barrier functions under model uncertainties}, volume = {22}, year = {2025}, }
• (2025). Stability analysis and control validation of DDEV in handling limit via SOSP: A strategy based on stability region. TASE, 22, 17240--17252. (www)

  @article{ , abstract = {The stability region is an important criterion in the active safety system of vehicle. Extensive literatures have developed various stability regions for the centralized driving vehicle (CDV). However, seldom of them make a thorough analysis of the stability for distributed driving electric vehicle (DDEV). Especially when the direct yaw moment control (DYC) intervenes, the stability region of DDEV shows a huge difference compared with CDV. So far, most researches on the stability control of DDEV are still based on the traditional CDV stability region, which leads to the conservation of controller design. To this end, a dynamic and analytical stability region of DDEV is firstly developed in this paper. By employing sum of square programming (SOSP) algorithm, we choose a high-order Lyapunov function to make a precise estimation of the stability region. A novel DC shape function is proposed to reduce the conservation of estimation when DYC involves. To ensure the real-time performance for control application, Long Short Term Memory (LSTM) neural network is employed to fit the coefficients of Lyapunov function, as well as enable the dynamic shifting with driving conditions. Based on the aforementioned stability region, we develop a MPC controller to ensure the stability and tracking performance during handling limit. Both simulations and road tests demonstrate that the developed stability region could effectively restraint the vehicle states from diverging, which enhances vehicle maneuverability while ensuring vehicle stability. Note to Practitioners—Handling stability is a crucial factor which concerns the safety of vehicle. Almost all vehicles should be equipped with active safety system, such as Electronic Stability Program (ESP) of Bosch and Electronic Stability Controller (ESC) of GM. They collect the real-time vehicle states by the Built-in sensors. Meanwhile, they calculate the current stability margin according to the driver’s control input and analyze the deviation of the vehicle states to decide whether the active control is needed to maintain the stability of vehicle. This system is established well in CDV. However, for DDEV, it has another input of DYC in addition to wheel steering angle, which changes the dynamic characteristic of vehicle. This means that the stability margin of traditional ESP is inaccurate when the DYC intervenes, which will lead to the misjudgment of instability or the unsafety of vehicle. This paper develops an analytic stability region of DDEV and describes the stability margin with different values of DYC, which could adapt to varying conditions. This technology fills the gap of the stability determination of DDEV and has a good application prospect in the active safety system of DDEV.}, archive = {J_TASE}, author = {Tong Shen and Fanxun Wang and Yongjun Yan and Mingzhuo Zhao and Yanjun Ren and Jinhao Liang and Guodong Yin}, doi = {10.1109/TASE.2023.3333393}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17240--17252}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stability analysis and control validation of DDEV in handling limit via SOSP: A strategy based on stability region}, volume = {22}, year = {2025}, }
• (2025). Data-driven injection attack against discrete-time intelligent automation systems with slowly time-varying delays. TASE, 22, 17227--17239. (www)

  @article{ , abstract = {This paper addresses data-driven injection attack against unknown intelligent automation systems (IASs) with slowly time-varying delays, which is a more general but also more challenging to deal with than the model-based real-time systems. Using the control input and system output measurements, several new data-driven injection attack strategies based on compact form dynamic linearization (CFDL) and incremental triangular dynamic linearization (ITDL) are proposed. The attack strategies are more general than the existing ones, taking into account the unknown model parameters and time-varying delays in control-to-actuator as well as sensor-to-controller data transmission channels. Consequently, the new design attack results are anticipated to have wider applicability. Based on the established attack models of CFDL and ITDL, the data-driven optimal parameter estimation algorithms are employed to overcome the difficulty of the unknown model. Furthermore, with the help of the principle of the linear regression equation, the problem of seeking partial derivatives for the attack inputs with time delays is avoided. Several examples are presented to illustrate the validity of the designed attack strategies. Note to Practitioners—The primary objective of this paper is to focus on the cyber security of discrete-time intelligent automation systems from the viewpoint of the attacker, which provides insight into the way of generating attack strategies under the data-driven framework. The majority of the available attack strategies against intelligent automation systems are based on a priori knowledge of the system, without delays or with single-sided fixed time delay, and thus in reality the attack strategies fail or are mostly ineffective due to inaccessibility of the attacked system parameters in conjunction with communication network time delays. This paper synthesizes parameter estimation, dynamic linearization, optimality principle and control theory to propose data-driven injection attack strategies based on CFDL and ITDL, respectively. Compared with the CFDL-based attack strategy, the latter is more applicable in the situation where the attacker suffers from restricted memory space occupation and hashrate. Specifically, the operation of attacker has two phases: eavesdropping estimation and strategy generation. In the eavesdropping estimation phase, the attacker continuously eavesdrops and stores the traffic data of the system for parameter estimation as well as optimizes the estimated parameters. In the strategy generation phase, the attacker generates strategies with the assistance of optimal parameters. Mathematically analytical forms of the proposed two data-driven injection attack strategies are presented in detail. Then, their effectiveness is verified through numerical simulation and experiments based on a leader-following form consisting of five quadcopters, but not tested in production. In the future, we are intending to design the data-driven attack strategies for heterogeneous intelligent automation systems with time-varying delays under the framework of game.}, archive = {J_TASE}, author = {Sheng Gao and Hao Zhang and Zhuping Wang and Haohan Huang and Huaicheng Yan}, doi = {10.1109/TASE.2023.3319828}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17227--17239}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven injection attack against discrete-time intelligent automation systems with slowly time-varying delays}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Emerging trends in safety-critical issues for intelligent automation systems. TASE, 22, 17221--17226. (www)

  @article{ , archive = {J_TASE}, author = {Chao Huang and Qinglai Wei and Huaguang Zhang and Andrey Savkine and Mohammed Chadli and Hailong Huang and Gaurav Pandey}, doi = {10.1109/TASE.2025.3575215}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17221--17226}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Emerging trends in safety-critical issues for intelligent automation systems}, volume = {22}, year = {2025}, }
• (2025). Performance-dependent privacy preserving for switched multiple systems under active disturbance rejection control. TASE, 22, 17209--17220. (www)

  @article{ , abstract = {Disturbances and uncertainties affect the performance of control systems and must be addressed. In addition, due to the potential privacy leakage caused by the informatization and networking of control systems, ensuring system privacy has become an important technical challenge. The topic of this paper is active disturbance rejection control (ADRC) with differential privacy preserving for switched multiple systems. By adding privacy noise with Laplace and uniform distribution to the system state, the novel privacy-preserving ADRC (PP-ADRC) framework ensures the required ADRC performance and privacy of system information in a non-ideal network environment. The innovation of this paper is embodied in three aspects: (i) The designed performance-dependent differential privacy preserving mechanism (PD-DPPM) under Laplace noise can adaptively balance system privacy and control performance. (ii) The presented privacy budget lower bound of uniform noise can determine the size of the uniform noise, making the PD-DPPM method more general in the field of privacy preservation. (iii) The newly constructed dual design of switching law and controller for PP-ADRC can ensure boundedness and $H_{\infty } $ performance. Lastly, illustrative examples are provided to explain effectiveness. Note to Practitioners—In practical applications, active disturbance rejection control (ADRC), which consists of extended state observer (ESO), is a practical and effective technique for dealing with nonlinearity and uncertainty. In addition, for multimodal systems, the switched systems consisting of multiple subsystems and control laws can adjust the working mode for achieving better control performance according to changes in real-time system state. Therefore, ADRC is adopted to improve the control performance and efficiency of the switched systems. In the switched ADRC systems, the feedback channel carrying sensitive information about the system state and operation is a key link in closed-loop control. Once it is leaked or stolen, the operation and security of the control system will be destroyed. Therefore, designing a privacy-preserving method for switched ADRC systems to prevent sensitive feedback information from being stolen and ensure the safe and stable operation of the system is of great significance. The designed performance-dependent differential privacy preserving mechanism (PD-DPPM) can achieve privacy and security of feedback information based on real-time system performance, and achieve a balance between privacy and control.}, archive = {J_TASE}, author = {Yiwen Qi and Chen Wang and Simeng Zhang and Choon Ki Ahn}, doi = {10.1109/TASE.2025.3583671}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17209--17220}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Performance-dependent privacy preserving for switched multiple systems under active disturbance rejection control}, volume = {22}, year = {2025}, }
• (2025). Pinning stabilization of logical networks based on deformation of the state transition matrix. TASE, 22, 17199--17208. (www)

  @article{ , abstract = {Boolean Networks (BNs) are a powerful model to describe the way in which genes work and holistically interact with one another in system biology. Based on the requirement to guide the systems to a desired state, the stabilization problem becomes an important issue in BNs. Since the normal state feedback control puts controllers to all nodes, the control cost is relatively high. In this paper, we investigate the stabilization problem of BNs under pinning control strategy. Meanwhile, to cut down the cost of control to the most, the set of pinned nodes is minimized. Specifically, a four-procedure method as well as the corresponding computationally feasible algorithms are proposed to determine a minimum set of controlled nodes based on a deformed state transition matrix and a constructed digraph. In comparison with the strategy that is traditionally based on the state transition matrix, the proposed approach can effectively reduce the computational complexity. Finally, gene networks are discussed as simulations, which demonstrate the effectiveness of the proposed method, minimizing the number of controlled nodes with lower time complexity. Note to Practitioners—BNs are commonly used to explore gene interactions and evolutions in modeling and analyzing cellular processes and neural networks. When it comes to biological systems or genetic networks, it’s critical to create therapeutic interventions that guide patients toward and sustain desired states, like health. Thus, the requirement to steer BNs to a preassigned state through control design is what motivates this study. In order to ensure the arbitrariness of the preassigned state and minimize the control cost, pinning control with minimum controlled nodes is a wise option. However, existing pinning strategies based on state transition matrix generally fall short of practice due to the high time complexity. To address the issue, the present work proposes a four-procedure pinning control method via a graph method to seek for the minimum set of pinned nodes with lower time complexity and verifies on different gene networks.}, archive = {J_TASE}, author = {Jiayang Liu and Amol Yerudkar and Shuting Sun and Yang Liu}, doi = {10.1109/TASE.2025.3583327}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17199--17208}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Pinning stabilization of logical networks based on deformation of the state transition matrix}, volume = {22}, year = {2025}, }
• (2025). Adaptive neural-network sliding mode admissible consensus for discrete-time singular multi-agent systems under stochastic topology. TASE, 22, 17189--17198. (www)

  @article{ , abstract = {The information interaction between agents in this paper is described using Markov switching topology, and an adaptive neural-network sliding mode (ANNSM) controller is proposed to address the admissible consensus problem for discrete-time nonlinear singular multi-agent systems (SMASs) with external disturbances. For the reason of approximating the nonlinear part online and overcoming the bounded difficulty, the radial basis function (RBF) based on neural network (NN) is introduced. When designing the state observer, a disturbance observer assisted by an adaptive update law is established for the purpose of estimating the external disturbance and the NN residual error. Next, an ANNSM control method related to the Markov switching topology is proposed based on the designed disturbance observer, and the Lyapunov stability theory demonstrates that the system achieves admissible bounded consensus. Finally, simulation studies on a numerical example and a distributed microgrid model to demonstrate the effectiveness of the proposed controller. Note to Practitioners—This paper was motivated by the effective control problem of microgrids. Microgrid has multiple distributed generators and could be modelled by multi-agent systems. Existing descriptions of the topology between the leader mircogrid and the following distributed generators are generally the fixed ones, which could not well depict the variable information interaction due to the external factors. Thus the Markov jumping topologies are introduced to the microgrid multi-agent model. Considering the various external disturbances to the voltage value, this paper establishes a disturbance observer to estimate the unknown disturbances, and suggests a new sliding mode control method based on the RBF neural network using an adaptive update control law to realize the leader-follower consensus problem. The feasibility of the proposed controller on the model is verified by solving and simulation.}, archive = {J_TASE}, author = {Jing Xie and Yongxin Qiu and Sa Cao and Baoping Jiang and Yonggui Kao}, doi = {10.1109/TASE.2025.3582766}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17189--17198}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neural-network sliding mode admissible consensus for discrete-time singular multi-agent systems under stochastic topology}, volume = {22}, year = {2025}, }
• (2025). Two-echelon arc routing problem with drones in traffic patrol. TASE, 22, 17170--17188. (www)

  @article{ , abstract = {Traditional patrolling methods relying on ground vehicles or fixed cameras are insufficient in adapting to dynamic changes in traffic flow due to their limited mobility and fixed positions, especially during emergencies like traffic accidents occur. In this paper, we investigate a new Truck-Drone collaborative Traffic Patrolling (TDTP) mode, where multiple trucks and drones work in a collaborative manner to enhance the efficiency of performing traffic patrolling tasks. The drone can be launched and recovered by trucks at certain points along the truck route, making it possible for both trucks and drones to perform patrol tasks simultaneously. A crucial aspect of TDTP is the coordinated arc routing scheduling of multiple trucks and drones, which is referred to as the Truck-Drone Arc Routing Problem (TD-ARP). We formulate the TD-ARP as a mixed-integer linear programming model with the objective of minimizing the total patrol time of all trucks and drones. Due to the difficulty of solving large instances of TD-ARP to optimality, we construct a divide and conquer based framework TA-TDAR (Task Allocation phase and Truck Drone fleet Arc Routing phase) for TD-ARP. TA involves the assignment of tasks to different truck-drone fleets, each comprising a truck and multiple drones, while TDAR focuses on path planning for the truck-drone fleets. Two phases are iteratively performed until the stopping criteria are met. In TA phase, a Task Allocation Heuristic method (TAH) is designed. Based on the task allocation results, an Adaptive Large Neighborhood Search algorithm with Local Search (ALNSLS) is embedded into TDAR phase to achieve a satisfactory scheduling scheme of each truck-drone fleet. Extensive experiments are conducted to validate the superiority of the proposed TAH-ALNSLS. The results show that our method has a potential improvement of up to 40% compared to the comparison methods, which stems from the divide and conquer framework and problem specific operators. Furthermore, several management implications are provided to assist practitioners in making efficient decisions. Note to Practitioners—This study is motivated by the need for developing effective and automated truck-drone collaborative traffic patrolling systems which can achieve complementary advantages to enhance the efficiency of performing traffic patrolling tasks. In truck-drone collaborative traffic patrolling systems, the drones provide flexible and efficient traffic patrol without the limitations of the road network, utilizing trucks as launch and recovery platforms to overcome the short battery endurance of drones. The successful implementation of truck-drone patrol system requires the coordinated arc routing scheduling between the trucks and drones. However, the truck-drone arc routing scheduling still remains largely unexplored. To establish truck-drone collaboration traffic patrol systems and obtain a high-quality patrol path, we formulate a mixed integrated model and develop a two-phase optimization method. The experimental results demonstrate the superiority of the proposed method. In addition, several meaningful insights are proposed based on experimental results, providing guidance for the practical application of the models and methods proposed in this paper.}, archive = {J_TASE}, author = {Jiaqi Cheng and Tianjun Liao and Guohua Wu and Zhongqiang Ma and Jianmai Shi}, doi = {10.1109/TASE.2025.3575171}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17170--17188}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Two-echelon arc routing problem with drones in traffic patrol}, volume = {22}, year = {2025}, }
• (2025). Robot dexterous grasping in cluttered scenes based on single-view point cloud. TASE, 22, 17160--17169. (www)

  @article{ , abstract = {Grasping is a basic but challenging task in intelligent robotic manipulation, and grasping pose detection is the key in this task. Most current work is shifting from 2D planar grasping to more flexible six-degree-of-freedom (6-DoF) grasping, and some significant progress has been made. However, there are still limitations such as low success rate and poor robustness in cluttered scenes. In this paper, we investigate 6-DoF grasping in cluttered scenes, and propose a cascaded multi-target learning network based on self-attention mechanism and multiscale sampling. The self-attention mechanism effectively improves the network’s attention to the correct grasping pose, while multiscale sampling improves the network’s adaptability to grasping objects of different sizes. In addition, we propose a dual-objective evaluation metric based on the force-closure metric and the center-of-mass distance metric, which can make a more reasonable and reliable evaluation of the grasping poses in the dataset. Our model is evaluated on large-scale benchmarks as well as the real robot system. The proposed method achieves state-of-the-art results on GraspNet-1Billion (8.8+AP), and shows 95.24% success rates in real cluttered scenes. Note to Practitioners—This paper presents a framework for 6-DoF grasping in cluttered scenes using single-view point clouds, improving success rates without requiring complete point clouds. By introducing a self-attention mechanism and multiscale sampling, the model adapts to varying object sizes and enhances focus on the correct grasping pose. A dual-objective evaluation metric based on force-closure and center-of-mass distance also refines dataset quality, balancing force and geometric criteria for more reliable grasping poses. This approach supports direct deployment in real-world robotic applications like automated sorting and assembly, with high robustness and efficiency in cluttered environments.}, archive = {J_TASE}, author = {Qingxing Zhao and Minhua Zheng and Zhaoxin Li and Shichang Huang and Wen Shi}, doi = {10.1109/TASE.2025.3575097}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17160--17169}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robot dexterous grasping in cluttered scenes based on single-view point cloud}, volume = {22}, year = {2025}, }
• (2025). Model reconstruction-based optimal adaptive prescribed time control for multi-stage precision robotic arms: Inequality constraints, uncertainties, disturbances. TASE, 22, 17147--17159. (www)

  @article{ , abstract = {There are always multi-stage and multi-type precision requirements for robotic arms in industrial manufacturing. Due to uncertainties and disturbances, the tracking error is forced to exceed the time-varying precision boundary and the system is even broken. Therefore, an intelligent optimization framework of fuzzy adaptive prescribed time control for robotic arms with multi-stage precision is developed in this paper. Firstly, the multi-stage precision requirements are explained as time-varying piecewise inequality constraints of the tracking error. Based on homeomorphic mapping, a model reconstruction method is proposed by designing the transformation function to form a reconstructed system. Then, based on the constraint force analysis of the reconstructed system, a leaky adaptive prescribed time control method is proposed, which simultaneously handles uncertainties, disturbances and inequality constraints. The error convergence is achieved within a prescribed time by setting piecewise constraints. The overcompensation is avoided via the leaky-type adaptive law. The proposed method is proven to be uniformly bounded and uniformly ultimately bounded. Furthermore, an optimization strategy is designed based on fuzzy set theory to optimize system performance and control cost, forming a set of general trade-off rules. Finally, the effectiveness of the proposed method is verified. The results show that the time-varying piecewise inequality constraints are fully satisfied at low control cost. Note to Practitioners—Robotic arms are widely used in collaborative robots, industrial robots and embodied intelligence. However, these scenarios always have multi-stage, multi-type and time-limited motion accuracy requirements, which are difficult to meet with conventional methods. Moreover, such nonlinear systems are easily affected by uncertainties such as modeluncertainty, external disturbances, and electronic noise. In addition, existing control strategies are difficult to provide engineers with a visual comprehensive optimization and trade-off criteria between control cost and system performance. This paper is dedicated to developing a fuzzy optimal adaptive time-limited control based on model reconstruction to simultaneously address the above limitations. 1) Engineers reconstruct the model based on the principle of homeomorphic mapping and the designed transformation function. Then the multi-stage accuracy requirements are set as inequality constraint boundaries; 2) Motion control is performed based on the proposed leakage adaptive control method. Leakage parameters avoid overcompensation, and high-order terms avoid additional control costs. 3) Based on the description of uncertainty by fuzzy sets, the optimal parameters are solved according to the optimization strategy. The trade-off criteria between performance and cost depend on the trade-off factors. The importance is visualized, so the optimization is intelligent. All the above steps use basic operations and are in analytical form. In future research, multi-parameter optimization strategies will be designed to comprehensively improve system performance.}, archive = {J_TASE}, author = {Like Zong and Cuiqing Jiang and Shengchao Zhen and Chao Ma}, doi = {10.1109/TASE.2025.3582955}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17147--17159}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model reconstruction-based optimal adaptive prescribed time control for multi-stage precision robotic arms: Inequality constraints, uncertainties, disturbances}, volume = {22}, year = {2025}, }
• (2025). Robust precision motion control for a pneumatic flexible manipulator with joint coupling and grasped loads based on a nonlinear ESO and a sliding mode TD. TASE, 22, 17134--17146. (www)

  @article{ , abstract = {In this paper, a motion control strategy is proposed for a pneumatic flexible manipulator with joint decoupling and a grasped load. A nonlinear extended state observer (ESO) is employed to estimate grasped load disturbances and joint coupling in the pneumatic flexible manipulator. An adaptive tornambe controller (TC) is proposed to provide further compensation and achieve robust precision motion control. A sliding mode tracking differentiator (TD) is designed to arrange a rapid transition process for the pneumatic flexible manipulator. Moreover, stability analyses of the sliding mode TD, the nonlinear ESO and the adaptive TC are conducted by Lyapunov methods. Experimental results are given to illustrate the feasibility of the proposed motion control strategy for the pneumatic flexible manipulator. Note to Practitioners—The motivation of this paper is to address the challenges faced by a pneumatic flexible manipulator with joint coupling and a grasping load during motion. Pneumatic flexible manipulators have strong adaptability and high safety, but they face three major challenges: 1. Pneumatic flexible manipulators exhibit high nonlinearity because of the hysteresis and creep characteristics of pneumatic artificial muscles (PAMs), which brings challenges for control performance enhancement. 2. There exists a complex coupling relationship between multiple PAMs in the pneumatic flexible manipulator, which causes difficulty in decoupling by model analysis. 3. Owing to the flexibility of the joints, the adverse effects of grasping loads on the motion control are not negligible. To address the above major challenges, a motion control strategy is proposed combined with an adaptive TC, a nonlinear ESO and a sliding mode TD in this paper. Notably, an adaptive control law is designed with TC to reduce the impact of the varying disturbances on the system control performance. In contrast to more complex decoupling techniques, the proposed adaptive TC is less computationally intensive and easier to implement in practical applications. Comparative experiment results show the feasibility and superiority of the motion control strategy. Future research will concentrate on improving the motion control strategy of the pneumatic flexible manipulators to decrease tracking errors.}, archive = {J_TASE}, author = {Ling Zhao and Rui Wang and Tao Wang and Peng Li}, doi = {10.1109/TASE.2025.3582910}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17134--17146}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust precision motion control for a pneumatic flexible manipulator with joint coupling and grasped loads based on a nonlinear ESO and a sliding mode TD}, volume = {22}, year = {2025}, }
• (2025). Event-triggered impulsive control for time-varying systems. TASE, 22, 17127--17133. (www)

  @article{ , abstract = {This paper addresses the stability of time-varying systems using event-triggered impulsive control. A new event-triggered mechanism with time-varying structure is proposed, ensuring global asymptotic stability and avoiding Zeno behavior. The relationship between impulse actions, time delay, and system dynamics is explored, with two examples confirming the results. Note to Practitioners—The motivation of this paper is the challenge of stabilizing time-varying systems with reduced control efforts, which is a common issue in automation and control systems. Specifically, the paper introduces an event-triggered impulsive control (ETIC) approach to address this problem. The novelty of this work lies in the development of a new event-triggered mechanism (ETM) that incorporates a time-varying structure. This enables the system to use its own time-varying dynamics to trigger control actions, reducing the frequency of impulse actions compared to previous mechanisms. This leads to a reduction in control costs without compromising system stability. The proposed method guarantees global asymptotic stability (GAS) while preventing Zeno behavior, which can be problematic in traditional event-triggered systems. The practical benefits of this research extend to various automation applications, such as robotic control systems, where minimizing control actions while ensuring stability is critical for efficiency and cost reduction. For example, the method is demonstrated in a wheeled mobile robots velocity tracking. Future extensions of this work could explore its application in more complex systems, including those with multiple interacting subsystems or with real-time disturbances. This approach can be further refined to handle non-linear dynamics and stochastic disturbances, potentially expanding its applicability to a broader range of automated processes and systems in industrial and robotic settings.}, archive = {J_TASE}, author = {Weilian Liu and Xiaodi Li}, doi = {10.1109/TASE.2025.3582888}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17127--17133}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered impulsive control for time-varying systems}, volume = {22}, year = {2025}, }
• (2025). Shared control with obstacle avoidance for UGVs. TASE, 22, 17109--17126. (www)

  @article{ , abstract = {Uncrewed ground vehicle (UGV) applications, such as warehouse operations, assembly-line production, infrastructure inspection, surveillance, precision farming, and search & rescue, can benefit from shared control, in which a human can semi-automatically control the UGV when needed and let it operate fully-automatically, when desired. Many algorithms have been developed to permit a UGV to semi-autonomously conduct tasks, either individually, or in a group. However, a complete semi-autonomous system that works wherever, and whenever, needed is far from being implemented. Here, we develop a human-robot shared controller for the supervisory control of one or more UGVs by a single person. The shared controller blends an automatic control input with a human control input. The automatic control input consists of a trajectory tracking controller and a control barrier function based input term for collision avoidance. A joystick is used to provide the human control input. Human intent is measured employing a Lyapunov-like storage function, which is used in a convex function based blending law that continuously varies the magnitude of the control inputs coming from the human and the machine. The approach permits us to theoretically prove the asymptotic stability of the closed-loop system. The shared controller is validated using both a physical robot in a cluttered indoor environment, and a hardware-in-the-loop simulated robot operating in virtual warehouse environment. Note to Practitioners—The methods presented in this paper can be applied to numerous low-speed applications that employ uncrewed ground vehicles. The approach is designed to work well for arbitrarily complex, smooth reference trajectories. The practitioner should develop an accurate simulation model and tune the controller parameters to optimize the system control gains. The simulation model should be used to ensure that the system can meet the requirements of a given application. The trajectory tracking controller can be tuned to minimize the mean-squared-error between the robot’s actual position and its desired position, where the positioning error depends on the robot’s kinematics and the accuracy of its localization system. The system is expected to work well with obstacles approximated with circles. To achieve good implementation, the system should have a reliable communication channel with a time lag that is considerably shorter than typical human reaction time.}, archive = {J_TASE}, author = {Cheikh Melainine El Bou and Florian Beck and Karl D. von Ellenrieder and Satyandra K. Gupta}, doi = {10.1109/TASE.2025.3582522}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17109--17126}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Shared control with obstacle avoidance for UGVs}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance control for multiple parametric strict feedback systems with unknown control coefficients. TASE, 22, 17098--17108. (www)

  @article{ , abstract = {This paper investigates the problem of prescribed performance control (PPC) for multiple parametric strict feedback (PSF) systems with unknown control coefficients in switching topologies. The synchronization error is regulated into the prescribed performance constraints within a user-given time, and agents’ outputs are synchronized accurately; furthermore, the results are independent of initial values. Compared with existing research with the same problem formulation, this paper investigates a more general case, reduces the communication load, and obtains a global result. The numerical simulations confirm the feasibility of the control algorithm presented. Note to Practitioners—This paper is motivated by the unsatisfactory performance during the control of multiple systems with unknown control coefficients, such as Norrbin systems and hypersonic vehicles with actuator gain failure. This is owing to the divergent properties of the Nussbaum function, which is considered an effective method for handling the unknown control coefficient. Most of the existing methods for studying such systems focus on the steady-state performance and ignore the transient progress. This paper unites prescribed performance control (PPC), backstepping control, and cooperative control to realize steady-state and transient control objectives simultaneously and removes the restrictions on initial values, which frequently arise in traditional PPC. Therefore, the result is global. The proposed controller guarantees that the synchronization errors are limited within a prescribed range and ultimately converge to 0; that is, the output synchronization is achieved precisely. Moreover, it only requires the output information of agents to be transmitted throughout the communication network and reduces the communication burden. We then show two simulation examples to demonstrate the validity of the algorithm. In future research, we will consider the impact of time delays.}, archive = {J_TASE}, author = {Junmin Peng and Jianbo Li and Chao Huang and Xin Wang}, doi = {10.1109/TASE.2025.3582230}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17098--17108}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance control for multiple parametric strict feedback systems with unknown control coefficients}, volume = {22}, year = {2025}, }
• (2025). Active power control and load suppression for wind turbines based on a variable-degree-of-freedom framework. TASE, 22, 17088--17097. (www)

  @article{ , abstract = {Structural load suppression of wind turbines is one of the important means to improve the economic efficiency of wind farm operation. However, wind turbines are complex systems coupled with multiple control variables. There is no comprehensive solution for the flexible and coordinated control of multiple controllable degrees of freedom. The main difficulty lies in the greater optimization burden of multi-degree-of-freedom systems. This paper proposes a variable degree-of-freedom control system for wind turbines, and uses multivariable model predictive control to design collaborative controllers for generator torque, pitch angle and yaw angle. To avoid the increase in complexity caused by the continuous intervention of yaw control, a yaw control cut-in/cut-out strategy is set, and the non-disturbance of switching is theoretically proved. Simulation results show that the proposed variable degree-of-freedom control system can effectively realize active power control and structural load suppression of wind turbines in a large wind speed range. Note to Practitioners—This paper focuses on efficiently using wind turbines’ control degrees of freedom to achieve better structural load suppression. The control strategy of existing commercial wind turbines is decentralized. This paper innovatively considers the organic coordination of generator torque, pitch, and yaw angles, and designs two- and three-degree-of-freedom controllers using model predictive control methods. Considering that the yaw angle control should be reduced in practical applications, this paper designs a switching strategy that can achieve disturbance-free switching between two- and three-degree-of-freedom. These works can provide a new development idea for the control framework of wind turbines and better serve the structural load suppression during the operation of wind turbines.}, archive = {J_TASE}, author = {Qi Yao and Qian Song and Yi Zhang and Yang Hu and Fang Fang}, doi = {10.1109/TASE.2025.3582138}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17088--17097}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active power control and load suppression for wind turbines based on a variable-degree-of-freedom framework}, volume = {22}, year = {2025}, }
• (2025). Koopman operator-based data-driven online learning control for an omni-directional mobile manipulator. TASE, 22, 17077--17087. (www)

  @article{ , abstract = {Omni-directional mobile manipulators (OMMs) are typically nonlinear, strongly coupled, multiple-input and multiple-output systems, for which the development of mechanistic models is often complex and time-consuming. Koopman operator theory is a fully data-driven modeling approach that leverages input-output data to generate high-dimensional linear models, but it often has modeling errors. In this paper, a completely data-driven online learning linear model predictive control (MPC) framework is proposed for an OMM, without using any prior knowledge of the robot system. A finite-dimensional approximate linear Koopman model is established for an OMM using the input-output data. The model errors (including external disturbances) are online learned by Gaussian process regression (GPR) using the collected data and compensated for in real time within the controller. Selective forgetting and incremental inverse computation methods are employed to reduce the computational cost of online GPR. Finally, a total of 11,400 experimental data pairs are generated for Koopman modeling of an OMM prototype, utilizing randomly generated control inputs under different initial states. Then experimental tests are carried out to verify the control performances and robustness of the proposed control scheme. Note to Practitioners—Most existing control design schemes for OMMs require mechanism models. However, the modeling process for OMMs is usually time-consuming, and the resulting model is often complex with parameter uncertainties. This paper proposes a completely data-driven modeling and control scheme for OMMs without using any prior knowledge of the robot system. Based on Koopman operator, we establish a finite-dimensional approximate linear dynamic model for an OMM via the input-output experimental data. Then we propose a simple linear MPC framework for OMM, while the model errors are online learned by GPR using the collected experimental data. The computational cost of online GPR is also considered. Experimental tests demonstrate that the proposed control scheme can achieve good tracking accuracy and robust performances.}, archive = {J_TASE}, author = {Chao Ren and Binjie Wang and Wendong Niu and Shuai Li and Shugen Ma}, doi = {10.1109/TASE.2025.3582032}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17077--17087}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Koopman operator-based data-driven online learning control for an omni-directional mobile manipulator}, volume = {22}, year = {2025}, }
• (2025). An adaptive grasping force tracking strategy for nonlinear and time-varying object behaviors. TASE, 22, 17063--17076. (www)

  @article{ , abstract = {Accurate grasp force control is one of the key skills for ensuring successful and damage-free robotic grasping of objects. Although existing methods have conducted in-depth research on slip detection and grasping force planning, they often overlook the issue of adaptive tracking of the actual force to the target force when handling objects with different material properties. The optimal parameters of a force tracking controller are significantly influenced by the object’s stiffness, and many adaptive force tracking algorithms rely on stiffness estimation. However, real-world objects often exhibit viscous, plastic, or other more complex nonlinear time-varying behaviors, and existing studies provide insufficient support for these materials in terms of stiffness definition and estimation. To address this, this paper introduces the concept of generalized stiffness, extending the definition of stiffness to nonlinear time-varying grasp system models, and proposes an online generalized stiffness estimator based on Long Short-Term Memory (LSTM) networks. Based on generalized stiffness, this paper proposes an adaptive parameter adjustment strategy using a PI controller as an example, enabling dynamic force tracking for objects with varying characteristics. Experimental results demonstrate that the proposed method achieves high precision and short probing time, while showing better adaptability to non-ideal objects compared to existing methods. The method effectively solves the problem of grasp force tracking in unknown, nonlinear, and time-varying grasp systems, demonstrating the generalization capability of our neural network and enhancing the robotic grasping ability in unstructured environments. Note to Practitioners—In unstructured scenarios such as daily life and agriculture, robots often need to adaptively control grasping force to safely handle objects with unknown properties like weight or hardness. Based on existing methods for determining the desired grasping force, this work focus on controlling actual grasping force to dynamically track the desired force. In designing this force tracking controller, object stiffness significantly impacts optimal parameters. Many adaptive controllers rely on stiffness estimation, but real-world objects often exhibit nonlinear, time-varying behaviors such as viscosity and plasticity, making existing estimators designed for ideal elastic bodies prone to errors. To address this, we propose a generalized stiffness concept suitable for non-ideal objects and a corresponding LSTM-based online generalized stiffness estimator, enabling practitioners to achieve adaptive parameter tuning for the force tracking controller. As an example, we implement this estimator on a PI controller. Experiments demonstrate our controller’s strong adaptability across elastic, viscoelastic, plastic, and variable-stiffness objects, achieving high accuracy and fast tracking speed in grasping tasks.}, archive = {J_TASE}, author = {Ziyang Cheng and Xiangyu Tian and Ruomin Sui and Tiemin Li and Yao Jiang}, doi = {10.1109/TASE.2025.3581688}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17063--17076}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptive grasping force tracking strategy for nonlinear and time-varying object behaviors}, volume = {22}, year = {2025}, }
• (2025). Noise adaptive filtering neural network under multiscale features. TASE, 22, 17050--17062. (www)

  @article{ , abstract = {Various uncertain disturbances in industrial processes bring noise to industrial process data, which brings great challenges to industrial soft sensor modeling. Traditional soft sensor models focused on removing noise in the process data, but it is almost impossible to remove all noise in actual engineering. Therefore, a novel noise adaptive filtering method integrating the multiscale neural network (NAF-MSNN) is proposed for the soft sensor, which can incorporate a noise processing mechanism that adaptively removes noise at different scales during feature extraction. The MSNN extracts overall trend and local trend features through the multiscale convolution. Then, the NAF converts multiscale features into frequency domain features, and constrains the noise filter matrix through proposed piecewise regularization to select important frequency domain components at different scales. Moreover, the multiscale fusion module controls denoised multiscale features exchange fusion between different scale based on the important measurement of each corresponding scale. Finally, the gated recurrent unit (GRU) establishes the dynamic relationships between the fused multiscale features and the key indicator. The proposed NAF-MSNN is compared with state-of-the-art soft sensor models in three datasets. In terms of R2 metrics, the accuracy improvement of NAF-MSNN reaches 4%, 8% and 3% in the public dataset and two industrial datasets. Note to Practitioners—Due to uncertain production environments, there may be a large amount of noise in industrial process data. The focus of this paper is to build a noise adaptive filtering method integrating the multiscale neural network (NAF-MSNN) to improve the robustness to noise. The proposed model can adaptively eliminate noise that has a large impact on the prediction results according to different scale features during the training process, thereby achieving high-precision soft sensing in an uncertain noise environment. In addition, we conducted comparative experiments and noise injection experiments on two real industrial data sets. The proposed method has higher soft sensing accuracy and stronger noise robustness than baseline methods.}, archive = {J_TASE}, author = {Xuan Hu and Peihao Zheng and Zhiqiang Geng and Yongming Han}, doi = {10.1109/TASE.2025.3579694}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17050--17062}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Noise adaptive filtering neural network under multiscale features}, volume = {22}, year = {2025}, }
• (2025). MG-SLAM: Structure gaussian splatting SLAM with manhattan world hypothesis. TASE, 22, 17034--17049. (www)

  @article{ , abstract = {Gaussian Splatting SLAMs have made significant advancements in improving the efficiency and fidelity of real-time reconstructions. However, these systems often encounter incomplete reconstructions in complex indoor environments, characterized by substantial holes due to unobserved geometry caused by obstacles or limited view angles. To address this challenge, we present Manhattan Gaussian SLAM, an RGB-D system that leverages the Manhattan World hypothesis to enhance geometric accuracy and completeness. By seamlessly integrating fused line segments derived from structured scenes, our method ensures robust tracking in textureless indoor areas. Moreover, The extracted lines and planar surface assumption allow strategic interpolation of new Gaussians in regions of missing geometry, enabling efficient scene completion. Extensive experiments conducted on both synthetic and real-world scenes demonstrate that these advancements enable our method to achieve state-of-the-art performance, marking a substantial improvement in the capabilities of Gaussian SLAM systems. Note to Practitioners—This paper was motivated by the limitations of Gaussian Splatting SLAM systems in complex indoor environments, where textureless surfaces and obstructed views often lead to substantial tracking errors and incomplete maps. While existing systems excel in high-fidelity reconstruction, they struggle with frame-to-frame or point-feature tracking in large-scale environments, particularly with significant camera rotations and obscured structures. In this paper, we enhance the neural dense SLAM by integrating the Manhattan World hypothesis, applying its parallel line and planar surface constraints for more robust tracking and mapping. We incorporate line segment features into both tracking and mapping to improve structural accuracy. Moreover, we propose a post-optimization method that interpolates new Gaussian primitives to effectively fill gaps on planar surfaces. Extensive experiments on multiple datasets demonstrate the superiority of our approach in large-scale indoor environments, resulting in more accurate tracking and mapping.}, archive = {J_TASE}, author = {Shuhong Liu and Tianchen Deng and Heng Zhou and Liuzhuozheng Li and Hongyu Wang and Danwei Wang and Mingrui Li}, doi = {10.1109/TASE.2025.3575772}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17034--17049}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MG-SLAM: Structure gaussian splatting SLAM with manhattan world hypothesis}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant mobile service offloading in mobile edge computing. TASE, 22, 17022--17033. (www)

  @article{ , abstract = {Mobile edge computing (MEC) is an evolving paradigm for rendering services through network-accessible resources deployed over Internet of Things (IoT) nodes at the edge. Nevertheless, an MEC environment usually employs thousands of physical machines connected via hundreds of switches/routers that communicate and coordinate to deliver computing service. In such complicated systems, faults caused by software, human errors, and hardware are often unavoidable. The edge of network presents a dynamic environment with great quantities of terminals, high mobility of mobile devices, heterogeneous applications, and intermittent traffic. In such an environment, MEC can suffer from unbalanced resource provisioning and interruptions of faults occurring at different levels, which further causes task faults and affects service quality. To address this challenge, this work proposes a novel fault-tolerant offloading method for handling faults by leveraging a reinforcement-learning-based service offloading decision model. The model synthesizes a Dueling Deep Q Network (DQN)-based algorithm for deciding user offloading behaviors and an adaptive checkpointing method for improving task execution reliability. For the purpose of model validation and comparison, extensive simulations are conducted. Numerical results clearly demonstrate that the proposed method is highly effective and outperforms existing methods. Note to Practitioners—Motivated by industrial demands for task offloading in a fault-prone mobile edge computing (MEC) scenarios, this work aims at achieving the high reliability and fault-tolerability of all offloadable tasks. The proposed method leverages a reinforcement learning-based Dueling DQN model for deciding user offloading behaviors and an adaptive checkpointing method for improving task execution reliability. In the experiments, compared with its peers, the proposed approach reduces the completion time of offloading tasks by 9.29% and fault rate by 48.69% on average, respectively. This is very significant to industrial MEC applications, e.g., smart cities.}, archive = {J_TASE}, author = {Tingyan Long and Yunni Xia and MengChu Zhou and Jianqi Li and Yong Ma and Yusuf Al-Turki}, doi = {10.1109/TASE.2025.3557934}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17022--17033}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant mobile service offloading in mobile edge computing}, volume = {22}, year = {2025}, }
• (2025). An experimental validation of nonlinear extended state observer controlled small hydro-PV array and battery storage-based microgrid for remote areas. TASE, 22, 17010--17021. (www)

  @article{ , abstract = {This paper presents a hybrid microgrid system designed for rural areas, integrating small hydro and photovoltaic (PV) arrays with battery storage to ensure uninterrupted power supply to local loads in isolated regions. A self-excited induction generator (SEIG) is employed for small hydro power generation, while power quality issues caused by nonlinear loads are addressed using the Volterra Filtered-X Least Mean Square (VFXLMS) technique. This method effectively suppresses dominant harmonic components in load currents and accurately estimates the fundamental frequency of load currents. The PV array is interconnected at the DC link of a voltage source converter (VSC) through a boost converter, while a bidirectional converter manages battery operations, facilitating efficient power management. The bidirectional converter supports the export and import of power from/to the battery, optimizing energy utilization within the microgrid. Conventional proportional-integral (PI) control has limitations in DC voltage tracking and mitigating overshoot and undershoot. To address this, a nonlinear extended state observer (NESO) is employed to mitigate disturbances. The NESO, combined with a disturbance observer, enhances the operational reliability of the microgrid. An experimental setup has been developed to validate the system’s performance under conditions of unbalanced loads and varying PV insolation levels. Note to Practitioners—The hybrid microgrid system proposed in this study offers a practical solution for addressing the energy needs of rural and isolated regions. By integrating small hydro and photovoltaic (PV) arrays with battery storage, the system ensures reliable power delivery to local loads. Practitioners working in rural electrification can benefit from the system’s ability to manage unbalanced loads and varying renewable energy sources effectively. A key feature of the proposed system is the use of a self-excited induction generator (SEIG) for small hydro power generation. This choice simplifies the system’s deployment and maintenance, making it suitable for remote areas. Additionally, the Volterra Filtered-X Least Mean Square (VFXLMS) technique effectively mitigates harmonic distortions caused by nonlinear loads, enhancing the power quality delivered to consumers. Energy management within the microgrid is streamlined through the integration of a boost converter for PV arrays and a bidirectional converter for battery control. The bidirectional converter optimizes energy utilization by facilitating power export and import, ensuring balanced energy flows during varying load and generation conditions. The study highlights the limitations of conventional proportional-integral (PI) controllers in handling dynamic disturbances. The implementation of a nonlinear extended state observer (NESO), combined with a disturbance observer, provides a robust solution for achieving precise DC voltage regulation, minimizing overshoot and undershoot, and enhancing overall system stability. This work also underscores the importance of experimental validation. The developed experimental rig demonstrates the system’s performance under real-world scenarios, such as unbalanced loads and fluctuating PV insolation, offering confidence in its practical applicability. Practitioners are encouraged to explore this hybrid microgrid design as a scalable and cost-effective approach to rural electrification, especially in areas with access to small hydro resources and solar potential. The system’s modular architecture and advanced control strategies ensure adaptability to various environmental and operational conditions.}, archive = {J_TASE}, author = {Shailendra Kumar and Bhim Singh and Souvik Das}, doi = {10.1109/TASE.2025.3581222}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {17010--17021}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An experimental validation of nonlinear extended state observer controlled small hydro-PV array and battery storage-based microgrid for remote areas}, volume = {22}, year = {2025}, }
• (2025). Dynamic memory event-triggered lag consensus of multi-UAV systems with hybrid attacks over stochastic switching topology. TASE, 22, 16999--17009. (www)

  @article{ , abstract = {The lag consensus problem of multi-uncrewed aerial vehicle (UAV) systems under hybrid attacks is investigated in this paper. First, a dynamic memory event-triggered control protocol is proposed for the multi-UAV system whose normal network communication would be hindered by denial-of-service (DoS) attacks. Different from traditional event-triggered mechanisms, we consider both historically transmitted data and dynamic threshold in the dynamic memory event-triggered protocol, and it will make less data transmissions and better control performance. Second, due to the communication structure is not fixed, we establish a switching-topology-based distributed control architecture. In view of the fact that communication delays among agents cannot be ignored, a distributed controller is proposed to achieves lag consensus of the multi-UAV system. And then, an estimator is designed to address the situation which the system state cannot be measured during the control process. Additionally, the practicality of the distributed control scheme is analyzed by ruling out Zeno behavior. Ultimately, the effectiveness and validity of the proposed control scheme are confirmed through a simulation example. Note to Practitioners—The motivation of this paper is the consensus issue of multi-uncrewed aerial vehicle (UAV) systems. In the collaborative operation of existing UAV systems, it is crucial to ensure that UAVs can maintain a certain level of coordination even in the face of hybrid attacks such as denial-of-service (DoS) attacks and deception attacks. This paper delves into the issue of lag consensus in multi-UAV system under hybrid attacks, the proposed dynamic memory event-triggered protocol can be embedded into the control software of the multi-UAV system. And the event-triggered protocol can avoid the drawbacks of continuous communication in traditional consistency control and better conserving resources. In this paper, the dynamic 3-Degrees-Of-Freedom (3DOF) motion model of UAV is constructed and we can use computers to design distributed controllers for the system model. Preliminary simulation experiments have demonstrated the feasibility of this strategy, but it has not yet been tested in actual multi-UAV systems. In future research, we will consider fixed time control strategies to accelerate the achievement of control objectives.}, archive = {J_TASE}, author = {Ji Ma and Xiangyong Chen and Guanghui Wen and Junyi Wang and Feng Zhao and Jianlong Qiu}, doi = {10.1109/TASE.2025.3580147}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16999--17009}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic memory event-triggered lag consensus of multi-UAV systems with hybrid attacks over stochastic switching topology}, volume = {22}, year = {2025}, }
• (2025). Distributed filtering for multi-rate nonlinear quantization systems with fading channels and sensor resolutions: Handling amplify-and-forward relays. TASE, 22, 16987--16998. (www)

  @article{ , abstract = {This paper addresses the distributed recursive filtering problem for a class of time-varying multi-rate nonlinear systems which are subject to fading channels, sensor resolutions and quantization effects. For the purpose of prolonging the transmission distance, the amplify-and-forward (AaF) relay is deployed in the wireless communication channel. The iterative method is employed to convert the underlying multi-rate system into a single-rate one. Before entering communication network for transmission, measurement outputs are quantized using a logarithmic quantizer. Moreover, a new equivalent delay-free measurement sequence is defined by the measurement reorganization technology to tackle random transmission delays. Our main aim is to design a distributed recursive filter such that an upper bound for the filtering error covariance is determined and subsequently minimized by properly choosing the filtering gains. Next, the performance of the proposed filtering algorithm is analyzed through establishing a sufficient condition for the boundedness of the filtering error covariance. Finally, a numerical simulation is used to testify the applicability and effectiveness of the developed filtering algorithm. Note to Practitioners—In practical process industries, signals from sensors are typically encapsulated into data packets for transmission to estimators via wireless communication network. As the limited transmission capability of the sensors, the transmission distance are limited, extending the propagation distance and enhancing the communication quality become a crucial design problem for filters in wireless network system. In this context, some efficient relay-based strategies have been introduced to facilitate signal transmission over long-distance communication channels. Up to now, relay-based communication plays an important role in 5G mobile network, autonomous aerial vehicle formation, navigation, and so on. Besides, the sensor resolutions, multi-rate samplings, fading channels, and quantization effects are also considered in this paper due to the technical/physical constraints, which may complicate the transmission behaviors of wireless communication. Therefore, in this paper, the distributed filtering problem is investigated in multi-rate nonlinear systems subject to sensor resolutions, fading channels and quantization effects under the relay-based mechanism. The proposed filtering algorithm is bounded under certain conditions, which ensures the reliability of the system. Future work is planned to extend the proposed filtering method to more complicated systems with cyber-attacks and bit rate constraints.}, archive = {J_TASE}, author = {Xiaohan Liu and Chunyan Han and Wei Wang}, doi = {10.1109/TASE.2025.3577977}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16987--16998}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed filtering for multi-rate nonlinear quantization systems with fading channels and sensor resolutions: Handling amplify-and-forward relays}, volume = {22}, year = {2025}, }
• (2025). Enhancing attitude tracking with self-learning control using tanh-type learning intensity. TASE, 22, 16976--16986. (www)

  @article{ , abstract = {This paper investigates the attitude tracking control problem for spacecraft. A tanh-type self-learning control (TSLC) approach with variable learning intensity (VLI) is proposed, which avoids saturation while overcoming previous algorithms’ long response time disadvantage. Unlike the previously introduced VLI method, the enhanced TSLC does not tweak the learning intensity based on the previous controller output. Instead, it relates learning intensity to an intermediate variable directly related to the system state and tunes the learning intensity using a tanh-type function. Since the system state reflects the tracking error in real-time, the transformed tanh-type function has a higher decay rate than the exponential function, which not only significantly reduces the saturation response but also improves the response speed and achieves higher steady-state accuracy. Simulation proved TSLC’s superiority, considering adverse actuator factors such as dead zone, bias torque, and saturation. The proposed approach has also been validated on the Quanser helicopter platform, confirming its better performance. Note to Practitioners—Unlike spacecraft control algorithms based on observers, adaptive methods, or neural networks, the proposed algorithm is simpler in control structure and consumes fewer computational resources. With the self-learning mechanism, the desired control accuracy can be achieved without needing an accurate spacecraft dynamics model. It is model-free and exceptionally easy to implement.}, archive = {J_TASE}, author = {Chengxi Zhang and Weijia Lu and Shunyi Zhao and Jin Wu and Xiaoyu Zhu and Zhijie Liu and Wei He}, doi = {10.1109/TASE.2025.3581953}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16976--16986}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing attitude tracking with self-learning control using tanh-type learning intensity}, volume = {22}, year = {2025}, }
• (2025). Evolution-guided Q-learning with dual swarm intelligence for model-free optimal control. TASE, 22, 16964--16975. (www)

  @article{ , abstract = {In this article, a novel accelerated evolution-guided Q-learning (EGQL) algorithm is introduced to address optimal control problems for unknown nonlinear systems. A novel adaptive evolutionary algorithm is introduced here to enhance problem-solving strategies in two key aspects, which is a concept referred to dual swarm intelligence. First, the novel evolutionary algorithm is incorporated into policy improvement, replacing the gradient descent approach in traditional Q-learning. The method eliminates the need for gradient information in the approximate Q-function, while also enabling more precise policy solutions. Second, the commonly used polynomial model is replaced with a neural network, which is further optimized by the evolutionary algorithm to approximate the critic network. The accuracy of the approximate Q-function is significantly enhanced, improving the precision of value function update. Furthermore, the monotonicity and convergence properties of the algorithm are thoroughly analyzed. Finally, the effectiveness of the EGQL method is validated through two simulation experiments, demonstrating its superiority over traditional Q-learning. Note to Practitioners—For many systems with unknown and nonlinear dynamics, it is challenging to solve optimal control problems. Inefficiencies and slow convergence tend to be exhibited by traditional Q-learning methods for such systems. The EGQL algorithm is introduced in this paper to overcome these challenges by integrating evolutionary computation with Q-learning. A novel evolutionary algorithm is employed in EGQL to improve policy improvement and optimize neural networks for approximating the critic network. By replacing traditional gradient descent with evolutionary strategies, policy accuracy is improved, and the need for gradient information is eliminated. Practitioners can leverage the EGQL algorithm to solve optimal control problems for unknown systems without relying on explicit system models. The algorithm is operated in an off-policy and model-free manner, with only real system data being utilized. The effectiveness of the method is demonstrated through two simulation experiments: a linear electric power system and a nonlinear torsional pendulum system, showcasing its superiority in both types of systems. Overall, an efficient solution for control problems in complex industrial settings is provided by the EGQL method. The integration of evolutionary strategies with reinforcement learning is emphasized in this work, offering a robust framework for adaptive control in nonlinear and unknown systems. The approach is especially useful in domains requiring decision-making in uncertain environments.}, archive = {J_TASE}, author = {Ding Wang and Zeqiang Yuan and Guohan Tang and Jiangyu Wang and Junfei Qiao}, doi = {10.1109/TASE.2025.3581315}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16964--16975}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Evolution-guided Q-learning with dual swarm intelligence for model-free optimal control}, volume = {22}, year = {2025}, }
• (2025). A two-phase planner for messenger routing problem in UAV-UGV coordination systems. TASE, 22, 16948--16963. (www)

  @article{ , abstract = {In this paper, a new Messenger Routing Problem (MRP) is studied, which is motivated by Uncrewed Aerial Vehicles (UAVs) accessing Uncrewed Ground Vehicles (UGVs) to deliver information in UAV-UGV coordination systems. The objective is to minimize the longest path among multiple messengers, ensuring fast and reliable information transmission. Two key challenges arise in tackling this problem. First, the targets are moving, incurring the travel cost between targets to vary with the travel process. Second, the messengers accessing the neighborhood of targets needs to satisfy the communication time constraint. Based on the idea of decoupling, a two-phase planner is proposed to sequentially determine global access sequence and optimize local access locations. In the first phase, a motion prediction module is introduced in the Adaptive Large Neighborhood Search (ALNS) framework to deal with the dynamic characteristics of MRP. In the second phase, an efficient bisection sampling method based on the prediction points is proposed to obtain a shorter access path while satisfying the communication time constraint. Finally, the effectiveness and efficiency of the proposed method are demonstrated by performance evaluation and comparison with the state-of-the-art algorithms. Note to Practitioners—This paper studies an emerging UAV routing problem, the Messenger Routing Problem, which is aimed at providing communication services for multiple mobile UGVs. These UGVs are tasked with large-scale operations such as firefighting, environmental monitoring, and disaster relief, where ground communication is often limited or unreliable. Due to their constrained communication capabilities, UGVs may struggle to obtain timely and essential environmental information. By employing UAVs as messengers to detect the scene and fly over the effective communication range of all mobile UGVs to deliver information, a cooperative air-ground network can be established, which allows for successful operation in environments with unavailable communication infrastructure and poor network connectivity, enabling large-scale collaboration among multiple UGVs. This paper proposes a two-phase planner to effectively search for UAV routes to deliver information in the shortest time. The planner first determines the global access sequence and then optimizes the local access locations. Simulation results demonstrate that the proposed method can generate high-quality tours for messenger UAVs effectively and efficiently. We also offer operational insights for real-world deployments and outline strategies for optimizing the system under hardware constraints. In future work, we aim to focus more on real-world scenarios involving uncertain UGV motion patterns.}, archive = {J_TASE}, author = {Zhao Zhang and Chen Chen and Lingda Wang and Yulong Ding and Fang Deng}, doi = {10.1109/TASE.2025.3580473}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16948--16963}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A two-phase planner for messenger routing problem in UAV-UGV coordination systems}, volume = {22}, year = {2025}, }
• (2025). Observer-based adaptive finite-time containment control for nonlinear multiagent systems subject to FDI attacks. TASE, 22, 16938--16947. (www)

  @article{ , abstract = {This paper investigates the adaptive finite-time containment control problem for a class of nonlinear multiagent systems (MASs) under false data injection (FDI) attacks. On account of the fact that instability is inevitable when the false datas are injected into the researched system, the backstepping technique based on the modified coordinate transformation is applied to eliminate the impact of the FDI attacks. Furthermore, the “dynamic surface control”(DSC) approach is utilized to overcome the issue of “explosion of complexity” caused by repeatedly taking derivatives for the virtual control laws. Then, by designing an observer to refactor the immeasurable states of the MASs, a finite-time adaptive output-feedback tracking containment controller is constructed. It is shown that all the signals in the closed-loop systems are semi-globally practical finite-time stability(SGPFS), and the outputs of all the followers converge to the convex hull spanned by the multiple leaders’s outputs. Besides, the observer errors and the containment errors can converge to a small neighborhood of the origin in finite time. Finally, the simulation results are presented to demonstrate the effectiveness of the proposed containment control protocol. Note to Practitioners—The containment control problem is a hot topic in the field of control, which plays an important role in practical engineering. Especially for this problem of nonlinear MASs, the mathematical models are difficult to be obtained accurately. This paper investigates the adaptive finite-time containment control problem for the nonlinear MASs, whose model can be extended to more complex engineering applications, such as UAV formations and intelligent traffic management. It is worth noting that network security and unmeasured state problems often exist in practical applications. Therefore, by designing an observer to refactor the immeasurable states of the MASs, it is ensured that all the signals in the closed-loop systems are SGPFS. To sum up, the paper proposes an adaptive finite-time containment control strategy, which contributes to the development of containment control for MASs in practical applications.}, archive = {J_TASE}, author = {Haobo Kang and Hongjun Ma and Ben Niu}, doi = {10.1109/TASE.2025.3562026}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16938--16947}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based adaptive finite-time containment control for nonlinear multiagent systems subject to FDI attacks}, volume = {22}, year = {2025}, }
• (2025). Multi-player pursuit-evasion game with interaction constraints: A cooperative game theoretic approach based on coalition structure. TASE, 22, 16925--16937. (www)

  @article{ , abstract = {This paper presents a comprehensive mathematical approach to address the multi-player and multi-objective pursuit-evasion games problem, incorporating coalition structure constraints from a cooperation-competition perspective. Social interaction networks are developed to approximate priority communication alliances based on individual pReferences, establishing a multi-connected topology and decision space for the games. An N-player variable-sum differential game model, featuring autonomous obstacle avoidance, is formulated by integrating kinematic constraints and the social forces method. Rigorous proofs are provided for the uniqueness of payoff distribution, the stability of alliance structures, and the convergence of many-to-many differential games to Nash Equilibrium. Simulation and experimental results are presented to validate the effectiveness and performance of the proposed method. Note to Practitioners—A structured stable alliance is essential for fostering collective intelligence, especially in scenarios marked by intense conflict and the necessity for cooperation. While traditional methods (e.g., optimal control, differential games) are widely studied, many rely on rigid, random, or preassigned large-scale alliances that miss dynamic group interactions and cooperative alliance dynamics, leading to inefficient resource allocation. Drawing inspiration from social force models, we propose a novel coalition-structure-based cooperative game theory framework. By integrating multi-agent interactions via a coupled differential game on interconnected topologies, it offers new perspectives for multi-agent pursuit-evasion with obstacles and coalition constraints, advancing collective intelligence research in military defense, law enforcement, and maritime surveillance—domains where group dynamics are critical.}, archive = {J_TASE}, author = {Xiwen Ma and Maolong Lv and Kairong Duan and Wei Xie and Jingsong Yang and Weidong Zhang}, doi = {10.1109/TASE.2025.3581695}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16925--16937}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-player pursuit-evasion game with interaction constraints: A cooperative game theoretic approach based on coalition structure}, volume = {22}, year = {2025}, }
• (2025). Adaptive tracking control of nonlinear multi-agent systems subject to multiple constraints via multi-dimensional taylor network. TASE, 22, 16913--16924. (www)

  @article{ , abstract = {This paper investigates the adaptive tracking control problem for nonlinear multi-agent systems operating under simultaneous input saturation and output performance constraints. To address asymmetric input saturation, an innovative auxiliary system is developed that generates compensatory signals based on the discrepancy between the input signal and the saturation function output. A central contribution is the introduction of a novel dynamic performance function (DPF), this function leverages signals from the auxiliary system to adaptively adjust performance boundaries, critically activating this adjustment only when input saturation occurs concurrently with synchronization errors exceeding predefined safety limits, thereby effectively resolving conflicts between the input and performance constraints. Furthermore, a first-order filter is employed within the backstepping control design to approximate virtual control derivatives, mitigating the “computational explosion” issue. An adaptive controller incorporating multi-dimensional Taylor network (MTN) is then synthesized based on this framework. Rigorous Lyapunov stability analysis confirms the boundedness of all signals within the closed-loop system. Supporting this theoretical finding, simulation results confirm the proposed control strategy’s effectiveness and feasibility, demonstrating enhanced synchronization performance and robustness under these multiple, potentially conflicting constraints. Note to Practitioners—Practitioners working with nonlinear multi-agent systems facing input and output constraints should consider the adaptive tracking control approach presented in this paper. The method innovatively addresses asymmetric input saturation by developing an auxiliary system that generates compensatory signals to mitigate its negative effects on system performance. To balance input saturation and output performance constraints, a dynamic performance function is introduced, ensuring that synchronization errors stay within acceptable ranges. This approach is particularly valuable for applications like drone swarms or automated transportation systems, where synchronization and constraint adherence are safety-critical.}, archive = {J_TASE}, author = {Wei-Jie Hao and Zhao-Yi Zong and Shu-Zhen Wei and Shan-Liang Zhu and Yu-Qun Han}, doi = {10.1109/TASE.2025.3581459}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16913--16924}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tracking control of nonlinear multi-agent systems subject to multiple constraints via multi-dimensional taylor network}, volume = {22}, year = {2025}, }
• (2025). Region-aware grasping for stacked workpieces: A 6D-wise label self-generation method and robust evaluation strategy. TASE, 22, 16899--16912. (www)

  @article{ , abstract = {The high-quality datasets and generalized network model combined with robust evaluation strategies serve as a significant benchmark for developing new policies for industrial bin-picking. In this paper, we propose the concept of region-aware grasping, a sim2real cutting-edge system to generate and evaluate 6D poses for robots to pick up novel workpieces from stacked environments. It consists of Region-Aware-Dataset, a large-scale synthetic point cloud grasp dataset; and Semantic-Graspnet, a 6D-wise affordance policy that predicts full 6D grasp pose for stacked workpieces. The introduced Semantic-Graspnet transforms the 6D pose prediction problem into semantic categorization via point cloud encoding and decoding. Meanwhile, we propose a robust evaluation strategy based on pose evaluation and mechanical grasping evaluation, which enhances the robot’s grasping success rate and sorting efficiency. In real industrial tasks, the robot achieves a grasp completion rate of 91.3% in cluttered scenes and 89.2% in densely stacked scenes, demonstrating state-of-the-art results in robotic picking-and-placing applications. Note to Practitioners—This work is inspired by the robotic universal AI technique, which holds an indispensable position in industrial as bin-picking tasks. Leveraging this technology, the robot can proficiently perform picking-and-placing for diverse workpieces. The prior robot grasping policies primarily centered on maximizing the optimization of grasp detection to ensure successful grasping. Nevertheless, limitations still exist regarding efficient datasets, generalized network models, and robust evaluation strategies. Therefore, in order to effectively curtail the development cycle of robot picking-and-placing applications, this paper puts forward a novel region-aware grasping policy for stacked workpieces. We have investigated an autonomous grasp label annotation method and constructed a large-scale synthetic point cloud grasp dataset. Subsequently, a semantic grasping prediction network model, Semantic-Graspnet, has been designed for 6D-wise affordance grasp prediction. More crucially, a proposed robust grasping evaluation strategy markedly improves the robot’s grasping performance. The developed technology even surpasses the state-of-the-art algorithms in multi-object stacked grasping. The innovative sim2real schema is well-suited for the rapid development of various workpieces sorting applications by robots.}, archive = {J_TASE}, author = {Xungao Zhong and Tao Gong and Junzhi Yu and Jiaguo Luo and Chengxian Zhou and Xunyu Zhong and Qiang Liu}, doi = {10.1109/TASE.2025.3580784}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16899--16912}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Region-aware grasping for stacked workpieces: A 6D-wise label self-generation method and robust evaluation strategy}, volume = {22}, year = {2025}, }
• (2025). A novel human-in-the-loop multimodal intention fusion method for Human–Robot interaction. TASE, 22, 16888--16898. (www)

  @article{ , abstract = {Understanding human intention plays a crucial role in the research of human-robot interaction (HRI) for service robots. Although multimodal sensing methods have shown promising results in certain scenarios, the fusion strategy cannot flexibly adapt to user preference and dynamic environment. To address this limitation, this paper proposed a human-in-the-loop multimodal intention fusion (HIL-MIF) algorithm that introduces weight factors to assess the importance of each modality. By dynamically adjusting these weight factors based on user feedback, we achieved high accuracy in intention understanding, enhancing the personalization and reliability of the interaction. In addition, a multimodal service robot system was proposed that supports three interaction modalities including gaze, voice and gestures, which can be flexibly configured into different combinations. Validation experiment was performed on target object grasping tasks and collected user subjective evaluations. The experimental results demonstrate that the HIL-MIF proposed in this paper is superior to other commonly used multimodal fusion methods in terms of both accuracy and reliability. This paper paves the way for the future tri-co intelligent robot deployment in the service industry. Note to Practitioners—In the field of service robot, accurately understanding user intentions is a critical issue, as it directly affects the success of interactions. Among the various tasks of intention understanding, we face numerous challenges. For instance, unimodal approaches are susceptible to environmental factors and have very limited applicability. Currently, multimodal intention understanding methods face challenges such as the difficulty of achieving a unified representation of heterogeneous data from multiple sources and relatively low accuracy. Additionally, users have personal pReferences regarding interaction modalities, and existing multimodal methods lack research on these preferences, making it difficult to provide highly reliable and personalized interaction experiences. To address these issues, this paper proposes a human-in-the-loop multimodal intention fusion (HIL-MIF), which integrates multimodal information from gaze, gestures, and speech. It introduces weight factors to accommodate user preferences, employs combinatorial rules to fuse multimodal data, and determines the final intent from the fused information. Furthermore, a human-in-the-loop strategy is adopted to dynamically update the weight factors based on user feedback regarding the fusion results. The primary research objective of This study is to tackle the challenges of low accuracy and poor personalization in multimodal intention understanding methods, while also adapting to interactive scenarios such as meetings and low-light environments. By utilizing the multimodal intent fusion approach to control a robotic arm for object grasping, we achieved an intention understanding accuracy of 94%. This method can be easily applied to other human-robot interaction systems. In the future, we plan to apply this method to more complex human-robot interaction tasks to accomplish more sophisticated intention understanding challenges.}, archive = {J_TASE}, author = {Shuo Jiang and Wenbo Zhang and Jiahang Liu and Haonan Li and Zhipeng Wang and Yanmin Zhou and Bin He}, doi = {10.1109/TASE.2025.3580469}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16888--16898}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel human-in-the-loop multimodal intention fusion method for Human–Robot interaction}, volume = {22}, year = {2025}, }
• (2025). Current-constrained finite-time control scheme for speed regulation of PMSM systems with unmatched disturbances. TASE, 22, 16877--16887. (www)

  @article{ , abstract = {The article investigates the speed regulation of permanent magnet synchronous motor (PMSM) systems. Existing control methods of the non-cascade structure suffer from the drawbacks of unsatisfactory anti-disturbance performance and slow convergence rate when the system is affected by disturbances, especially unmatched disturbances. Meanwhile, the requirements of current constraint and fast dynamics cannot be effectively balanced in the single-loop structure of speed and current using traditional control methods such as the PID controller. Because large transient currents induced by fast dynamics may damage the hardware of the system. Therefore, a current-constrained finite-time control approach is proposed. Specifically, a robust finite-time control scheme is developed with the assistance of the improved finite-time observer technique. The proposed method is capable of actively suppressing both matched and unmatched disturbances in non-cascade control systems. Simultaneously, an effective penalty mechanism is established to incorporate a specific gain function into the designed controller. This approach restricts the q-axis current to a predefined safe range without solving an optimization problem. Finally, comparative experiment results indicate that the newly proposed finite-time control method outperforms the baseline control methods in terms of disturbance rejection, convergence rate, and current constraint. Note to Practitioners—This paper addresses practical challenges in controlling permanent magnet synchronous motor (PMSM) systems, such as ensuring robust disturbance rejection while limiting transient currents to protect the hardware. Traditional methods often fail to balance fast dynamics with current constraints, leading to inefficiencies and potential damage. To tackle these issues, the proposed finite-time control approach introduces a practical solution that achieves robust disturbance rejection while ensuring that transient currents remain within safe operational limits. The use of an improved finite-time observer allows for real-time estimation of system disturbances, while the novel penalty mechanism restricts the q-axis current, preventing hardware damage. This method is particularly suited for applications in robotics, electric vehicles, and automated manufacturing, where precise speed regulation and reliability are crucial.}, archive = {J_TASE}, author = {Huiming Wang and Zhize Zhang and Yingjie Luo and Junxiao Wang and Jiankun Sun and Yunda Yan and Xianlun Tang}, doi = {10.1109/TASE.2025.3580362}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16877--16887}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Current-constrained finite-time control scheme for speed regulation of PMSM systems with unmatched disturbances}, volume = {22}, year = {2025}, }
• (2025). Time-varying optimal sliding-mode lag formation control for high-order nonlinear multiagent systems based on reinforcement learning. TASE, 22, 16864--16876. (www)

  @article{ , abstract = {This paper considers a type of high-order nonlinear multiagent systems (MASs), and the time-varying optimal lag formation control problem for such MAS is addressed. To eliminate the influences of the unknown system dynamics and external disturbance on single agent, a neural network (NN) identifier and disturbance observer based compensation controller is developed. Then, an optimal control law to ensure the practical time-varying lag formation for the MAS is introduced by exploiting the sliding-mode control (SMC) method and reinforcement learning (RL) technique. Finally, a numerical example is given to validate the effectiveness of the proposed optimal control law. Note to Practitioners—Recently, leader-following approach has been widely utilized to investigate the time-varying formation (TVF) control problem for various MASs. In fact, time delay is unavoidable during the process of the follower tracking the leader, it is more meaningful and reasonable to address the time-varying lag formation control problem of MASs. Moreover, most of the existing TVF control results ignore the optimal performance. Considering that a balance between system performance and control resources, it is also very significant and challenging to further tackle the optimal TVF control problem of MASs. In this paper, an RL-based time-varying optimal sliding-mode lag formation control strategy for high-order nonlinear MASs is developed, and an MAS consisting of four unmanned surface vehicles (USVs) is utilized to validate the effectiveness of the presented control strategy.}, archive = {J_TASE}, author = {Qinhong Fu and Jin-Liang Wang and Tingwen Huang}, doi = {10.1109/TASE.2025.3579744}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16864--16876}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time-varying optimal sliding-mode lag formation control for high-order nonlinear multiagent systems based on reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Fault diagnosis of discrete event systems modeled by time-interval automaton. TASE, 22, 16851--16863. (www)

  @article{ , abstract = {We address in this paper the problem of diagnosability of time-interval discrete event systems (TIDES), a class of discrete event systems that has a single clock structure and whose event occurrences take place within a time interval after the previous event occurrence. The idea behind the diagnosability of TIDES is to leverage time information to distinguish fault traces from non-faulty ones, increasing the accuracy of the fault diagnosis system. For this purpose, we use a recently proposed timed model called time-interval automaton (TIA), and, based on this model, we present necessary and sufficient conditions for the diagnosability of discrete event systems modeled by TIA and an algorithm for its verification. Examples illustrate all of the results presented in the paper. Note to Practitioners—Unlike the usual assumption on discrete-event systems, whose evolution is dictated by the asynchronous occurrence of events, in practice, there is always some knowledge on the time associated with event occurrence; e.g., in a manufacturing system, it is also possible to have an estimation on the time the parts take to go from the beginning of a conveyor belt up to different sensors located along the conveyor. In this paper, the time information is given by the time interval within which an event must occur after the system reaches some state, and we take advantage of this information to improve the accuracy of the fault diagnosis system.}, archive = {J_TASE}, author = {João C. Basilio and Christiano H. Rezende and Gustavo S. Viana}, doi = {10.1109/TASE.2025.3579140}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16851--16863}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault diagnosis of discrete event systems modeled by time-interval automaton}, volume = {22}, year = {2025}, }
• (2025). Practical prescribed-time bipartite time-varying formation control for multiagent systems with adaptive self-triggered. TASE, 22, 16838--16850. (www)

  @article{ , abstract = {This paper investigates the problem of practical prescribed-time fuzzy bipartite time-varying formation control for nonlinear multiagent systems. To achieve practical prescribed-time control, a bounded and continuous prescribed-time evolution function is constructed, and a new stability lemma is derived, proving that the proposed controller avoids singularities and ensures synchronization errors converge to a neighborhood independent of initial conditions within the prescribed time. Considering cooperative-competitive interactions among the agents, a fuzzy bipartite time-varying formation controller is designed to achieve formation control within a specified time while keeping all closed-loop signals bounded. Furthermore, to reduce communication costs, an adaptive self-triggered mechanism is introduced. This mechanism maps the prescribed time to a dynamic triggering function, enabling the controller to adjust triggering conditions based on the user-defined convergence time, thereby balancing speed and communication costs. Finally, the effectiveness of the proposed method is validated through a simulation case involving five uncrewed ground vehicles.Note to Practitioners—This study addresses the formation control problem in nonlinear multiagent systems and proposes a practical prescribed-time bipartite formation control strategy, particularly suited for complex cooperative and competitive relationships. Traditional methods often face issues related to communication overhead and dynamic uncertainties, while the proposed approach significantly optimizes system performance by introducing a prescribed-time evolution function and an adaptive self-triggered mechanism, particularly in scenarios with limited communication resources. The self-triggered mechanism enables the system to adaptively adjust the triggering conditions based on the user-defined convergence time, thereby ensuring rapid convergence while reducing unnecessary communication. This approach is particularly applicable to scenarios that require fast responses and low communication overhead, offering broad industrial application potential.}, archive = {J_TASE}, author = {Shuxing Xuan and Hongjing Liang and Tieshan Li and Chenguang Yang}, doi = {10.1109/TASE.2025.3579029}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16838--16850}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical prescribed-time bipartite time-varying formation control for multiagent systems with adaptive self-triggered}, volume = {22}, year = {2025}, }
• (2025). Fault location in power distribution networks with massive missing data: A graph-based imputation and contrastive learning approach. TASE, 22, 16825--16837. (www)

  @article{ , abstract = {Accurate fault location is essential for reliable operation of distribution networks, yet it remains challenging due to incomplete data caused by insufficient sensor deployment, sensor malfunctions, and communication failures during fault events. Existing Graph Neural Network (GNN)-based methods often assume complete data availability or resort to zero-fill imputation. This work proposes a two-stage fault location pipeline for data deficiency environments. A training-free graph-based imputation algorithm (GDIA) is proposed, capable of restoring missing values by directly minimizing graph Dirichlet energy without relying on prior data distribution. Additionally, a theoretical bound is established, linking changes in Dirichlet energy to imputation error. Subsequently, GNN models are employed to perform fault location using the imputed data. To address class imbalance arising from data deficiencies, a graph contrastive learning technique is integrated, further enhancing fault location accuracy. Experimental results demonstrate the proposed method’s robustness and effectiveness, highlighting its potential to enhance fault location accuracy in practical distribution network applications. Note to Practitioners—This paper addresses the critical challenge of fault location in distribution networks under massive missing data. Traditional methods relying on Graph Neural Networks (GNNs) often assume complete data availability, making them impractical for real-world scenarios. The paper suggests a novel way to address the issue by proposing a training-free graph-based imputation algorithm (GDIA) for data imputation, following the graph constrastive learning framework (GCL) for fault location considering class imbalance. GDIA efficiently reconstructs missing data by minimizing the Dirichlet energy on the network graph, ensuring structural consistency while avoiding reliance on prior data distributions. Theoretical analysis shows that higher Dirichlet energy correlates with reduced imputation accuracy, guiding practical implementation. For data imputation, since the low deployment of sensors and traditional deep learning based imputation methods cannot perform without the prior distribution, this paper proposed GDIA, which minimize dirichlet energy on graph to impute missing data and theoretically prove its viability that when graph Dirichlet energy becomes larger, the imputation accuracy would correspondly becomes lower. For fault location, GCL tackles class imbalance presented in fault datasets by training the GNN model to recognize the nodes that exhibit similar patterns to their neighbors, while faults occurence disrupt these patterns. By amplifying discrepancies between faulty and normal nodes, GCL can address the issue and enhance the fault location accuracy. Validated on the IEEE 33-bus system under massive data missing ( $\geq 60$ %), our GDIA-GCL framework achieves significantly higher fault location accuracy compared to existing methods. This approach enables utilities to enhance grid reliability without costly sensor upgrades, as it operates effectively on existing infrastructure. While the current method assumes homogeneous network, ongoing work focuses on adapting to heterogeneous grid reconfigurations and scaling to larger networks.}, archive = {J_TASE}, author = {Luliang Zhang and Dingyan Hua and Tianyao Ji and Tong Qian and Jian Wang}, doi = {10.1109/TASE.2025.3580658}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16825--16837}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault location in power distribution networks with massive missing data: A graph-based imputation and contrastive learning approach}, volume = {22}, year = {2025}, }
• (2025). Robotic motion optimization for tactile object recognition learned from human behaviors. TASE, 22, 16814--16824. (www)

  @article{ , abstract = {Touch is one of the most important human senses. With the development of artificial intelligence, an increasing number of scholars are investigating how robots can be endowed with the sense of touch. Tactile information acquisition relies heavily on action patterns, as touch is generated by direct contact. This paper proposes a robot active tactile perception framework inspired by human behaviors for object recognition. Human recognition experiments were designed to explore the action patterns used to recognize unknown objects. The analysis of these experiments revealed three primary action patterns: pressing, sliding, and rubbing. Robotic motions were conducted to collect tactile data from the same objects using these actions. Given the limited number of available datasets, which restricts the application of active perception learning methods inspired by human behavior patterns, this work aims to address this challenge by developing a novel framework. A tactile convolutional neural network was established for object recognition using the collected data. Additionally, a human-behavior-pattern-inspired tactile feedback control unit was integrated to optimize robotic motions and improve recognition accuracy. The proposed method achieved a recognition accuracy of 97.5%, surpassing the human experiment. This work provides valuable insights into the active tactile perception of robots, offering a reference for future research in this field. Note to Practitioners—This paper was motivated to propose a novel perspective for enhancing the tactile object recognition capabilities of robots on limited datasets. Current approaches focused on using neural networks to extract features from data for classification. This paper posits that robots should not only learn from tactile data but also to learn the methods of data collection, ultimately being able to autonomously determine a data collection motion pattern that is suitable for themselves. In this study, the robot learns motion patterns for data collection from human recognition experiments, and then utilizes the action feedback optimization system proposed in this paper to dynamically confirm its own suitable motion patterns from human identification experiences and previous training experiences. Finally, it can maintain a relatively stable action ratio and a high recognition rate. Preliminary experiments of this method indicate that this perspective is beneficial for improving the tactile recognition performance of robots, although the selected motion patterns are relatively simple. In future experiments, we plan to design complex tactile sensors and incorporate more complex tactile actions into the recognition process, making humanoid robots more human-like in logic.}, archive = {J_TASE}, author = {Yafei Wang and Yanmin Zhou and Ping Lu and Zhipeng Wang and Bin He}, doi = {10.1109/TASE.2025.3580573}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16814--16824}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic motion optimization for tactile object recognition learned from human behaviors}, volume = {22}, year = {2025}, }
• (2025). Two-step event-triggered data driven model predictive control for trajectory tracking of unmanned surface vessel under environmental disturbances. TASE, 22, 16801--16813. (www)

  @article{ , abstract = {Data driven model predictive control (DMPC) has become an effective trajectory tracking approach for unmanned surface vessel (USV). However, its offline-trained prediction models and event-triggered (ET) mechanisms result in reduced control accuracy and efficiency under time-varying ocean disturbances. To effectively improve the control accuracy and efficiency of the DMPC method under environmental influences, a two-step ET DMPC (TETDMPC) algorithm is proposed. Traditional mechanism prediction models fail to effectively account for time-varying USV dynamics, which can degrade the accuracy of control algorithms. To enhance the control accuracy of the algorithm, a v-support vector regression (vSVR)-based data driven model is introduced into the model predictive control algorithm to replace mechanism prediction model. However, the rolling horizon calculations of the DMPC and the online updates of the vSVR suffer from extensive computation. To improve the control efficiency and enable vSVR model online updates, a two-step ET mechanism is designed. The performance of the online vSVR model depends on the quality of the training data. Therefore, a spatiotemporal data updating algorithm is also developed to construct the model training dataset online. The trajectory tracking performance of the proposed TETDMPC algorithm is validated based on the CyberShip II and Dolphin 1 USVs, respectively. The results demonstrate that the TETDMPC algorithm has desirable control performance. Note to Practitioners—Our research has achieved a breakthrough. A two-step event-triggered data driven model predictive control (TETDMPC) algorithm has been specifically designed to improve the control efficiency, accuracy and reliability of the unmanned surface vehicle (USV) systems under environmental disturbances. This innovation is of great significance, especially for the trajectory tracking of USVs under environmental disturbances. Since traditional model predictive control (MPC) has difficulty taking into account the decline in control performance caused by inaccurate models, and data driven MPC (DMPC) methods are unable to maintain good control efficiency under environmental disturbances, the need for effectively saving computational resources and maintaining good control accuracy has emerged for DMPC. TETDMPC has demonstrated its ability to solve these problems in this paper, making it a powerful tool for improving control performance. However, TETDMPC does not consider how to perform effective obstacle avoidance operations. In future research, we will address the integration issue of the TETDMPC and obstacle avoidance algorithms.}, archive = {J_TASE}, author = {Lichao Jiang and Chuanrui Wang and Xiaobing Shang and Zhi Zhang}, doi = {10.1109/TASE.2025.3579396}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16801--16813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Two-step event-triggered data driven model predictive control for trajectory tracking of unmanned surface vessel under environmental disturbances}, volume = {22}, year = {2025}, }
• (2025). Optimal deception attacks on remote state estimation under interval constraints. TASE, 22, 16790--16800. (www)

  @article{ , abstract = {In this paper, the problem of optimal deception attacks on remote state estimation is studied, where an interval $\chi ^{2}$ detector is set up to verify the validity of the data packets received by the remote state estimator. Malicious attackers are not only able to intercept the original measurements transmitted on the wireless network, but also obtain side information about the system states sensed by an extra sensor. First, By fusing these two types of information, an innovation-based deception attack model with combined information is proposed. Then, we present a novel stealthiness constraint and derive the covariance at the final instant of the attack interval to characterize the attack performance. Furthermore, the closed-form optimal deception attack schemes are obtained by utilizing the Lagrange method. Finally, numerical simulations and experiment results confirm the effectiveness of the proposed attack scheme. Note to Practitioners—Kalman filter is widely used in Cyber-physical systems such as smart grids, smart transportation, and the Internet of Vehicles for trajectory tracking and data fusion. However, the open communication environments in these applications introduce information security risks to the system, which cannot be ignored. To lay the groundwork for formulating more targeted defense measures, this paper examines the optimal deception attack scheme from the attacker’s perspective. Specifically, considering a scenario where the attacker can obtain system side information in real time, this paper investigates an attack strategy using combined information under interval concealment constraints for the $\chi ^{2}$ interval anomaly detector. The closed-form solution for the optimal deception attack scheme for both stable and unstable systems is obtained. Compared to previous methods that utilize optimization problem toolboxes, the approach proposed in this paper improves real-time performance and provides a feasible strategy for industrial applications.}, archive = {J_TASE}, author = {Hongbo Yuan and Wen Yang and Wenjie Ding and Yun Liu and Yang Tang}, doi = {10.1109/TASE.2025.3579380}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16790--16800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal deception attacks on remote state estimation under interval constraints}, volume = {22}, year = {2025}, }
• (2025). Robust forecasting-aided state estimation of active distribution network with multiple distributed generators. TASE, 22, 16780--16789. (www)

  @article{ , abstract = {This paper develops a robust variational Bayesian unscented Kalman filter (RVBUKF) to track the state of active distribution networks with distributed energy resources. First, a spherical simplex-based unscented transform (UT) strategy is used to select Sigma points to reduce the computational burden. Subsequently, based on unscented Kalman filter and variational Bayesian theory, a variational update form of the covariance matrix is derived by using the inverse Wishart probability density function. To mitigate the impact of outliers, the Gaussian process regression model trained by historical data is used to synchronously predict observation values to replace abnormal measurements. Finally, simulation experiments are conducted on the distribution system containing distributed generations (DGs). The numerical results demonstrate that the developed method can accurately track state changes in uncertain scenarios while reducing the computational burden of UT by almost half. Note to Practitioners—In practical power systems, due to the randomness and fluctuation of distributed generation output, operating states of the active distribution network may frequently change, which places new demands on the dynamic state estimation of active distribution networks. Therefore, in this paper, a new forecasting-aided state estimation method that combines robust Kalman filtering with machine learning techniques based on Gaussian process regression is proposed. The effectiveness of the proposed forecasting-aided state estimation was verified in the constructed active distribution network. Preliminary simulation experiments have shown that the number of Sigma points selected by SSUT is almost 50% of that of traditional UT, and the proposed method is capable of accurately tracking the dynamic changes of active distribution network states in various uncertain scenarios. In future research, we will address the problem of active distribution network forecasting-aided state estimation based on non-parametric modeling and attempt to use actual power grid data for testing.}, archive = {J_TASE}, author = {Dongchen Hou and Yonghui Sun and Kang Sun and Venkata Dinavahi and Yi Wang}, doi = {10.1109/TASE.2025.3578493}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16780--16789}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust forecasting-aided state estimation of active distribution network with multiple distributed generators}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time consensus of nonlinear multi-agent systems by dynamic event-triggered and self-triggered protocol. TASE, 22, 16768--16779. (www)

  @article{ , abstract = {This article proposes the dynamic event-triggered and self-triggered protocols to achieve the prescribed-time consensus of a class of nonlinear multi-agent systems with a trade-off performance. Firstly, a dynamic event-triggered control protocol and a time-varying parameter are jointly designed to accomplish the prescribed-time consensus. Secondly, a dynamic self-triggered control protocol is further designed to avoid continuous monitoring of the system states. The designed control protocols both can ensure the avoidance of the Zeno phenomenon and allow for dynamic adjustment between the specified time and the inter-event time. Specially, the minimum inter-event time can be designed as an arbitrarily large bounded constant when every agent is an integrator system. Finally, an example system is utilized to demonstrate the effectiveness of the designed algorithms. Note to Practitioners—Since this type of nonlinear system can describe the nonlinear characteristics in the system dynamics, some complex practical systems such as single-link manipulator systems and electrical subsystem dynamics system can be modeled as such systems. As control systems demand increasing precision, prescribed-time control, which allows for arbitrarily predefined convergence times, has been applied in many time-critical fields, such as missile guidance and emergency braking. On the other hand, in engineering applications, multi-agent systems are often constrained by limited communication resources. To address this issue, this paper introduces a dynamic event-triggered mechanism. Compared with general control strategies, the dynamic event-triggered mechanism can effectively conserve communication resources. Additionally, to avoid continuous monitoring of system states, a corresponding self-triggered mechanism is also designed.}, archive = {J_TASE}, author = {Meilin Li and Kai Zhang and Yang Liu and Fazhi Song and Tieshan Li}, doi = {10.1109/TASE.2025.3578418}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16768--16779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time consensus of nonlinear multi-agent systems by dynamic event-triggered and self-triggered protocol}, volume = {22}, year = {2025}, }
• (2025). Learning autonomous surgical irrigation and suction with the da vinci research kit using reinforcement learning. TASE, 22, 16753--16767. (www)

  @article{ , abstract = {The irrigation-suction process is a common procedure to rinse and clean up the surgical field in minimally invasive surgery (MIS). In this process, surgeons first irrigate liquid, typically saline, into the surgical scene for rinsing and diluting the contaminant, and then suction the liquid out of the surgical field. While recent advances have shown promising results in the application of reinforcement learning (RL) for automating surgical subtasks, fewer studies have explored the automation of fluid-related tasks. In this work, we explore the automation of both steps in the irrigation-suction procedure and train two vision-based RL agents to complete irrigation and suction autonomously. To achieve this, a platform is developed for creating simulated surgical robot learning environments and for training agents, and two simulated learning environments are built for irrigation and suction with visually plausible fluid rendering capabilities. With techniques such as domain randomization (DR) and imitation learning, two agents are trained in the simulator and transferred to the real world. Individual evaluations of both agents show satisfactory real-world results. With an initial amount of around 5 grams of contaminants, the irrigation agent ultimately achieved an average of 2.21 grams remaining after a manual suction. As a comparison, fully manual operation by a human results in 1.90 grams remaining. The suction agent achieved 2.64 and 2.24 grams of liquid remaining across two trial groups with more than 20 and 30 grams of initial liquid in the container. Fully autonomous irrigation-suction trials reduce the contaminant in the container from around 5 grams to an average of 2.42 grams, although yielding a higher total weight remaining (4.40) due to residual liquid not suctioned. Further information about the project is available at https://tbs-ualberta.github.io/CRESSim/ Note to Practitioners—The irrigation-suction process is a surgical procedure for rinsing and cleaning surgical fields. This work tackles automating the process to reduce the workload for surgeons. Our approach is based on two customized simulation environments that can simulate the irrigation and suction process realistically. Two autonomous agents are trained using robot learning approaches in the environments for completing irrigation and suction, respectively, and then transferred to the real world. The agents autonomously control the surgical robot by interpreting the images captured from an RGB camera and the robot’s current state, and generate joint movements of the robot. This approach has been tested in physical settings, showing promising results in terms of the individual and combined performance of the agents in executing the full irrigation-suction process. Future work is needed to extend the approach to more practical surgical settings, evaluate the performance under diverse conditions, and enhance integration with existing surgical robot platforms.}, archive = {J_TASE}, author = {Yafei Ou and Mahdi Tavakoli}, doi = {10.1109/TASE.2025.3576731}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16753--16767}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning autonomous surgical irrigation and suction with the da vinci research kit using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Multi-target stochastic configuration network model based on collaborative optimization and its applications. TASE, 22, 16741--16752. (www)

  @article{ , abstract = {A multi-target model can simultaneously monitor the changing trends of multiple related parameters, which plays a crucial role in achieving intelligent perception and decision-making in the production process. However, efficiently learning the non-linearity between input and output variables and exploring the correlations between multiple targets to construct high-performance monitoring models have always been challenging. This paper proposes a multi-target stochastic configuration network (SCN) model from the perspectives of network structure design and parameter group sparse constraints. This model assigns hidden layer input weights and biases using a weighted supervisory mechanism and adds a latent layer to the network structure. It facilitates multi-target balanced learning while exploiting inter-target correlations to improve modeling accuracy. The output weights of the hidden and latent layers are collaboratively constrained by constructing a generalized sparse matrix elastic net to induce their group sparsification. An alternating optimization algorithm based on matrix decomposition is designed to iteratively solve the model parameters. The performance of the proposed model is experimentally validated on benchmark multi-target datasets and historical data of municipal solid waste incineration process. The results demonstrate that the proposed model achieves superior accuracy and exhibits potential for multi-target modeling applications. Note to Practitioners—This paper proposes a high-performance multi-target SCN model for modeling multiple related parameters in the production process. Through the collaborative optimization of the network structure and parameters, this model can not only capture the non-linear relationship between inputs and outputs but also explore the potential correlations between multiple targets to improve the modeling accuracy. Moreover, we have designed an alternating optimization algorithm to determine the model parameters. Experimental results demonstrate that the proposed multi-target SCN model exhibits high accuracy and can be applied to tasks such as equipment performance monitoring, product quality analysis, and energy consumption control.}, archive = {J_TASE}, author = {Aijun Yan and Kaicheng Hu}, doi = {10.1109/TASE.2025.3579737}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16741--16752}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-target stochastic configuration network model based on collaborative optimization and its applications}, volume = {22}, year = {2025}, }
• (2025). Autonomous and adaptive role selection for multi-robot collaborative area search based on deep reinforcement learning. TASE, 22, 16728--16740. (www)

  @article{ , abstract = {In the tasks of multi-robot collaborative area search, we propose the unified approach for simultaneous mapping for sensing more targets (exploration) while searching and locating the targets (coverage). Specifically, we implement a hierarchical multi-agent reinforcement learning algorithm to decouple task planning from task execution. The role concept is integrated into the upper-level task planning for role selection, which enables robots to learn the role based on the state status from the upper-view. Besides, an intelligent role switching mechanism enables the role selection module to function between two timesteps, promoting both exploration and coverage interchangeably. Then, the primitive policy learns how to plan based on their assigned roles and local observation for sub-task execution. The well-designed experiments show the scalability and generalization of our method compared with state-of-the-art approaches in scenes with varying complexity and numbers of robots. Our code is released at https://github.com/linaug/Role_selection Note to Practitioners—In the realm of area search tasks, the collaboration among multiple robots in a multi-robot system brings about a reduction in labor costs and an enhancement in task completion efficiency. Recognizing this, the development of an intelligent multi-robot collaborative decision-making algorithm becomes crucial-especially within complex and diverse environments, ensuring better adaptability to various real-world scenarios. This paper introduces a hierarchical framework designed for the simultaneous mapping and searching of targets in multi-robot area search tasks. The upper-level role selection framework employs adaptive role learning through deep reinforcement learning, while the low-level primitive policy focuses on planning actions based on the designated roles. As a result, the proposed approach exhibits notable advantages in terms of both the percentage and efficiency of task completion, particularly evident in well-designed scenes with varying complexities. We believe this framework holds significant potential for real-world applications, addressing challenges in multi-robot collaboration across diverse environments.}, archive = {J_TASE}, author = {Lina Zhu and Jiyu Cheng and Hao Zhang and Zhichao Cui and Wei Zhang and Yuehu Liu}, doi = {10.1109/TASE.2025.3579736}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16728--16740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous and adaptive role selection for multi-robot collaborative area search based on deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Holistic construction automation with modular robots: From high-level task specification to execution. TASE, 22, 16716--16727. (www)

  @article{ , abstract = {In situ robotic automation in construction is challenging due to constantly changing environments, a shortage of robotic experts, and a lack of standardized frameworks bridging robotics and construction practices. This work proposes a holistic framework for construction task specification, optimization of robot morphology, and mission execution using a mobile modular reconfigurable robot. Users can specify and monitor the desired robot behavior through a graphical interface. In contrast to existing, monolithic solutions, we automatically identify a new task-tailored robot for every task by integrating Building Information Modeling (BIM). Our framework leverages modular robot components that enable the fast adaption of robot hardware to the specific demands of the construction task. Other than previous works on modular robot optimization, we consider multiple competing objectives, which allow us to explicitly model the challenges of real-world transfer, such as calibration errors. We demonstrate our framework in simulation by optimizing robots for drilling and spray painting. Finally, experimental validation demonstrates that our approach robustly enables the autonomous execution of robotic drilling. Note to Practitioners—Construction sites are unpredictable and constantly changing, making automation difficult. To tackle this, we propose a modular robotic system that can adapt to specific tasks. In this work, we address the challenge of automating robotic drilling and introduce a general framework that also supports other construction tasks. We use Building Information Modelling (BIM) data to automatically determine the best robot configuration by optimizing factors such as setup time and robustness to positioning errors. Through an intuitive interface, workers can specify tasks, after which they assemble the robot following the provided instructions. The result of this work is a robotic system that can be deployed on diverse construction sites while requiring only readily available BIM data and minimal user inputs. Our approach enables the same hardware to be deployed across various scenarios while accounting for site-specific challenges. A core innovation lies in combining modular robotics with automated optimization, task planning, and motion planning to minimize the need for user intervention while delivering robots specifically tailored to each task. To achieve this, we extend methods for modular robot design to account for multiple conflicting objectives in the optimization process. Experimental results demonstrate the ability of the system to perform autonomous drilling with high precision. Beyond drilling, this framework has the potential to support a wide range of applications, and we show its applicability to spray painting in simulation. Overall, it offers a practical and adaptable solution for diverse construction tasks and sites.}, archive = {J_TASE}, author = {Jonathan Külz and Michael Terzer and Marco Magri and Andrea Giusti and Matthias Althoff}, doi = {10.1109/TASE.2025.3579720}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16716--16727}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Holistic construction automation with modular robots: From high-level task specification to execution}, volume = {22}, year = {2025}, }
• (2025). Environment-adaptive motion planning via reinforcement learning-based trajectory optimization. TASE, 22, 16704--16715. (www)

  @article{ , abstract = {This paper proposes a novel environment-adaptive motion planning framework for mobile robots, which utilizes deep reinforcement learning to dynamically adjust optimization objectives according to various environmental and robot-ego characteristics, greatly enhancing the adaptability and robustness compared to existing motion planning strategies. Our approach features a two-stage trajectory optimization algorithm that optimizes for smoothness, safety, and efficiency-elements critical in practical applications. Firstly, we propose a reinforcement learning algorithm that dynamically adjusts optimization objectives based on the environmental context, which carefully encodes the environment, coarse initial path and robot information into the observation space. Additionally, two techniques are designed to reduce the sim-to-real gap: 1) integrating classical optimization framework as the motion planning backbone; 2) using low-dimensional input in the learning component, which minimizes discrepancies between simulated and real-world conditions. With the hybrid strategy, not only the interpretability and stability of the classical motion planning pipeline is preserved, but also the adaptive capability of emerging DRL techniques is fully utilized. The efficacy of the proposed solution is demonstrated through extensive simulations and real-world experiments, showcasing superior performance in terms of safety and efficiency across various testing scenarios. Especially in unknown and cluttered real office environments, our approach significantly improve safety (17% increase in success rate) and efficiency (29% reduction in time cost), compared to the popular traditional methods. (Supplementary video link: https://youtu.be/ph0pDGpI864) Note to Practitioners—This paper is motivated by the need for reliable and adaptable motion planning in various environments where mobile robots must navigate safely and efficiently. Traditional motion planning methods often struggle with the adaptability required in these settings due to their fixed optimization objectives and lack of flexibility. On the other hand, end-to-end deep reinforcement learning (DRL) approaches can adapt to changing environments but may suffer from catastrophic failures and a lack of interpretability. This paper introduces a hybrid approach that combines DRL with classical optimization techniques. The proposed method allows the robot to adjust its optimization objectives dynamically, enhancing its ability to handle unforeseen obstacles and changes in the environment. This is particularly beneficial in applications where robots must operate autonomously for extended periods without human intervention. The hybrid approach presented has shown promising results in simulations and real-world tests, demonstrating both stability and reliability.}, archive = {J_TASE}, author = {Zhejin Zhu and Runhua Wang and Yisong Wang and Yaonan Wang and Xuebo Zhang}, doi = {10.1109/TASE.2025.3579573}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16704--16715}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Environment-adaptive motion planning via reinforcement learning-based trajectory optimization}, volume = {22}, year = {2025}, }
• (2025). Elliptical magnetic nanoparticle swarm formation by an artificial neural network-based rotating magnetic field. TASE, 22, 16692--16703. (www)

  @article{ , abstract = {Biomedical applications of magnetic nanoparticle (MNP) swarm control rely on precise manipulation of the swarm using magnetic fields. This approach improves the accessibility and maneuverability of the swarm and enhances targeted therapies and drug delivery. In this study, we applied artificial intelligence to control the shape of elliptical MNP swarms by adjusting the parameters of a rotating magnetic field (RMF). The resulting RMF controlled the major axis and slant angle of the ellipse-shaped swarm. We constructed a dataset of diverse swarm shapes and corresponding RMF parameters, and applied it to train an artificial neural network (ANN). The trained ANN was used to predict the experimental RMF parameters required to achieve the desired shapes of elliptical MNP swarms. After ANN training, differences between the predicted and actual root mean square error (RMSE) values were evaluated to verify the model. The RMSEs of the major and minor axis lengths and slant angle were 0.21 mm, 0.06 mm, and 6.01°, respectively. This simple, inference-based approach to swarm formation yielded higher swarm formation accuracy and allowed fine control of the swarm. Note to Practitioners—Magnetic nanoparticle swarms have shown good potential for precise and non-invasive targeted drug delivery and therapies. The motivation for this work originates from the need for a faster and more efficient process of generating magnetic nanoparticle swarms and their shape control. Currently used approaches for swarm formation rely on generating a table of rotating field parameters corresponding to different swarm shapes and size or automatic closed loop controllers. These processes are slow and can result in collapse of swarm shape due to instantaneous change in the rotating magnetic field parameters. To tackle these drawbacks, we propose a machine learning based model for swarm formation and shape control. The trained artificial neural network not only enabled faster swarm formation due to its fast inference time but also allowed for easier shape control, as the user could select any elliptical swarm shape parameter to obtain the required rotating magnetic field parameters.}, archive = {J_TASE}, author = {Jongwon Park and Uijeong Lee and Yunjae Hwang and Dahye Lee and Sarmad Ahmad Abbasi and Hongsoo Choi}, doi = {10.1109/TASE.2025.3579309}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16692--16703}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Elliptical magnetic nanoparticle swarm formation by an artificial neural network-based rotating magnetic field}, volume = {22}, year = {2025}, }
• (2025). Adaptive sliding mode synchronous control for complex networks with amplify-and-forward relays. TASE, 22, 16682--16691. (www)

  @article{ , abstract = {This paper aims to the remote synchronous control for complex networks with amplify-and-forward (AF) relays. Firstly, the AF relay dynamic with random noise is integrated into the control signal transmission. Secondly, the complex networks incorporating an AF relay is converted into the sliding control mode (SMC), which robustly addresses unknown disturbances among the nodes to achieve synchronous states. Thirdly, an adaptive gain mechanism based on the equivalent value of sign function of SMC is proposed to estimate the disturbance amplitude, hence mitigating the chattering of SMC. Finally, the illustrative simulation demonstrated the effectiveness of proposed methodology. Note to Practitioners—This paper is motivated by the synchronous control of the hydraulic supports in underground coal mining, where hundreds of hydraulic supports form a complex network, often with the centralized controller situated remotely from emulsion pump station. In actual engineering application, the incorporation of amplify-and-forward (AF) relays in the control signal transmission process is advisable. However, the AF relays would introduce noise and exacerbates system uncertainty. Consequently, the primary objective of this paper is to design a robust controller tailored for this complex network with AF relays. The existing research seldom addresses this particular challenge. To tackle this issue, this paper design a sliding mode controller (SMC). Within the framework of SMC, the transmission and AF relaying processes are transformed into a sliding variable dynamic; an adaptive mechanism for sign function is proposed to mitigate the SMC chattering. We mathematically prove the reachability of the sliding mode, the asymptotic stability of the control system, and the convergence of the adaptive gain. Preliminary simulation experiments demonstrate the effectiveness of our proposed methodology. In future research, we aim to address the constraint faced by some transmission channels of AF relays in the synchronous control of complex networks.}, archive = {J_TASE}, author = {Yang Yi and Qian Wei and Wang Tian and Wang Keping}, doi = {10.1109/TASE.2025.3578652}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16682--16691}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive sliding mode synchronous control for complex networks with amplify-and-forward relays}, volume = {22}, year = {2025}, }
• (2025). Coding-decoding protocol-based data-driven adaptive sliding mode control under energy-constrained DoS attacks. TASE, 22, 16668--16681. (www)

  @article{ , abstract = {The security tracking control problem is addressed for nonlinear discrete-time networked control systems (NCSs) subject to energy-constrained denial-of-service (EC-DoS) attacks. A probabilistic quantization-based coding-decoding protocol (CDP) is designed to mitigate bandwidth limitations and enhance transmission security. Then, a CDP-based data-driven adaptive sliding mode control (DDASMC) method is presented on the basis of a linear data model (LDM) of the nonlinear NCSs. A detection algorithm is designed to detect EC-DoS attacks so that the proposed CDP-based DDASMC can automatically switch between two cases with and without EC-DoS attacks, respectively. In the duration that the attack does not occur, an integral sliding function and a reaching law are designed to achieve fast convergence while ensuring a good control performance. In the duration that the attack occurs, a prediction mechanism is introduced to construct a predictive LDM for estimating future control information. Further, a sliding mode predictive control is developed to address the EC-DoS attacks and enhance the control performance. All the algorithms of the proposed DDASMC are computed only using the input-output (I/O) data instead of any other information of the physical models. Simulation study validate the results. Note to Practitioners—NCSs have been extensively applied in numerous fields, such as smart grid and aerospace. However, the open network environment provides attackers with many opportunities for cyber-attacks, resulting in the deterioration of control performance. Accordingly, it is urgent to analyze the security problems of NCSs that are vulnerable to cyber-attacks. This study considers the security tracking control problem of NCSs under EC-DoS attacks. A probabilistic quantization-based CDP is introduced for the feedback channel to reduce the bandwidth of communication networks and improve the security of data transmission by compressing the data before transmission. Then, a detection mechanism is proposed to determine whether NCSs are attacked, and a DDASMC with a switching mechanism is developed to address both cases with and without EC-DoS attacks. The presented control method is data-driven and includes an adaptive switching mechanism that not only saves resources but is also more suitable for practical applications.}, archive = {J_TASE}, author = {Lina Chang and Ronghu Chi and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3577992}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16668--16681}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Coding-decoding protocol-based data-driven adaptive sliding mode control under energy-constrained DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Saturation function-based finite-time synchronization control for fractional-order coupled neural networks. TASE, 22, 16658--16667. (www)

  @article{ , abstract = {In the existing research on finite-time synchronization (FTS) control for fractional-order coupled neural networks (FOCNNs), signum function plays a crucial role in controller design. The discontinuity of the signum function causes the chattering phenomenon to worsen the performance of controlled system. In this paper, a saturation function is utilized instead of signum function in controller design, overcoming the shortcomings of previous control schemes. Due to the introduction of the saturation function, the system exhibits different dynamic behaviors within and outside the boundary of the saturation function. To further analyze this effect, the two-stage fractional-order nonlinear differential inequalities (TFNDIs) are established, which provides an effective tool for handling saturation function-based FTS control for FOCNNs. At last, the validity of proposed theoretical results is demonstrated through numerical simulations, which show that the chattering has been significantly suppressed.Note to Practitioners—The FTS problems have practical applications in control and engineering, such as multi-robot collaboration, microgrid synchronization and satellite formation flight, etc. However, there is a fact that signum function may cause chattering phenomenon in finite-time control, which makes it difficult to implement in the engineering application of high precision systems. In this paper, a saturation function-based finite-time control scheme has been constructed for FOCNNs to suppress the chattering, but it is difficult to apply the stability method based on integer-order systems directly to fractional-order ones. Therefore, we propose a new TFNDIs method, and give a rigorous proof for FTS. The superiority of proposed method has been demonstrated in the numerical example.}, archive = {J_TASE}, author = {Zhanshan Wang and Bibo Zheng}, doi = {10.1109/TASE.2025.3577901}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16658--16667}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Saturation function-based finite-time synchronization control for fractional-order coupled neural networks}, volume = {22}, year = {2025}, }
• (2025). Concurrent analysis of local structure and global temporality: Observable and controllable dynamic graph for RUL prediction and evolution illustration. TASE, 22, 16644--16657. (www)

  @article{ , abstract = {The accurate prediction of bearing remaining useful life (RUL) is crucial for industrial safety and cost reduction. However, existing deep learning methods often lack interpretability and struggle to integrate local-global and structural-temporal information, limiting their effectiveness. To address this issue, an observable and controllable RUL prediction method is proposed, which visualizes the bearing degradation process through spatio-temporal dynamic graph (STDG), and fully extracts structural information and temporal features to achieve precise RUL prediction. First, canonical variable fluctuation analysis (CVFA) is employed to maximize past-future data correlation, enabling information distillation by analyzing static deviations and dynamic fluctuations. Second, restraint graph attention network (RGAT) is proposed to extract local structural information, generating dynamic factors that feedback into the spatio-temporal static graph (STSG) to construct the STDG for observation and control of experts. Thirdly, to address degradation variability and long sequence dependency, the reversible instance normalization Transformer (RevIN-Former) explores the temporal evolution of the STDG from a global perspective, achieving accurate RUL prediction. Finally, the effectiveness of the proposed method is validated through three accelerated life experiments on bearings, which reduces the RUL prediction error by 88% compared to other methods. Note to Practitioners—Bearings are critical components of rotating machinery. Accurate prediction of their RUL is essential for ensuring industrial safety and reducing costs. Existing deep learning methods face challenges related to incomplete degradation information extraction and poor interpretability. To address these challenges, this paper proposes a deep learning approach that effectively leverages local structural and global temporal information through spatio-temporal dynamic graphs to comprehensively learn and visualize the bearing degradation process. The proposed method has been validated on three public datasets, and the results indicate a significant reduction in prediction error.}, archive = {J_TASE}, author = {Shumei Zhang and Sirui Du and Shuai Tan and Feng Dong}, doi = {10.1109/TASE.2025.3579438}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16644--16657}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Concurrent analysis of local structure and global temporality: Observable and controllable dynamic graph for RUL prediction and evolution illustration}, volume = {22}, year = {2025}, }
• (2025). Dynamic decentralized event-triggered tracking control of continuous stirred tank reactor systems. TASE, 22, 16632--16643. (www)

  @article{ , abstract = {We develop a dynamic decentralized event-triggered tracking control strategy for cascade interconnected continuous stirred tank reactor (CSTR) systems subject to asymmetric input limits. Initially, we construct auxiliary augmented subsystems related to the cascade interconnected CSTR systems. Then, by introducing modified nonquadratic cost functions for the auxiliary augmented subsystems, we convert the decentralized constrained tracking control problem into an array of unconstrained optimal regulation problems. After that, with the construction of dynamic event-triggering mechanisms, we propose the event-triggered Hamilton-Jacobi-Bellman equations (ET-HJBEs) associated with the transformed optimal regulation problems. To approximately solve the ET-HJBEs, we design critic neural networks (CNNs) in the adaptive dynamic programming framework with the CNNs’ weights being updated through the gradient descent method. Furthermore, we use Lyapunov method to prove that the CNNs’ weight estimation errors and the tracking error are stable in the sense of uniform ultimate boundedness. Finally, simulation results of the three-reactor cascade interconnected CSTR systems are provided to validate the present dynamic decentralized event-triggered tracking control scheme. Note to Practitioners—Cascade interconnected continuous stirred tank reactor (CSTR) systems widely emerge in chemical industries. They are sensitive to variations in the reactant concentration and the reaction temperature. Therefore, it is necessary to regulate the concentration of reactants and the reaction temperature via designing an online tracking controller for cascade interconnected CSTR systems. However, the heat transfer limitations and the computational cost must be considered for designing such a tracking controller, which leads to big challenges. This article designs an online-learning decentralized tracking controller in a dynamic event-triggering mechanism for the cascade interconnected CSTR systems within the framework of adaptive dynamic programming. The present decentralized dynamic event-triggered tracking controller is not only able to conquer input constraints but distinctly cut down the computational cost in comparison with the static event-triggered tracking controllers designed for cascade interconnected CSTR systems.}, archive = {J_TASE}, author = {Xiong Yang and Jianling Meng and Qinglai Wei}, doi = {10.1109/TASE.2025.3579230}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16632--16643}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic decentralized event-triggered tracking control of continuous stirred tank reactor systems}, volume = {22}, year = {2025}, }
• (2025). Remaining useful life prediction for hybrid systems under intermittent fault using doubly stochastic process model. TASE, 22, 16616--16631. (www)

  @article{ , abstract = {This paper proposes a stochastic process based remaining useful life (RUL) prediction method for hybrid systems in the presence of intermittent fault. The failure of an intermittently faulty component is determined as a fault appearance with its severity exceeding failure threshold. This failure process is referred to as an event-triggered cooperative failure process (ETCFP). To depict the ETCFP in the hybrid system, a doubly stochastic process model is developed. In this model, the fault occurrence process is characterized by a compound non-homogeneous Poisson process with mode- and degradation-dependent fault occurrence rate and random external impacts, while the degradation process is composed of the external impacts and a Wiener process with the mode-dependent drift coefficient. Then, a multi-stage expectation maximization algorithm is proposed to estimate the unknown parameters in the doubly stochastic process model. In this approach, the stochastic integrals are approximated by the numerical integration with progressively refined step sizes at each stage. After that, the formulation of RUL distribution is derived based on the failure mechanism of ETCFP. To solve the RUL distribution that consists of stochastic integrals, a particle filter based numerical method is employed. Finally, the experimental study is carried out on a hybrid circuit system to show the effectiveness of the proposed method. Note to Practitioners—This paper is motivated by the remaining useful life (RUL) prediction problem of hybrid systems under intermittent fault. The existing degradation models for intermittent fault are deterministic models and the RUL is predicted by extrapolating the degradation trajectory to the failure threshold. However, the random occurrence characteristic of intermittent fault challenges the applicability of these approaches, as the fault may disappear when the extrapolated degradation trajectory reaches the failure threshold, meaning that the intermittently faulty component will not fail immediately. In practice, failure under intermittent fault occurs only when two conditions are met simultaneously: (1) the degradation process exceeds the preset threshold, and (2) a fault appears. This failure process, termed the event-triggered cooperative failure process (ETCFP), results from the interaction between a fault occurrence process and a degradation process. Given the limitations in capturing the stochastic nature of fault occurrence, the deterministic model based methods may not be the suitable solutions. To solve this problem, a doubly stochastic process model is developed, which could characterize the fault occurrence process and the degradation process simultaneously. After that, a multi-stage expectation maximization algorithm is proposed to estimate the unknown parameters in the doubly stochastic process model, using numerical integration to approximate stochastic integrals with progressively refined step sizes at each stage. Finally, the reliability analysis is conducted based on the failure mechanism of the ETCFP, and the RUL distribution is derived from the reliability function. A particle filter-based numerical method is employed to solve this distribution in the presence of stochastic integrals. In this way, both the probability of the degradation trajectory exceeding the failure threshold and the fault occurrence rate can be considered in the derivation of RUL distribution, by which an improved RUL prediction performance can be expected.}, archive = {J_TASE}, author = {Junjie Chen and Ming Yu and Bin Zhang and Rensheng Zhu}, doi = {10.1109/TASE.2025.3579044}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16616--16631}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Remaining useful life prediction for hybrid systems under intermittent fault using doubly stochastic process model}, volume = {22}, year = {2025}, }
• (2025). An interpretable data-driven fuzzy petri net method for industrial domain knowledge modeling of energy efficiency management. TASE, 22, 16603--16615. (www)

  @article{ , abstract = {Industrial domain knowledge modeling aims to integrate multi-source industrial information to build a knowledge model that facilitates industrial system construction. Some studies have sought solutions by fuzzy Petri nets (FPNs), which graphically represent domain knowledge to realize interpretable knowledge modeling. However, these methods cannot meet the intelligent development demands of current industrial systems due to laborious manual modeling, inflexible knowledge representation, and non-adaptive parameter determination. In this paper, we propose an interpretable data-driven FPN (IDFPN) method to realize industrial domain knowledge modeling. Unlike traditional FPN modeling paradigm, IDFPN develops a data mining-based explicit knowledge acquisition (DEKA) to automatically explore the FPN structure. Then, a novel q-rung orthopair fuzzy Petri net (q-ROFPN) model and an adaptive model learning algorithm are proposed to dynamically adjust the knowledge representation ranges and adaptively determine q-ROFPN parameters. We apply IDFPN to address a real-world knowledge modeling problem in energy efficiency management of the data center cooling system. Experimental results demonstrate that the proposed IDFPN can fulfill the industrial domain knowledge modeling task automatically and generate an interpretable q-ROFPN model with good parameter learning capabilities and reasonable inferences regarding system behavior statuses. Note to Practitioners—This paper was motivated by the problem of modeling industrial domain knowledge, which is critical for understanding the operational mechanisms of industrial systems and enabling their intelligent maintenance. Although fuzzy Petri nets (FPNs) can successfully model industrial system domain knowledge in a graphical and interpretable manner, they are usually constructed by manually collecting expert knowledge, which limits their practicality in complex systems. This paper suggests a new method to realize interpretable and fully automated industrial domain knowledge modeling. The proposed method can automatically extract interpretable domain knowledge from industrial data and build the FPN-based industrial knowledge model. The model can visually represent operational rules, optimization strategies, and can complex internal relationships while accurately inferring the industrial system’s operating status. A real-world application of energy efficiency management in data centers suggests that the proposed method is feasible and can be expanded to other fields, such as fault diagnosis, process control, and system management.}, archive = {J_TASE}, author = {Kaiyuan Bai and Wenyu Zhang and Shiping Wen and Dan Jia and Weiye Meng}, doi = {10.1109/TASE.2025.3578598}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16603--16615}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An interpretable data-driven fuzzy petri net method for industrial domain knowledge modeling of energy efficiency management}, volume = {22}, year = {2025}, }
• (2025). Human–Robot coordination control for sit-to-stand assistance in hemiparetic patients with supernumerary robotic leg. TASE, 22, 16591--16602. (www)

  @article{ , abstract = {In light of global aging and prevalent stroke-related hemiplegia, this study addresses challenges in robot-assisted Sit-to-Stand (STS) movements, a daily activity prone to falls. Supernumerary Robotic Legs (SRL) serve as independent support, enhancing stability and limb movement range. Existing coordination control methods lack personalization for STS assistance, requiring solutions for human intent transmission and rapidly optimize coordination control challenges in the non-coupled human-robot system. The proposed human-SRL coordination control algorithm, grounded in personalized SRL-human coupling models, incorporates surface electromyography (sEMG) signals to design an intent-driven variable stiffness impedance control. The inclusion of incremental learning enables rapid optimization of impedance parameters, facilitating real-time adjustments in SRL assistance for adaptive coupling with users. Practical experiments involving both healthy participants and hemiparetic patients validate the algorithm’s effectiveness during STS. The results validate substantial reductions in STS time (39.54%) and muscle activity (28.01%), highlighting the efficacy of the proposed algorithm-controlled SRL support for hemiparetic individuals. Note to Practitioners—To prevent misalignment between the SRL and natural limb movements, which can impose additional strain on the body, this study combines the principles of joint bio-mimicry to develop a personalized human-robot coupling model, and also propose an sEMG-based adaptive impedance control algorithm. Our research fills a gap in the nuanced analysis of robot-assisted support effects on both the affected and unaffected sides of hemiparetic patients, while also possessing the potential for transfer or extension to other types of rehabilitation robotic platforms. Although the SRL shows promising results, it is important to note the limitations of a small sample size and the need for further exploration of the needs of diverse patient populations. Future research could build on our findings by integrating additional sensing technologies and optimizing control strategies to enhance adaptability. This work has the potential to significantly improve the daily lives of individuals with mobility impairments, paving the way for more personalized and effective rehabilitation solutions.}, archive = {J_TASE}, author = {Jie Zuo and Jun Huo and Xiling Xiao and Yanzhao Zhang and Jian Huang}, doi = {10.1109/TASE.2025.3578594}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16591--16602}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human–Robot coordination control for sit-to-stand assistance in hemiparetic patients with supernumerary robotic leg}, volume = {22}, year = {2025}, }
• (2025). Confidence-aware robust dynamical distance constrained reinforcement learning for social robot navigation. TASE, 22, 16572--16590. (www)

  @article{ , abstract = {Navigating in a crowded social environment without collisions or freezing is a crucial and challenging task. Recent studies have demonstrated considerable success using Deep Reinforcement Learning for social robot navigation. However, the lack of safety guarantees remains a prevalent problem among these methods. To address this limitation, we present a comprehensive safety framework that builds upon a novel multi-agent environment single-agent decision-making formulation and the strengths of Safe Reinforcement Learning (Safe RL) methods. First, considering the uncertainty of pedestrians and the inaccuracy of trajectory predictions, we explicitly embed confidence-weighted uncertainty and risk measurement in the interaction graph network to induce adaptive policies. Next, we replace the state-wise robust-form unshaped cost with a cumulative robust dynamical distance constraint, which establishes a stronger and essential safety objective while ensuring the invariance of the optimal policy. Then we use reachability-based Safe RL to ensure state-wise safety assurance of the feasible part and the risk minimization of the infeasible part. Simulation and virtual-reality evaluation results demonstrate that our proposed method outperforms the state-of-the-art methods in various metrics, such as success rate and collision energy, achieving the best trade-off between safety and efficiency. Our proposed method yields a policy that exhibits desired collision avoidance behavior in both simulation and real-world experiments. Note to Practitioners—The motivation of this paper is to enhance the safety of robots navigating in human-filled environments without being overly conservative. Typical application scenarios include mobile robots navigating through crowds and autonomous vehicles moving through unprotected intersections. On one hand, existing deep reinforcement learning-based navigation methods have shown remarkable capabilities in complex dynamic environments but often lack guaranteed safety constraints. On the other hand, due to the uncertainty introduced by other uncontrollable agents in the environment, unreasonable safety constraints can lead to suboptimal optimization problems, resulting in overly conservative policies. This paper starts from the high-level intrinsic safety objective and proposes a comprehensive safety framework that balances safety and efficiency. For clarity, the proposed approach is structured into several key modules, including confidence-weighted uncertainty, robust dynamical distance constraints, and safe reinforcement learning, among others. It is important to recognize that these modules form a tightly coupled system. We validate the feasibility and effectiveness of our proposed approach through theoretical analysis, simulations, virtual-reality tests, and real-world experiments. In future work, this approach will be extended to incorporate technologies such as sim-to-real transfer and static obstacle handling, enabling more effective deployment in real-world social environments.}, archive = {J_TASE}, author = {Kai Zhu and Tao Xue and Tao Zhang}, doi = {10.1109/TASE.2025.3578326}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16572--16590}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Confidence-aware robust dynamical distance constrained reinforcement learning for social robot navigation}, volume = {22}, year = {2025}, }
• (2025). Fully distributed consensus control under dual-channel dynamic event-triggered strategies. TASE, 22, 16560--16571. (www)

  @article{ , abstract = {This paper intends to propose a fully distributed consensus protocol under the dual-channel dynamic event-triggered strategies. First, an observer is devised to estimate the unmeasurable system state, and a dual-channel triggering strategies are established to save the network resources. Further, a fully distributed control protocol is devised to achieve consensus control without requiring the global topology information. The major superiority of the proposed method is that the controller, the observer, and their respective triggering conditions can be designed separately without affecting each other. Finally, simulations are presented for illustrating the validity of the theoretical result. Note to Practitioners—Recently, networked control systems have garnered growing interest in automation and control engineering due to their ability to provide enhanced operational efficiency, reduced implementation costs, and simplified system integration capabilities. However, an important factor that restricts the advancement of the networked control systems is the issue of the limited network communication resources. Therefore, how to complete the desired control task while saving network communication resources is a widely studied problem. This paper designs a novel fully distributed consensus protocol with the intermittent communication, which has the feature that the controller, the observer and their respective triggering conditions can be designed separately without affecting each other. Simulation results indicate that the presented method effectively saves the network communication resources.}, archive = {J_TASE}, author = {Yong-Sheng Ma and Wei-Wei Che}, doi = {10.1109/TASE.2025.3578045}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16560--16571}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully distributed consensus control under dual-channel dynamic event-triggered strategies}, volume = {22}, year = {2025}, }
• (2025). Cognitive UAV tracking: Leveraging DRL and hybrid curriculum learning for target reacquisition. TASE, 22, 16547--16559. (www)

  @article{ , abstract = {Tracking a moving uncrewed ground vehicle (UGV) with an autonomous Uncrewed Aerial Vehicle (UAV) is challenging, particularly in GNSS-denied indoor environments where reacquiring the UGV after losing track poses a significant obstacle. This paper presents a novel learning framework designed to address these challenges, enabling a quadrotor UAV to effectively chase a moving UGV and regain tracking in an indoor environment. The proposed framework encompasses two primary components: the Track-HCL and the Tracking Vision System (TVS). The TVS leverages a lightweight tracker to offer real-time recognition and localization of the UGV. Additionally, the Chronological Ghosting (CG) method is employed to describe the UGV’s motion trend within a single frame. The Track-HCL component introduces a hybrid curriculum strategy to guide policy learning for the Deep Reinforcement Learning (DRL) agent. The Track-HCL enables the agent to learn the tracking policy conducive to target chasing and proficient reacquisition. We demonstrate the effectiveness of the proposed method in both simulation and field experiments.Note to Practitioners: This research addresses the challenge of autonomously tracking a moving UGV with an UAV in environments where GPS signals may not be available, such as indoors. The solution is particularly valuable for applications requiring precise location tracking, such as logistics in large warehouses, surveillance, or search and rescue missions where real-time data and reactivity are crucial. Our approach combines a novel learning framework using DRL with a hybrid curriculum strategy, enhancing the UAV’s ability to adapt and respond to dynamic changes in the UGV’s path even when visual contact is temporarily lost. The TVS we developed is key to this adaptability, providing real-time recognition and localization of the UGV, which allows the UAV to maintain tracking or quickly reacquire the target after disruptions caused by obstacles or signal interference. The practical implementation of our system shows that it can significantly improve the efficiency and reliability of UAV tracking operations. It reduces the need for manual intervention in tracking processes and increases operational uptime in environments where traditional tracking methods may fail. However, there are limitations, such as the dependency on visual cues, which might be obstructed in highly cluttered environments. Future enhancements could involve integrating multisensory data to mitigate this. The potential applications of this technology extend beyond the tested scenarios. Industries that operate in complex and dynamic environments, like urban logistics or emergency response in disaster-stricken areas, could benefit from deploying UAVs equipped with our tracking system. This would not only improve operational efficiency but also enhance safety and response times.}, archive = {J_TASE}, author = {Jiaqing Wang and Baichuan Zeng and Lan Deng and Ze Ji and Changyun Wei and Zheng Zeng}, doi = {10.1109/TASE.2025.3577984}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16547--16559}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cognitive UAV tracking: Leveraging DRL and hybrid curriculum learning for target reacquisition}, volume = {22}, year = {2025}, }
• (2025). Robust IMC for time delay attack compensation of renewable supported power system. TASE, 22, 16532--16546. (www)

  @article{ , abstract = {This study presents the design of a robust control strategy utilizing an internal model control (IMC)-based proportional integral-one plus derivative with filter (PI-(1+DF)) controller to mitigate the impact of time-delay attacks (TDAs) on renewable-integrated power system (PS). Before implementing mitigation, an adaptive recursive least square filter with a forgetting factor (ARLS-FF) is employed as an online detection mechanism for TDAs. The design methodology incorporates Kharitonov’s stability theorem to identify the worst-case plant, for which the proposed controller parameters are obtained using the IMC framework. A critical feature of IMC structure is the filter coefficient $({\mu })$ , which is determined based on the maximum inverse sensitivity ${(\boldsymbol {M}_{\boldsymbol {T}}}\boldsymbol {)}$ and the characteristics of considered TDA, underscoring the proposed control scheme’s enhanced efficacy. Stability under time-varying attacks is rigorously analyzed through the Lyapunov-Krasovskii function (LKF) and linear matrix inequalities (LMIs) based approach with vivid robustness assessment ensuring resilience to parametric uncertainties. Finally, benchmarking on the IEEE-39 and IEEE-118 bus systems validates the proposed controller’s superiority on large-scale and realistic interconnected PSs. Note to Practitioners—In PS operations, TDAs are among the most serious cyber threats, with the potential to destabilize the system and even cause blackouts in extreme cases. As a result, TDAs have become a growing focus in PS research. This study introduces IMC-based robust load frequency control (LFC) designed to counteract TDAs, thereby improving the effectiveness of current LFC techniques. The controller is designed for a TDA of 2s but demonstrates robust performance up to 4s of TDA. The proposed IMC-PI-(1+DF) control architecture’s effectiveness is evaluated under various practical scenarios, including constant TDAs, stochastic time-delayed cyber-attacks (STCAs), parametric uncertainties, nonlinearities, step load perturbations (SLPs), renewable energy (solar and wind) integrations, and in a realistic interconnected PS setting, exemplified by the IEEE-39 and IEEE-118 bus systems. Results confirm that the proposed robust control scheme maintains frequency stability within the permissible range under these diverse conditions, underscoring its applicability in practical PS engineering.}, archive = {J_TASE}, author = {Pawan Kumar Pathak and Anil Kumar Yadav and Innocent Kamwa}, doi = {10.1109/TASE.2025.3577725}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16532--16546}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust IMC for time delay attack compensation of renewable supported power system}, volume = {22}, year = {2025}, }
• (2025). Predictor-based adaptive iterative learning control of MASs with distributed error compensation. TASE, 22, 16522--16531. (www)

  @article{ , abstract = {This paper proposes an adaptive iterative learning control (AILC) scheme for multiagent systems (MASs) to improve the containment control performance. To deal with the uncertain nonlinearity, a neural network (NN)-based iterative predictor with the same structure of MASs is constructed, where an auxiliary approximation term is further incorporated to handle the NN reconstruction error. By utilizing the output of iterative predictor, an AILC-based containment control scheme is designed under the backstepping framework. To overcome the initial error problem, distributed compensation signals are developed to compensate for containment errors and dynamic surface errors. Instead of using dynamic information of neighbors, only the output of neighbors is used in each follower such that the communication load is reduced. It is established that as iteration index approaches infinity, the iteratively convergence of containment errors is achieved, and the stability of MASs is ensured. Simulation results on manipulators verify the effectiveness of both the iterative predictor and the adaptive iterative learning containment controller. Note to Practitioners—This paper is motivated by practical applications of AILC-based containment control for MASs, such as robot collaboration, intelligent transportation, smart grids, etc. Distributed containment control allows each agent to communicate only with its neighbors, thereby reducing unnecessary communication overhead. However, the overall communication network is unknown for each agent, which leads to the inevitable initial containment error. To address this issue, a distributed compensation signal is introduced to adjust the containment error. Additionally, an NN-based iterative predictor is designed to indirectly modify the control input to improve control performance. Since the transmission and reception of dynamic information from neighbors occupy communication resources, the proposed method can only utilize the output of neighbors to design the distributed containment controller. This approach further reduces communication load, which makes the AILC-based containment control method more flexible and suitable in practical applications.}, archive = {J_TASE}, author = {Zhiqiang Li and Yang Liu and Hongru Ren and Hongyi Li}, doi = {10.1109/TASE.2025.3577609}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16522--16531}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predictor-based adaptive iterative learning control of MASs with distributed error compensation}, volume = {22}, year = {2025}, }
• (2025). Cascade finite-time adaptive control for stand-alone inverters with load disturbances. TASE, 22, 16512--16521. (www)

  @article{ , abstract = {Three-phase inverters have been widely implemented for stand-alone power conversion applications where the utility grid is not available. In these applications, high-quality output voltage regulation of inverters is crucial for the reliable operation of local loads. However, critical load conditions (e.g., unbalanced loads, nonlinear loads, and load variations) bring time-varying load disturbances, deteriorating the steady-state and transient performance of the output voltage. To address this issue, a finite-time adaptive control (FTAC) with a cascade structure is proposed in this article. Firstly, novel adaptive laws are designed to estimate time-varying load disturbances. By incorporating the designed adaptive laws, cascade finite-time controllers are then constructed for both the outer voltage loop and inner current loop. The stability analysis shows that the voltage tracking errors tend to an arbitrarily small neighborhood of zero within a finite time, enabling fast and accurate output voltage control of stand-alone inverters under load disturbances. Meanwhile, all the signals in the closed-loop system are bounded. Simulation and experiment results validate the effectiveness and superiority of the FTAC strategy. Note to Practitioners—Stand-alone inverters are key power interfaces in distributed generation systems, providing stable output voltage support for local loads. Nevertheless, a challenging problem is how to mitigate the impact of load disturbances on the voltage regulation performance. Using additional load current sensors can detect and compensate for load disturbances, but it increases costs and reduces reliability from the perspective of practical applications. This paper proposes an FTAC strategy to guarantee high-quality voltage regulation in the presence of time-varying load disturbances, without using load current sensors.}, archive = {J_TASE}, author = {Cheng Fu and Jiapeng Liu and Zicheng Zhang and Qing-Guo Wang and Jinpeng Yu}, doi = {10.1109/TASE.2025.3577455}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16512--16521}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cascade finite-time adaptive control for stand-alone inverters with load disturbances}, volume = {22}, year = {2025}, }
• (2025). Exact algorithm for batch scheduling with many-to-many job-tool matching and job release time constraints. TASE, 22, 16499--16511. (www)

  @article{ , abstract = {Job-dependent tool switching is necessary in many batch processing systems (BPSs). Heterogeneous tool demand and extra time consumption for tool switches bring great challenges for high-performance production scheduling in BPSs. In this study, we present a novel batch scheduling problem derived from a satellite vibration test system. Different from basic scheduling problems in BPSs with tool switching, it has complications caused by many-to-many task-tool matching and job release time constraints. Specifically, each job has predetermined candidate tools. One of them should be selected and equipped to process a given job. Unlike sequence-dependent setup time that is only related to two adjacent jobs, tool switching time is additionally dependent on the tools selected by both jobs from their respective candidate tools. Both job sequence and tool selections need to be determined to minimize makespan. Mixed-integer linear programming (MILP) and constraint programming (CP) models are developed for formulating the problem. We prove that the problem is $\mathcal {NP}$ -hard. To obtain exact solutions, we propose a novel branch-and-bound algorithm (B&B) based on the analysis and extraction of problem properties. It integrates a domain reduction mechanism, branching strategies, and dominance rules. Among them, the domain reduction mechanism reduces the number of jobs to be scheduled; the branching strategies accelerate solving speed; and the dominance rules save the memory consumption. Ablation studies and key influencing factor experiments are performed. The results show that the proposed B&B can optimally solve much larger instances than state-of-the-art methods (including some heuristics). It can optimally solve industrial-size problems in a short time, implying that it can well satisfy the scheduling need of a satellite vibration test system and thus advance the field of BPS scheduling. Note to Practitioners—This work deals with a new batch scheduling problem arising from a satellite vibration test system. It has unique properties that there is a many-to-many mapping between jobs and tools while job’s release time affects its availability. Mixed-integer linear programming and constraint programming models are formulated for this problem. Afterward, a branch-and-bound algorithm is developed to exactly solve the problem. We embed a domain reduction mechanism, branching strategies, and dominance rules into it, which together contribute to its high performance. Ablation studies and key influencing factor experimental results show that it can find optimal solutions more efficiently than its peers. It can obtain an optimal solution of the concerned problem with 50 jobs in less than 16 seconds on average and even 1 second for some representative industrial cases, thus making it immediately applicable to a satellite vibration test system.}, archive = {J_TASE}, author = {Shengchao Li and Shixin Liu and Ziyan Zhao and MengChu Zhou}, doi = {10.1109/TASE.2025.3575531}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16499--16511}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exact algorithm for batch scheduling with many-to-many job-tool matching and job release time constraints}, volume = {22}, year = {2025}, }
• (2025). Dynamic self-triggered intelligent path tracking control for autonomous agricultural vehicles via reinforcement learning. TASE, 22, 16487--16498. (www)

  @article{ , abstract = {This paper investigates the path tracking control of autonomous agricultural vehicles with disturbances under a dynamic self-triggered mechanism via reinforcement learning (RL). To begin with, a path tracking offset system is constructed based on the kinematic model of the autonomous agricultural vehicle, which transforms the path tracking control problem into an optimal control problem. Subsequently, a novel dynamic self-triggered second-order integral sliding mode control policy is developed to mitigate the impact of disturbances and to derive a nominal path tracking offset model. Afterward, to further alleviate the computing and communication burdens, a novel dynamic self-triggered mechanism is proposed for the optimal control policy. It can predict the next update time based on current information, thus avoiding the need for continuous monitoring of the triggering condition. Furthermore, a single critic network architecture is constructed to obtain an approximate path tracking control policy, and it is proven by Lyapunov stability theory that this policy ensures autonomous agricultural vehicles can maintain the predefined working path in the presence of disturbances. Finally, the effectiveness of the proposed path tracking control method is demonstrated by simulation experiments. Note to Practitioners—Autonomous agricultural vehicles serve as the core equipment in autonomous farms, and path tracking control is a critical technology for achieving precise and efficient operations. However, existing path tracking control methods for autonomous agricultural vehicles primarily focus on system stability, without considering the control cost. Moreover, traditional methods are typically time-triggered, leading to substantial consumption of computing and communication resources. To overcome the above bottleneck, this paper proposes an RL-based dynamic self-triggered path tracking control method for autonomous agricultural vehicles with disturbances. The proposed method comprises two components, that is, a second-order integral sliding mode control policy to mitigate the impact of disturbances and an optimal control policy to ensure that the autonomous agricultural vehicle follows the desired path. Notably, both control policies are implemented within the self-triggered mechanism, which predicts the next update time based on current information, such that the need for continuous state monitoring is eliminated. Simulation results demonstrate that the proposed method enhances the control performance and enables the completion of operational tasks in an efficient and energy-saving manner.}, archive = {J_TASE}, author = {Yongwei Zhang and Derong Liu}, doi = {10.1109/TASE.2025.3575353}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16487--16498}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic self-triggered intelligent path tracking control for autonomous agricultural vehicles via reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Event-triggered prescribed-time tracking control for UAVs using polynomial error trajectories. TASE, 22, 16476--16486. (www)

  @article{ , abstract = {In this paper, we propose a novel event-triggered prescribed-time tracking (PTT) control method for an underactuated unmanned aerial vehicle (UAV). Unlike existing PTT control methods, where the transient behavior of the system depends solely on the controller gains, we design a novel reference error trajectory (RET) using a polynomial to guide the convergence of the tracking error, particularly during the transient phase. Based on this RET and within the backstepping framework, we design a thrust reference. Additionally, we introduce a novel event-triggered mechanism based on the tracking error to update the torque applied to the UAV. To streamline intricate mathematical expressions and improve robustness to external disturbances, a second-order linear system is used as a low-pass filter within the backstepping design. Finally, simulation results are provided to demonstrate the effectiveness of the proposed approach and validate the accuracy of the theoretical predictions. Note to Practitioners—This paper introduces a novel prescribed-time control method for an underactuated UAV, incorporating an event-triggered technique. This approach aims to enhance tracking accuracy while reducing control signal updates, thereby lowering communication overhead and increasing system efficiency. A reference error trajectory is designed for the position error, guiding the UAV to converge quickly to the reference trajectory within a prescribed time. This ensures precise tracking even in dynamic and uncertain environments, making it suitable for real-world applications. Additionally, a second-order filter is introduced to reduce the computational burden from repeated differentiation of the virtual controller in the backstepping design. Our overall approach provides an effective solution for critical tasks such as autonomous navigation, search-and-rescue, and surveillance, making it a promising technology for UAV applications in dynamic, high-stakes environments.}, archive = {J_TASE}, author = {Linghuan Kong and Joel Reis and Wei He and Carlos Silvestre}, doi = {10.1109/TASE.2025.3574809}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16476--16486}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered prescribed-time tracking control for UAVs using polynomial error trajectories}, volume = {22}, year = {2025}, }
• (2025). Adaptive tracking control with disturbance rejection for cable-driven exoskeleton robot with compliant actuators. TASE, 22, 16462--16475. (www)

  @article{ , abstract = {The cable-driven upper-limb exoskeleton robot with compliant actuators presents notable advantages when applied to rehabilitation training, yet introduces challenges in position tracking control. Traditional control methods inadequately address hysteresis issues caused by cable drives and the performance degradation resulting from compliant actuators. To solve these common issues in such robotic systems, this study introduces a unified dual closed-loop control strategy that aims to enhance system tracking performance. In the outer loop, an adaptive hysteresis compensator is intricately designed to address position errors induced by cable-driven motion from a kinematic perspective. Simultaneously, the inner loop employs a high-order controller utilizing the backstepping method to control the dynamics of the robotic system with flexible joints effectively. Additionally, a dedicated high-order super-twisting observer (STO) is integrated to estimate unknown dynamic models and external disturbance, enabling the generation of disturbance rejection items for the inner-loop controller design. Experimental results validate the efficacy of the proposed control strategy, demonstrating commendable tracking performance when compared with alternative methods. Note to Practitioners—This study presents a novel approach to address challenges in position tracking control for cable-driven upper-limb exoskeleton robots with compliant actuators. By introducing a unified dual closed-loop control strategy, we provide an effective solution to mitigate hysteresis issues from cable drives and performance degradation caused by flexible joints, which traditional methods have overlooked. The proposed control method is suitable for achieving precise position tracking, such as in rehabilitation robots for providing motion training to stroke patients or in exoskeleton robots for accomplishing specific motion tasks for users. Additionally, the proposed inner-loop backstepping controller method offers a design paradigm for controlling a high-order system with both matched and mismatched disturbances. Importantly, the adaptability of our approach holds potential applicability in various other cable-driven robotic systems with compliant actuators or flexible joints.}, archive = {J_TASE}, author = {Xiaoyu Wu and Jun Yang and Haoyong Yu}, doi = {10.1109/TASE.2025.3573762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16462--16475}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tracking control with disturbance rejection for cable-driven exoskeleton robot with compliant actuators}, volume = {22}, year = {2025}, }
• (2025). Sliding mode control for IT2 fuzzy systems under sensor fault estimation. TASE, 22, 16451--16461. (www)

  @article{ , abstract = {This paper delves into the output feedback sliding mode control (SMC) problems in a class of Interval Type-2 (IT2) fuzzy systems affected by sensor faults. Firstly, a Markovian chain is employed to depict transmission channels vulnerable to stochastic deception attacks, providing a more accurate representation of the real-world scenario. Secondly, by leveraging historical estimates of fault signals, a novel learning-based observer integrated with a Saturation Function is proposed to reconstruct sensor faults. This observer incorporates a dynamic saturation level aimed at enhancing filtering performance. Subsequently, based on the estimated fault information, an output feedback SMC law is designed, and its finite-time boundedness is analyzed. Finally, the proposed approach and its associated theoretical findings are extended to an inverted pendulum model. Note to Practitioners—The proposed method introduces a novel sliding surface design based on real-time fault estimation and utilizes an adaptive saturation function to enhance system robustness. This approach proves especially effective when dealing with nonlinear, uncertain, and fault - prone systems across industrial automation, robotics, and aerospace engineering. Unlike conventional SMC methods, the proposed framework explicitly incorporates sensor fault dynamics, allowing for real-time compensation and ensuring finite-time boundedness of the system’s states. By implementing this control strategy, system performance can be improved, the impact of faults can be reduced, and operational stability can be enhanced, even in the presence of uncertainties and faults. This study provides practical insights and theoretical guarantees, bridging the gap between advanced control theory and its industrial applications.}, archive = {J_TASE}, author = {Qiongwen Zhang and Huaguang Zhang and Lei Wan and Zhihong Liang}, doi = {10.1109/TASE.2025.3571379}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16451--16461}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sliding mode control for IT2 fuzzy systems under sensor fault estimation}, volume = {22}, year = {2025}, }
• (2025). Fully distributed secure consensus control for cyber-physical systems against actuator fault and denial-of-service attack. TASE, 22, 16440--16450. (www)

  @article{ , abstract = {The cooperative control problem for cyber-physical systems under actuator faults and denial-of-service attacks is investigated, where each subsystem can be modeled by an agent and denial-of-service attacks are viewed as attacks against the inter-agent communication. Specifically, a scenario is considered in which the communication topology loses connectivity following a denial-of-service attack. To address the above problem, a fully distributed secure control strategy is proposed. This strategy combines a distributed observer with an observer-based fault-tolerant consensus control method. Therein, the assumptions that the leader’s communication links are unbreakable and the switched topology contains a spanning tree are further relaxed based on the idea of a jointly connected topology. In addition, this paper also has the following features: Firstly, the proposed scheme eliminates the need for all agents to have prior knowledge of the leader’s dynamics, making the scenario considered in this work more general and applicable. Secondly, the control method presented in this paper does not rely on any global topology information, allowing it to be implemented in a completely distributed manner. Building on the proposed scheme, even under denial-of-service attacks and actuator faults, the consensus problem for cyber-physical systems can be realized. Finally, simulations are provided to demonstrate the effectiveness of the proposed method. Note to Practitioners—CPSs are regarded as the core technology for the next generation of manufacturing, with widespread applications in fields such as autonomous driving systems, smart grids, and intelligent manufacturing. The advantage of CPSs lies in their ability to tightly integrate physical processes with computational and communication technologies, enabling efficient and intelligent system operations. However, in engineering practice, while the application of network technologies has brought great convenience and significantly improved operational efficiency, their open nature also makes CPSs vulnerable to cyber-attacks. Additionally, during long-term operation, CPSs inevitably face fault issues, which can negatively affect system performance, potentially leading to partial system failure or a decline in stability. Based on the above analysis, a fully distributed secure consensus control strategy is proposed, which incorporates a distributed observer and an observer-based fault-tolerant consensus control method. The proposed strategy does not require all agents to have prior knowledge of the leader’s dynamics information, offering greater flexibility and adaptability. More importantly, the assumptions that the leader’s communication links are unbreakable and that the switched topology contains a spanning tree are further relaxed. This makes the solution applicable to a broader range of real-world scenarios.}, archive = {J_TASE}, author = {Yue Long and Ximing Yang and Tieshan Li and Hongjing Liang}, doi = {10.1109/TASE.2025.3577599}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16440--16450}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully distributed secure consensus control for cyber-physical systems against actuator fault and denial-of-service attack}, volume = {22}, year = {2025}, }
• (2025). Tuning function-based light computational adaptive fixed-time control for overhead cranes with multiple uncertainites. TASE, 22, 16428--16439. (www)

  @article{ , abstract = {Overhead cranes are important transportation equipments in practice, however, their existing control methods have encountered many difficulties in applications due to the underactuation, input limitation and computation complexity. This paper proposes an adaptive fixed-time control scheme for the underactuated overhead crane with multiple uncertainties to deal with the above challenges simultaneously. A coordinate change is employed to address the underactuated structure by reformulating the crane dynamics as a strict-feedback system. A series of time-varying tuning functions are designed to guarantee the input signal varies within a small range to meet the practical input requirement of the overhead crane system. Moreover, a second-order nonlinear tracking differentiator (NLTD) is set up to avoid the repetitive derivative calculation of the virtual controllers. Then, an adaptive law is designed to tackle multiple uncertainties with no need of introducing any other control algorithms but only the single one of itself. Further, a light computational adaptive fixed-time control scheme is proposed by consisting of the tuning functions, NLTD, and the adaption law to achieve a fast location of the overhead crane system. The simulation experiments illustrate the effectiveness of the presented method. Note to Practitioners—This work is to develop an adaptive fixed-time control method for a kind of overhead cranes, which is of important significance in the field of ocean engineering. The time-varying tuning function is proposed to guarantee an even start and a desired swing range of the crane, which is necessary in practice. Furthermore, a second-order nonlinear tracking differentiator (NLTD) is built to eliminate the repetitive derivative of the virtual control signals. Only the single adaptive law is designed to deal with multiple uncertainties with no need of introducing any other control techniques. Therefore, the presented adaptive fixed-time control algorithm has a light computation burden, which is more suitable for practical applications. The designed scheme can achieve a fast location of the overhead crane system, which saves the control cost to some extent. Moreover, the light computation method is also suitable for the platform with weak computing ability. The elementary experiments are carried out based on the self-built crane hardware test platform. In the future, the designed control scheme will be further applied to an industrial overhead crane system to enhance the efficiency of the presented algorithm.}, archive = {J_TASE}, author = {Jia-Ke Wang and Yang Liu and Ronghu Chi and Xuhui Bu and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3576775}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16428--16439}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tuning function-based light computational adaptive fixed-time control for overhead cranes with multiple uncertainites}, volume = {22}, year = {2025}, }
• (2025). Reward-based collision-free algorithm for trajectory planning of autonomous robots. TASE, 22, 16415--16427. (www)

  @article{ , abstract = {This paper proposes a novel mission planning algorithm for autonomous robots that selects an optimal waypoint sequence from a predefined set to maximize total reward while satisfying obstacle avoidance, state, input, mission time, and distance constraints. The formulation extends the prize-collecting traveling salesman problem. A tailored genetic algorithm evolves candidate solutions using a fitness function, Crossover, and mutation, with constraint enforcement via a penalty method. Differential flatness and clothoid curves are employed to penalize infeasible trajectories efficiently, while the Euler spiral method ensures curvature-continuous trajectories with bounded curvature, enhancing dynamic feasibility and mitigating oscillations typical of minimum-jerk and snap parameterizations. Due to the discrete variable length optimization space, crossover is performed using a dynamic time-warping-based method and extended convex combination with projection. The algorithm’s performance is validated through experiments with a ground vehicle, quadrotor, and quadruped, supported by benchmarking and time-complexity analysis. Note to Practitioners—This work addresses the challenge of managing multiple tasks within limited resources in cyber-physical systems, where tasks vary in priority and maximizing total reward is critical. We propose an evolutionary-inspired algorithm that optimizes task selection and prioritization in a motion planning context, generating trajectories that respect robot capabilities while favoring higher-priority tasks and potentially bypassing lower-priority ones to fulfill constraints. The algorithm integrates with traditional motion planning techniques, ensuring safe, resource-aware trajectories. It has been validated on a ground vehicle, a quadruped robot, and a quadrotor drone, demonstrating adaptability across platforms with varying constraints. Practical contributions include the flexibility of the method to accommodate different resource limitations, such as time, distance, and energy. The current formulation assumes static rewards and constraints; future work will explore real-time adaptation and extensions to broader classes of robots, including collaborative scenarios involving human-robot interaction.}, archive = {J_TASE}, author = {Jose D. Hoyos and Tianyu Zhou and Zehui Lu and Shaoshuai Mou}, doi = {10.1109/TASE.2025.3575762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16415--16427}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reward-based collision-free algorithm for trajectory planning of autonomous robots}, volume = {22}, year = {2025}, }
• (2025). A filter-based extended state observer with arbitrary noise suppression ability and its application to drag-free spacecraft system. TASE, 22, 16403--16414. (www)

  @article{ , abstract = {In this work, we investigate the measurement noise sensitivity and peaking problem of the extended state observer (ESO) under high-gain characteristics. For a class of $ n $ -order nonlinear systems with measurement noise, a new variable-order filter and the corresponding filtered ESO (FESO) are constructed to achieve arbitrary suppression ability of measurement noise. Such a salient feature lies in the establishment of the variable-order filter with independent parameters such that the observer poles and the order of the proposed filter can be set arbitrarily by the designer, thereby allowing the user to predefine the suppression strength in prior by adjusting parameters. In addition, a peaking-free FESO is further proposed to eliminate the undesired peaking phenomenon during the transient state. Finally, a comprehensive numerical comparison is presented, and the proposed peaking-free FESO is applied to the drag-free spacecraft hardware in-loop platform. The simulation results reveal the advantages of the proposed scheme in measurement noise suppression and peaking elimination. Note to Practitioners—Drag-free spacecraft, as a tool for detecting low-frequency gravitational waves, plays a crucial role in deep-space exploration. This paper focuses on the measurement noise sensitivity and peaking problem of the extended state observer in disturbance rejection control. We propose a novel filtered ESO (FESO) utilizing the variable-order filter to achieve the arbitrary suppression ability of high-frequency measurement noise. Meanwhile, a peaking-free FESO is further provided to eliminate the undesirable peaking by inserting the saturation function inside the observer. Finally, the proposed peaking-free FESO is applied to the drag-free spacecraft hardware in-loop platform “Taiji Pathfinder”. The experimental results show that the proposed scheme not only reduces the high-frequency fluctuation of the system output but also decreases the control input. These innovative noise suppression features would also be valuable for other drag-free control missions such as “Taiji”, “Tianqin” and “LISA”.}, archive = {J_TASE}, author = {Guotao Zhao and Bing Cui and Pengcheng Wang and Yuanqing Xia and Yonghe Zhang}, doi = {10.1109/TASE.2025.3574422}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16403--16414}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A filter-based extended state observer with arbitrary noise suppression ability and its application to drag-free spacecraft system}, volume = {22}, year = {2025}, }
• (2025). Novel looped functionals in designing output feedback controllers for aperiodic sampled-data control systems. TASE, 22, 16397--16402. (www)

  @article{ , abstract = {This paper addresses the problem of stabilizing aperiodic sampled-data systems using output feedback control. First, a novel two-sided looped functional is constructed, providing a sufficient condition to ensure the asymptotic stability of the resulting closed-loop system. This condition is then adapted for control design. Using a cone complementary linearization algorithm with a stringent iteration criterion, a practical approach is developed to compute the control gain. Finally, three examples, including an inverted pendulum system and a one-area load-frequency control system, demonstrate the effectiveness and superiority of the proposed method over others in the literature. Note to Practitioners—This paper tackles the challenge of stabilizing aperiodic sampled data systems through output feedback control. We introduce a novel two-sided looped function that establishes a sufficient condition for ensuring the asymptotic stability of the closed-loop system. This condition is then refined for practical control design. To illustrate the advantages of our approach, an inverted pendulum system and a one-area load-frequency control system are provided, demonstrating its efficacy and advantages. The presented method can streamline the design process of networked control systems, saving network communication resources and thus reducing the requirements for network bandwidth.}, archive = {J_TASE}, author = {Wei Wang and Jin-Ming Liang and Hong-Bing Zeng and Xian-Ming Zhang}, doi = {10.1109/TASE.2025.3573304}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16397--16402}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Novel looped functionals in designing output feedback controllers for aperiodic sampled-data control systems}, volume = {22}, year = {2025}, }
• (2025). System-level efficient performance of EMLA-driven heavy-duty manipulators via bilevel optimization framework with a leader-follower scenario. TASE, 22, 16373--16396. (www)

  @article{ , abstract = {The global push for sustainability and energy efficiency is driving significant advancements across various industries, including the development of electrified solutions for heavy-duty mobile manipulators (HDMMs). Electromechanical linear actuators (EMLAs), powered by permanent magnet synchronous motors, present an all-electric alternative to traditional internal combustion engine (ICE)-powered hydraulic actuators, offering a promising path toward an eco-friendly future for HDMMs. However, the limited operational range of electrified HDMMs, closely tied to battery capacity, highlights the need to fully exploit the potential of EMLAs that drive the manipulators. This goal is contingent upon a deep understanding of the harmonious interplay between EMLA mechanisms and the dynamic behavior of heavy-duty manipulators. To this end, this paper introduces a bilevel multi-objective optimization framework, conceptualizing the EMLA-actuated manipulator of an electrified HDMM as a leader-follower scenario. At the leader level, the optimization algorithm maximizes EMLA efficiency by considering electrical and mechanical constraints, while the follower level optimizes the manipulator’s motion through a trajectory reference generator that adheres to manipulator limits. This optimization approach ensures that the system operates with a synergistic trade-off between the most efficient operating region of the actuation system, achieving a total efficiency of 70.3%. Furthermore, to complement this framework and ensure precise tracking of the generated optimal trajectories, a robust decomposed system control (RDSC) strategy is developed with accurate control and exponential stability. The proposed methodologies are validated on a 3-degrees-of-freedoms (DoFs) manipulator, demonstrating significant efficiency improvements while maintaining high-performance operation. Finally, experiments are conducted on an EMLA test bed under predefined optimal trajectories, simulating dynamic load conditions of the manipulator’s lift joint and controlled with the developed RDSC. The results validate the effectiveness of the optimization framework and the control strategy. Note to Practitioners—The increasing demand for energy efficiency and environmental sustainability is driving the transition in the working machine industry toward electric actuation systems. For heavy-duty mobile manipulators (HDMMs), this means replacing conventional hydraulic and internal combustion engine (ICE)-powered hydraulic actuators with fully-electric electromechanical linear actuators (EMLAs) driven by advanced electric motors such as permanent magnet synchronous motors. This new actuation mechanism represents a clean-tech solution that helps reduce the environmental impact of these working machines. However, one major challenge with EMLA-equipped HDMMs is their limited operating time, constrained by battery capacity. To address this challenge, this paper presents a bilevel optimization framework that focuses on EMLA-driven manipulators, aiming to maximize actuator efficiency while also ensuring that manipulator performance criteria are taken into account. The bilevel optimization aligns the manipulator’s trajectory with the EMLAs’ most efficient operating regions, improving overall actuation system efficiency to 70.3%. This approach ensures lower energy consumption, which could pave the way to extend operational time for the HDMMs. To integrate the control system with the optimization framework for precise trajectory tracking, a robust, adaptive, and modular control design is developed in this paper. The modularity of the control system not only reduces the complexity but also simplifies troubleshooting and upgrades, thereby contributing to lower maintenance costs and facilitating seamless integration into existing industrial setups. Experimental validation on a 1-DoF EMLA test bed confirms the effectiveness of the proposed optimization and control approach.}, archive = {J_TASE}, author = {Mohammad Bahari and Álvaro Paz Anaya and Mehdi Heydari Shahna and Pauli Mustalahti and Jouni Mattila}, doi = {10.1109/TASE.2025.3572307}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16373--16396}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {System-level efficient performance of EMLA-driven heavy-duty manipulators via bilevel optimization framework with a leader-follower scenario}, volume = {22}, year = {2025}, }
• (2025). Adaptive sliding mode control with chebyshev neural disturbance observer for active power filter. TASE, 22, 16355--16372. (www)

  @article{ , abstract = {For the purpose of mitigating harmonics caused by power electronic devices, a novel adaptive sliding mode control (NASMC) strategy utilizing a Chebyshev neural disturbance observer (CNDO) is introduced for controlling an active power filter (APF). By adopting a novel adaptive law, NASMC scheme effectively achieves superior tracking performance while minimizing the impact of chattering, where the arbitrarily small vicinity of sliding manifold is taken into consideration and the derivative of switching gain is divided into three states. Such adaptive law enables extraordinary ability to rapidly adapt and effectively reduce the emergence of chattering and singularity in close proximity to the sliding manifold. Moreover, an adaptive recursive Chebyshev fuzzy neural network (ARCFNN) has been integrated into the disturbance observer to alleviate the dependence on APF parameters. Concurrently, the CNDO can accurately estimate and compensate for inherent modelling nondeterminacy and exterior perturbation within the APF system, thus reducing the vibrating and reinforcing overall system robustness. The proposed control scheme has been effectively validated through a series of simulations and experiments, demonstrating superior performance over other methods. Note to Practitioners—This paper was motivated by the problem of power quality control using active power filter. A novel adaptive sliding mode control strategy utilizing Chebyshev neural disturbance observer is introduced for controlling an active power filter. An adaptive recursive Chebyshev fuzzy neural network (ARCFNN) has been integrated into the disturbance observer. From both simulation and experiments, the proposed CNDO-NASMC strategy has been unequivocally demonstrated to exhibit superior current tracking performance, increased resilience against disturbances, as well as heightened robustness.}, archive = {J_TASE}, author = {Jiahao Yang and Juntao Fei}, doi = {10.1109/TASE.2025.3571991}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16355--16372}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive sliding mode control with chebyshev neural disturbance observer for active power filter}, volume = {22}, year = {2025}, }
• (2025). Bridging in vitro and in vivo: Decellularized plant-based vascular networks for magnetically guided microswarm manipulation. TASE, 22, 16346--16354. (www)

  @article{ , abstract = {An ex vivo platform that can simulate physiological vascular environments while maintaining experimental accessibility is essential for developing effective magnetically guided microswarm delivery systems and transitioning them from a laboratory concept to a practical therapeutic tool. In this article, we present a novel platform that uses decellularized spinach leaves as engineered vascular networks to manipulate magnetic microswarm. Through quantitative characterization, including structural integrity analysis, optical transparency measurements, and perfusion studies, we validate the platform’s suitability for microswarm manipulation. Integrating a customized permanent magnetic control system and machine learning-based visual tracking, we demonstrate four key maneuvers for targeted therapeutic applications within the proposed platform, including forward motion, backward motion, selective branch navigation, and localized aggregation at targeted sites. This physiologically relevant yet accessible platform establishes a crucial bridge between in vitro models and in vivo systems. Specifically, it enables the systematic development, quantitative evaluation, and control strategy implementation of magnetically guided microswarm, potentially accelerating the development of magnetic microswarm-based targeted therapy systems. Note to Practitioners—Magnetic microswarm holds great potential for high-throughput targeted therapies. This article presents an easy-to-process, cost-effective, physiologically relevant, and accessible alternative to facilitate research on microswarm behaviors and accelerate related applications.}, archive = {J_TASE}, author = {Fuzhou Niu and Xinyang He and Quhao Xue and Qing Cao and Hao Yang and Dong Han and Huayong Yang}, doi = {10.1109/TASE.2025.3577478}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16346--16354}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bridging in vitro and in vivo: Decellularized plant-based vascular networks for magnetically guided microswarm manipulation}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based admittance control for physical Human–Robot interaction with output constraints. TASE, 22, 16334--16345. (www)

  @article{ , abstract = {Focused on the scientific issues of collision avoidance and compliant operation of physical human-robot interaction (pHRI) systems, this paper proposes a reinforcement learning (RL) strategy based on admittance control to achieve compliant collision avoidance and accurate trajectory tracking of pHRI. Firstly, a differentiable reference trajectory is generated using a soft saturation function with an admittance model. Subsequently, a reinforcement learning strategy based on an actor-critic structure is implemented to address dynamic uncertainty and enhance tracking performance and compliance. Different from existing studies, a reinforcement learning admittance controller containing an integral barrier Lyapunov function (IBLF) is constructed to attain accurate tracking while ensuring that the end-effector achieves the position constraints. Lyapunov stability theory is employed to proof that all states of the closed-loop system remain semiglobally uniformly ultimately bounded (SGUUB). Finally, a suite of tests on Baxter robot experimental platform have been conducted to validate the superiority of the proposed algorithm compared with adaptive impedance control and conventional admittance control. Note to Practitioners—This paper is motivated by solving the safety problem of interaction between collaborative robots and operators, the proposed method is applicable to service industries and industrial sectors such as medical rehabilitation and collaborative assembly. In current human-robot interaction methods, robots typically operate without positional constraints. This absence of restrictions can lead to potential collisions with obstacles or other participants during collaborative tasks between humans and intelligent robots, raising significant safety concerns. This paper presents a novel control method for human-robot interaction. Firstly, admittance control is employed as a foundational approach to enhance the flexibility of human-robot interactions. Subsequently, a RL strategy is implemented to model the uncertainties in the robotic system and to improve tracking performance during interactions. Lastly, a differentiable soft saturation function is utilized for trajectory shaping, and the output of the physical human-robot interaction (pHRI) system is constrained through the integral barrier Lyapunov function (IBLF) to prevent collisions by regulating the robot’s position. Preliminary experiments on the human-robot interaction platform demonstrate that the proposed method is viable, and the controller has application value in many fields. The proposed controller can be applied in industrial robot assembly tasks to ensure that the industrial robot operates within the designated task area. In certain industrial applications, end-effectors of robots are equipped with hazardous or delicate tools. When the end-effector surpasses the defined task area, it may pose a potential risk of injury to both the human operator and the robotic system. Controller can also be utilized in rehabilitation robots to prevent harm to patients by restricting their movements. In some rehabilitation tasks, the controller parameters could be adjusted in accordance with the patient’s feedback to adapt the autonomy of the rehabilitation task. In future research, we will design control for more complex human-robot interactions, such as human-robot collaborative handling and exoskeleton robots.}, archive = {J_TASE}, author = {Hejia Gao and Yang Yang and Jiangxu Liu and Changyin Sun}, doi = {10.1109/TASE.2025.3576586}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16334--16345}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based admittance control for physical Human–Robot interaction with output constraints}, volume = {22}, year = {2025}, }
• (2025). Adaptive neuro-fuzzy sliding mode tracking for quadrotor UAVs. TASE, 22, 16322--16333. (www)

  @article{ , abstract = {Quadrotor systems offer significant potential, yet precise trajectory tracking remains challenging due to their complex dynamics and nonlinear characteristics. In particular, the unmodeled dynamics of the translational subsystem and the nonlinearity of the rotational subsystem present substantial obstacles. This study proposes a dual-loop structure integrating neural networks (NN) and fuzzy control techniques. By combining the sliding-mode control (SMC) method with adaptive laws, the approach achieves a fast tracking response, improved transient performance, and reduced chattering. An adaptive neural control based on a switching surface addresses the unmodeled dynamics of the outer loop and performs position control. Meanwhile, a dissipative fuzzy sliding-mode controller for the inner loop manages nonlinearity and input uncertainty, ensuring robust attitude control. The designed controllers’ stability was verified using Lyapunov theory, and the simulation of various trajectories demonstrated the robustness and potential of the proposed dual-loop neuro-fuzzy tracking controller. Note to Practitioners—In many real-world applications, PID controllers are often preferred over model-based controllers due to the significant nonlinearity and unmodeled dynamics of quadrotors. To address these challenges and ensure robustness against disturbances, we developed a novel controller that integrates SMC, NN, fuzzy logic, and dissipative performance. Unlike PID controllers, which require extensive tuning and optimization for specific tasks, the adaptive and modeling capabilities of fuzzy logic and NN enable stable flight without additional adjustments. This approach offers a cost-effective solution for managing quadrotors in applications such as search and rescue, transportation, and other tasks involving complex trajectories, unknown dynamics, and external disturbances.}, archive = {J_TASE}, author = {Hyun Duck Choi and Kwan Soo Kim and Peng Shi and Choon Ki Ahn}, doi = {10.1109/TASE.2025.3576292}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16322--16333}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neuro-fuzzy sliding mode tracking for quadrotor UAVs}, volume = {22}, year = {2025}, }
• (2025). Semi-supervised probabilistic learning network for soft sensor modeling with partially labeled data. TASE, 22, 16309--16321. (www)

  @article{ , abstract = {Deep probabilistic learning networks have been applied in industrial soft sensors. However, they face significant challenges in latent variable inference, deep learning backend implementation, and labeled data scarcity. The first challenge arises when covariates directly infer the latent variable, potentially leading to inaccuracies. The second stems from discrepancies between the theoretical probabilistic distribution and practical instance-based deep learning backends. The third is commonly encountered in soft sensor applications, where unlabeled data often go unused, reducing accuracy. To address these challenges, this work proposes a novel semi-supervised probabilistic learning network (SS-PLN) for soft sensor modeling with partially labeled data. The first issue is addressed by formulating an optimization problem as the model’s learning objective. This optimization is efficiently solved by investigating the input of inference network through analyzing the optimal solution’s structure. For the second issue, mean and covariance equations are used to represent probabilistic distributions, ensuring effective deep learning backends. The third issue is addressed by integrating supervised and unsupervised probabilistic learning networks to form the SS-PLN model, thereby maximizing the use of unlabeled data and enhancing the soft sensing performance. The feasibility and effectiveness of the proposed SS-PLN are validated through comparisons with recent semi-supervised learning methods, using data from two industrial processes. Note to Practitioners—Soft sensors are extensively used to predict key variables in industrial processes, especially when direct measurement is challenging or infeasible. While deep learning models have shown promise for this purpose, several challenges remain. Directly inferring latent variables from covariates often results in inaccuracies due to the complexity of their relationships. Additionally, discrepancies between theoretical probabilistic distributions and practical instance-based implementations complicate soft sensor design, particularly under partially labeled data conditions. This paper proposes a novel semi-supervised probabilistic learning network (SS-PLN) to address these challenges. The SS-PLN leverages both labeled and unlabeled data to enhance prediction accuracy. By formulating a learnable optimization problem, the accuracy of inferring latent variable is improved. The mean and covariance expressions are utilized to represent the probabilistic distributions, facilitating numerical implementation. A supervised and an unsupervised probabilistic networks are integrated to construct SS-PLN, enabling effective modeling of partially labeled data. Experimental results from a sugar crystallization process and a steel-making process demonstrate the effective and suitable soft sensing performance of SS-PLN.}, archive = {J_TASE}, author = {Xudong Shi and Ronghuan Li and Humberto Morales and Adriana Amicarelli and Wangya Huang and Weili Xiong}, doi = {10.1109/TASE.2025.3576122}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16309--16321}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semi-supervised probabilistic learning network for soft sensor modeling with partially labeled data}, volume = {22}, year = {2025}, }
• (2025). Smooth surface-to-surface contact control for rope-base soft-tip manipulator. TASE, 22, 16297--16308. (www)

  @article{ , abstract = {A new control pipeline has been proposed for the Rope-Base Soft-tip Manipulator (RBSM) to execute the surface contact task to prevent the jamming and slipping problems. The control pipeline enables smooth surface-to-surface contact for the RBSM using only force sensors, eliminating the dependence on additional pose measurement of the window surface plane and soft-tip deformation information. The pipeline consists of three steps: free contact step implemented by an exponential force shape controller to avoid force overshoot to the window surface; orientation refinement step implemented by a force and torque combined controller to make the RBSM cleaning head surface stable adapt to the smooth window surface; and finally, a release normal force step to reduce head jamming and region covering with a pre-defined vibration-less cleaning trajectory for smooth cleaning on the slippery window surface. The proposed pipeline has been validated in a Rope base Cleaning Manipulator prototype to clean a common window surface. The force and velocity curves during the cleaning experiment show that the proposed method achieves smooth scraping and cleaning under unknown initial significant errors in surface orientation. Note to Practitioners—The current robotic rope base high-rise exterior work, especially cleaning, is not efficient due to the instability of the base and the influence of the soft-tip on the window and end-effector contact. A new control pipeline has been proposed for the Rope-Base Soft-tip Manipulator (RBSM) to clean the window surface smoothly to avoid jamming, slipping, and leaving problems in the contact process. The control pipeline realizes smooth surface-to-surface contact of RBSM with only force sensors without additional pose measurement of the window surface plane. The pipeline consists of three steps: free contact step implemented by an exponential force shape controller to avoid force overshoot to the window surface; orientation refinement step implemented by a force and torque combined controller to make the RBSM cleaning head surface stable adapt to the smooth window surface; and finally, a release normal force step to reduce head jamming and region covering with a pre-defined vibration-less cleaning trajectory for smooth cleaning on the slippery window surface. This setting can effectively guarantee the RBSM contact process and can also avoid physical damage to the window. Therefore, it lays the foundation for autonomous robotic rope base manipulation.}, archive = {J_TASE}, author = {Guangli Sun and Fangxun Zhong and Peng Li and Linzhu Yue and Zhi Chen and Xiang Li and Yun-Hui Liu}, doi = {10.1109/TASE.2025.3574853}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16297--16308}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Smooth surface-to-surface contact control for rope-base soft-tip manipulator}, volume = {22}, year = {2025}, }
• (2025). Event-triggered dynamic output feedback control for switched singular systems with asynchronous switching. TASE, 22, 16287--16296. (www)

  @article{ , abstract = {This manuscript investigates the output-based event-triggered control problem for switched singular systems via the dynamic output feedback controller. To enhance communication efficiency and reduce transmission load, a discrete-time event-triggered scheme based on the output is proposed. Considering the asynchronous switching between the system and the controller due to the event-triggered conditions, the closed-loop system (CLS) is formulated by incorporating an augmented switching signal through the signal merging method. By utilizing the mode-dependent Lyapunov function method and the average dwell time (ADT) technique, sufficient criteria are established to ensure the CLS be regular, impulse-free and E-uniformly ultimately bounded (E-UUB). Additionally, by employing an improved matrix decoupling approach, solvable linear matrix inequality (LMI) conditions are derived, and the co-design strategies for controller gains and event-triggered scheme parameters are presented. Finally, two examples are provided to demonstrate the advantages of the proposed approaches. Note to Practitioners—Switched singular systems are widely applied in current engineering applications, such as power grids, robotics, and intelligent transportation. In network transmission, simultaneous data transmission from a large number of nodes can lead to network delays and packet losses, subsequently degrading system performance and potentially causing instability. By reducing the number of control task executions, event-triggered control can improve communication efficiency. However, event-triggered conditions usually give rise to asynchronous switching between controller and system, which is rarely considered in existing results. Therefore, this paper considers the dynamic output feedback control problem for switched singular systems and proposes an output-based event-triggered control strategy, taking into account the impact of asynchronous switching. Finally, the advantages of the approaches are verified through a circuit example and a numerical example.}, archive = {J_TASE}, author = {Zhuan Liu and Yuchen Sun and Shuping Ma}, doi = {10.1109/TASE.2025.3574527}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16287--16296}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered dynamic output feedback control for switched singular systems with asynchronous switching}, volume = {22}, year = {2025}, }
• (2025). Multi-agent reinforcement learning for connected and automated vehicles control: Recent advancements and future prospects. TASE, 22, 16266--16286. (www)

  @article{ , abstract = {Connected and automated vehicles (CAVs) have emerged as a potential solution to the future challenges of developing safe, efficient, and eco-friendly transportation systems. However, CAV control presents significant challenges significant challenges due to the complexity of interconnectivity and coordination required among vehicles. Multi-agent reinforcement learning (MARL), which has shown notable advancements in addressing complex problems in autonomous driving, robotics, and human-vehicle interaction, emerges as a promising tool to enhance CAV capabilities. Despite its potential, there is a notable absence of current reviews on mainstream MARL algorithms for CAVs. To fill this gap, this paper offers a comprehensive review of MARL’s application in CAV control. The paper begins with an introduction to MARL, explaining its unique advantages in handling complex and multi-agent scenarios. It then presents a detailed survey of MARL applications across various control dimensions for CAVs, including critical scenarios such as platooning control, lane-changing, and unsignalized intersections. Additionally, the paper reviews prominent simulation platforms essential for developing and testing MARL algorithms for CAVs. Lastly, it examines the current challenges in deploying MARL for CAV control, including safety, communication, mixed traffic, and sim-to-real challenges. Potential solutions discussed include hierarchical MARL, decentralized MARL, adaptive interactions, and offline MARL. The work has been summarized in MARL_in_CAV_Control_Repository. Note to Practitioners—This paper explores the application of MARL for controlling CAVs in complex traffic scenarios such as platooning, lane-changing, and intersections. MARL provides an adaptive and decentralized control approach, offering advantages over traditional rule-based and optimization-based methods in handling dynamic interactions, improving traffic flow, enhancing safety, and optimizing fuel efficiency. The paper reviews state-of-the-art MARL algorithms and simulation platforms, serving as a resource for practitioners looking to implement these advanced control strategies. However, real-world deployment remains challenging due to communication reliability, real-time decision-making constraints, and the complexities of mixed traffic environments involving both automated and human-driven vehicles. Future research should focus on ensuring robust inter-agent communication, developing safety-aware MARL frameworks, and addressing the sim-to-real transfer gap. This paper provides insights to help practitioners bridge the gap between research and deployment, facilitating the development of more scalable, adaptive, and reliable CAV control.}, archive = {J_TASE}, author = {Min Hua and Xinda Qi and Dong Chen and Kun Jiang and Zemin Eitan Liu and Hongyu Sun and Quan Zhou and Hongming Xu}, doi = {10.1109/TASE.2025.3574280}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16266--16286}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-agent reinforcement learning for connected and automated vehicles control: Recent advancements and future prospects}, volume = {22}, year = {2025}, }
• (2025). Robust real-time sampling-based motion planner for autonomous vehicles in narrow environments. TASE, 22, 16250--16265. (www)

  @article{ , abstract = {Real-time sampling-based planners increasingly use learned sampling distributions for faster planning in autonomous vehicles. These planners employ a neural network to predict the optimal path and bias some samples toward the path. However, inherent prediction inaccuracies of the network often lead to suboptimal paths, especially in narrow spaces. Learned samples should be used carefully based on accuracy, as inaccurate samples can degrade planning performance. To address this problem, this paper proposes Learned Adaptive Anytime TargetTree-RRT* (LA3T*) algorithm. The proposed planner introduces the adaptive biasing ratio. The approach learns to assess the reliability of the learned distribution using the network’s confidence. This confidence approximates a proper ratio of learned samples used, thereby adaptively maximizing planning performance while considering a level of prediction accuracy. Furthermore, the LA3T* algorithm incorporates the target tree algorithm. The goal pose is replaced with a set (target tree) of pre-defined optimal path segments, reducing computational efforts in narrow regions. Experiments in various driving tasks explore the benefits of each component through ablation studies. The proposed algorithm significantly increases the success rate and reduces the path length in simulated and real-world scenarios compared to other sampling-based methods. Note to Practitioners—This work was motivated by the need to develop a practical real-time motion planning algorithm for autonomous vehicles in narrow environments. Although learning-based sampling methods have been actively explored, their planning performance often degrades due to inaccurate predictions from the learned model, limiting their practicability and robustness. The Learned Adaptive Anytime TargetTree-RRT* (LA3T*) algorithm addresses this problem by adaptively leveraging the learned model based on its reliability and integrating it with the target tree algorithm. Experiments conducted in various driving and parking scenarios demonstrate its practicality and robustness, notably improving success rates in worst-case situations and reducing path length. Furthermore, experiments in sensor noise and dynamic environments—particularly when combined with existing methods—highlight the algorithm’s extensibility. LA3T* can be applied to various autonomous systems needing efficient, real-time motion planning in narrow spaces, such as other autonomous vehicles or robots. The algorithm has not yet been investigated in highly out-of-distribution scenarios. Future work will focus on improving performance in such scenarios and better handling more dynamic environments.}, archive = {J_TASE}, author = {Minsoo Kim and Arthur Esquerre-Pourtère and Jaeheung Park}, doi = {10.1109/TASE.2025.3574262}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16250--16265}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust real-time sampling-based motion planner for autonomous vehicles in narrow environments}, volume = {22}, year = {2025}, }
• (2025). Operation analysis and fault path recognition of complex industrial systems based on boolean networks. TASE, 22, 16237--16249. (www)

  @article{ , abstract = {In modern Industrial Production Processes(IPP), the increasing complexity poses challenges for operation analysis and fault path recognition. This study proposes a dynamic modeling approach using Boolean Networks (BNs) for IPP operation analysis, offering greater dynamism and interpretability compared to existing approaches. Furthermore, different abnormal events lead to the same fault. To bridge this gap, we propose using BNs attractor cycle and semi tensor product (STP) bijective technique to convert fault paths into data sequences. This enables precise identification of fault paths and impacts even when different root causes lead to similar failures. Subsequently, we provide strategies to mitigate fault impacts and use BNs nodes fault propagation topology graphs to prove the interpretability. The results show this method is effective in handling BNs complexity and fault path length. Note to Practitioners—Our work provides plant operators with a dynamic tool for operation analysis and fault path recognition through logical reasoning rather than static rule-based systems. By transforming intricate fault propagation patterns into visualizable data sequences, the method simplifies decision-making and reveals the root causes for maintenance teams. Additionally, formulated criteria for mitigating fault impacts can be applied directly in industrial environments. The proposed topology graphs provide a systematic and visual representation of fault paths, and application in the Preheater case study demonstrates a significant reduction in system complexity and fault path length.}, archive = {J_TASE}, author = {Shenglin Zhang and Yan Wang and Xiang Liu and Zhicheng Ji}, doi = {10.1109/TASE.2025.3574174}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16237--16249}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Operation analysis and fault path recognition of complex industrial systems based on boolean networks}, volume = {22}, year = {2025}, }
• (2025). Versatile bipedal locomotion and walking-running transition: Coordinating supervised learning and nonlinear optimization. TASE, 22, 16224--16236. (www)

  @article{ , abstract = {Online gait planning plays a crucial role for the locomotion of humanoid robots. While simplified models often fail to capture critical dynamic features of the robot’s motion, making online gait modifications with constrained nonlinear optimization in complex models is highly challenging with current computational power. This paper introduces a gait planning method that leverages supervised learning to expedite the gait planning and optimization process. Building upon our previous work, this paper extends a three-body model to the three-dimensional (3D) case. This model incorporates the angular momentum and height variation of body as well as the influence of leg motions, thus facilitating the generation of omnidirectional walking and running patterns. Due to the complexity of the model, an online gait planning modification is impractical. Therefore, supervised-learning is employed to train a policy derived from the model-based gait planning approach. This policy is then implemented online to produce versatile locomotion. Furthermore, the gradient of the trained neural network is utilized for nonlinear optimization of gait parameters, significantly improving the robot’s balance against external perturbations. The effectiveness of the proposed method is validated through a series of experiments conducted in simulation and on the real robot BHR-T, confirming its capability to generate adaptive walking and running motions in response to varying demands and disturbances. Note to Practitioners—The gait planning method presented in this paper addresses the practical challenge of real-time locomotion control in humanoid robots. This approach is beneficial for applications requiring quick adaptability to external disturbances and varying terrain. By integrating supervised learning with model-based nonlinear optimization, this method allows for the efficient generation and optimization of stable walking and running patterns. The trained neural network not only enhances the robot’s balance but also reduces the computational burden, making it feasible to implement in real-world scenarios. This method has been validated in both simulations and real-world tests, demonstrating its potential to improve the performance of humanoid robots in practical deployments.}, archive = {J_TASE}, author = {Huanzhong Chen and Gao Huang and Xuechao Chen and Zhangguo Yu and Chencheng Dong and Qingqing Li and Qiang Huang}, doi = {10.1109/TASE.2025.3573028}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16224--16236}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Versatile bipedal locomotion and walking-running transition: Coordinating supervised learning and nonlinear optimization}, volume = {22}, year = {2025}, }
• (2025). Optimal h∞ event-triggered control for markov jump linear systems based on actor-critic Q-learning approach. TASE, 22, 16213--16223. (www)

  @article{ , abstract = {In this paper, the optimal ${\mathcal {H}}_{\infty } $ control problem is investigated for a class of Markov jump linear systems, under event-triggering mechanism. To this end, a two-player zero-sum differential game-theoretic problem is formulated and a mode-dependent event-triggered version of Hamiltonian function is established and compared with its time-triggered counterpart. Based on this, the corresponding triggering event is given to ensure the stochastic stability of the considered system with a predefined ${\mathcal {H}}_{\infty } $ noise attenuation level index and the existence of a saddle point for the defined cost function is proved. The difference in the optimal cost between the time-triggered case and the event-triggered one is also shown. Further to derive an online solution without the full knowledge of the system model, a Q-learning framework is adopted to approximate the optimal cost and the optimal event-triggered control law combined with an actor-critic network. The error dynamics of the learning algorithm are analyzed and the convergence is proved. The stochastic stability of the resulting closed-loop system is also confirmed. Finally, a numerical example is given to show the effectiveness of the proposed results. Note to Practitioners—The wide applications of Markov jump system have gained its reputation in modeling systems with random changes from internal structure or external parameters. This work is motivated by the optimal ${\mathcal {H}}_{\infty } $ event-triggered control problem for Markov jump system. Existing results are mostly based on linear matrix inequality approach which is strictly model-based and offline thus vulnerable to modeling uncertainties. The optimality of the event-triggered controller has rarely been discussed. In this paper we adopt a Q-learning based actor-critic approach to perform the optimal ${\mathcal {H}}_{\infty } $ event-triggered controller design with respect to a user-defined triggering condition. The stability of the closed-loop system is proved as well as the convergence of the learning algorithm. It is data-based and easy to be implemented online without the full knowledge of the system parameters. Examples are given in the end to demonstrate the performance of the obtained results. In future research, network-induced disturbances such as time-delays, malicious attacks will be considered and discussed under the proposed method.}, archive = {J_TASE}, author = {Zhaowen Xu and Zheng-Guang Wu}, doi = {10.1109/TASE.2025.3573908}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16213--16223}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal h∞ event-triggered control for markov jump linear systems based on actor-critic Q-learning approach}, volume = {22}, year = {2025}, }
• (2025). Robust model predictive control of position and orientation of redundant manipulators. TASE, 22, 16201--16212. (www)

  @article{ , abstract = {Redundant manipulators are explored and utilized across a wide range of domains. Investigating effective trajectory tracking of redundant manipulators is pivotal for their practical applications. This paper constructs a model predictive control of position and orientation (MPCPO) scheme, which controls the position and orientation of the end-effector simultaneously while minimizing norms of the tracking error, joint velocity, and joint acceleration. Additionally, the MPCPO scheme addresses different levels of physical constraints while preserving the feasible region of decision variables. Moreover, an integration enhanced gradient-based recurrent neural network (IEGRNN) model is designed to solve the MPCPO scheme in noisy environments. Theoretical analyses demonstrate the convergence and robustness of the IEGRNN model. The effectiveness, accuracy, flexibility, and noise tolerance of the MPCPO scheme solved by the IEGRNN model are illustrated through simulations and experiments. Note to Practitioners—This paper is motivated by the goal of enhancing trajectory tracking performance in redundant manipulators. Many schemes only consider controlling the position of the end-effector of a redundant manipulator while neglecting orientation. Besides, the commonly used schemes address multiple levels of physical constraints by using conversion techniques, which diminishes the feasible region of decision variables. To overcome these weaknesses, a model predictive control of position and orientation (MPCPO) scheme is constructed in this paper, which controls the position and orientation of the end-effector simultaneously. Additionally, the MPCPO scheme addresses different levels of physical constraints while preserving the feasible region. Moreover, most existing MPC solvers do not account for the influence of noises, which reduces their reliability in real-world scenarios. To remedy this, an integration enhanced gradient-based recurrent neural network (IEGRNN) model is designed to solve the MPCPO scheme, with theoretical guarantees on its convergence and robustness in noisy environments. Simulations and experiments validate the effectiveness of the proposed method. Its superiority is further demonstrated through comparisons with other schemes.}, archive = {J_TASE}, author = {Jinfu Tang and Zhenming Su and Juntao Liu and Long Jin}, doi = {10.1109/TASE.2025.3572441}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16201--16212}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust model predictive control of position and orientation of redundant manipulators}, volume = {22}, year = {2025}, }
• (2025). Online receding horizon safe-critical control for complex nonlinear system under uncertainty: A biological brain-inspired dual learning approach. TASE, 22, 16185--16200. (www)

  @article{ , abstract = {This research introduces a real-time, computationally efficient receding horizon control (RHC) framework with adaptive safety mechanisms under uncertainty. Achieving real-time optimal performance and safety with uncertain nonlinear systems and environments is a challenging task due to the heavy computational complexity involved in learning optimal solutions and managing system uncertainties. To address this, a novel RHC-based safe critical mechanism is designed, enhancing the classical RHC by integrating performance with a real-time safety framework that adapts to environmental uncertainties. Particularly, a human brain-inspired dual-learning algorithm is developed with a slow learning phase that gradually learns the system and environmental uncertainties and further achieves the optimal RHC solution by employing a situation-aware physics-informed neural network (SA-PINN). A fast-learning mechanism is then developed to maintain system safety with quick decision-making capabilities, utilizing a fast-learned adaptive control barrier (FLA-CBF) function that incorporates an adaptive bound relaxation. This approach enables the slow-learning phase to gradually learn the optimal solutions and adapt to uncertainties, thus enhancing the fast-learning phase, which supports real-time control execution with guaranteed safety. In addition, Lyapunov stability analysis and a series of experimental results have been provided to showcase the efficacy of the developed algorithm. Note to Practitioners—This paper investigates the critical challenge of ensuring safety and performance in real-time control of complex nonlinear systems with uncertainties, emphasizing the need to balance control effectiveness and safety in dynamic real-world environments. The approach detailed in the paper has potential applications across various fields that involve dynamic systems needing robust safe control and performance under uncertainty, like robotics, autonomous vehicles, and aerospace engineering. A key potential of this research lies in its novel dual-learning algorithm, which mimics human cognitive processes to balance safety and performance dynamically. This is particularly advantageous in environments where fast safety decisions are needed, and system responsiveness is critical. Here, the optimal control and model learning from the slow learning part are transferred to the fast learning phase, which quickly adapts to environmental changes using a modified CBF, known as FLA-CBF, with an initial strict safety threshold that is gradually relaxed. However, the selection of this bound threshold depends on the environment and requires further exploration. Also, the computation intensity of the algorithm needs to be improved for real-time application, especially in scenarios involving multiple agents.}, archive = {J_TASE}, author = {Shawon Dey and Hao Xu}, doi = {10.1109/TASE.2025.3567610}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16185--16200}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online receding horizon safe-critical control for complex nonlinear system under uncertainty: A biological brain-inspired dual learning approach}, volume = {22}, year = {2025}, }
• (2025). A nonlinear elasticity model and feedforward compensation method to increase positioning accuracy of industrial robots. TASE, 22, 16168--16184. (www)

  @article{ , abstract = {Large industrial robots are inexpensive compared to their workspace. Due to flexibility, the positioning accuracy is compromised. This paper presents a simple yet effective static calibration method for robots using a laser tracker that can be applied to processes with no change in external load and without high processing forces. The calibration method is based on a novel elastokinematic model that captures the nonlinear gearbox and link compliance and also accounts for the effects of pose dependent joint friction and nonlinearities due to the gravity compensation mechanism. Furthermore, a feedforward concept is presented that separates kinematics from compliance to reduce computation time. The calibration method was implemented on industrial hardware and experimentally validated on a KUKA KR210–2. In comparison to state of the art compliance modeling, accuracy is improved by 23.1 %. Finally, the feedforward model reduces the mean positioning error in the entire workspace by 54.9 % compared to the purely kinematically calibrated model, emphasizing the need for compliance compensation. Note to Practitioners—Industrial processes such as bonding, welding, quality inspection, and aircraft assembly require high path and positioning accuracy. This is often accomplished by programming paths offline. Due to inaccuracies, these programs often require manual corrections, which is undesirable in an era when production aims for greater customization. Another possibility is to utilize additional, costly sensors. Especially for large industrial robots, the inaccuracy increases as the lever arms and component masses increase due to the additional load on the drivetrains. Therefore, robot manufacturers specify repeatability rather than absolute positioning accuracy, because these effects depend on the robot’s pose. To overcome this issue, a practical compensation technique that minimizes the pose dependent error is presented. This is achieved by using a laser tracker to identify an improved model of the robot. The calibration procedure can then be automated without the need for any information other than that provided in the robot’s manual. Finally, the compensation algorithm is numerically inexpensive, enabling it to be implemented in robot control when there is a possibility of manipulating the path planning angles. A short step by step explanation is given on how to implement our methods for other robots. With that, it is possible to expand the spectrum of robotic applications without external sensors, where no large process forces act on the axes.}, archive = {J_TASE}, author = {Marcel Dzubba and Michael Neubauer and Christoph Hinze and Armin Lechler and Alexander Verl}, doi = {10.1109/TASE.2025.3559673}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16168--16184}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A nonlinear elasticity model and feedforward compensation method to increase positioning accuracy of industrial robots}, volume = {22}, year = {2025}, }
• (2025). Fault-estimation and fault-tolerant control for T-S fuzzy descriptor systems with disturbances. TASE, 22, 16157--16167. (www)

  @article{ , abstract = {The article investigates the problems of sensor fault estimation and fault-tolerant control for T-S fuzzy descriptor systems with disturbance. A novel regular form is devised for the fault estimation of T-S fuzzy descriptor systems, transforming the systems into a block structure. This transformation not only separates the observable and unobservable parts but also converts the singular matrix of the systems into a symmetric form, thereby simplifying the observer design and eliminating the equality constraint arising from the presence of the singular matrix. Subsequently, a controller in terms of the sliding mode observer is developed. Sufficient conditions are provided by the linear matrix inequality method for achieving simultaneous estimation of sensor faults and system states, as well as ensuring the stability of the overall closed-loop system. Furthermore, the sliding mode input of the sliding mode observer is designed to satisfy the reachability condition. Finally, three simulation examples are provided that illustrate the effectiveness of the proposed results. Note to Practitioners—Safety-critical systems (e.g., aerospace and nuclear power) require that the control systems can handle various abnormal conditions, including failures, loss of control, and equipment damage. The focus of this article is on fault estimation and fault-tolerant control, aiming to ensure acceptable performance and stability levels for safety-critical systems in the event of a fault occurrence. T-S fuzzy descriptor model is considered, T-S fuzzy model can convert intricate nonlinear problems into linear ones, and the descriptor form can encompass a broader spectrum of dynamic models than those of normal systems by taking into account the static constraints among variables. The new regular form proposed in this article ensures the feasibility of the sliding mode observer for T-S descriptor systems, and the method is also suitable for normal systems. Moreover, the reduced-order characteristic of the designed observer is advantageous for increasingly complex industrial systems.}, archive = {J_TASE}, author = {Longfei Yu and Kaiqing Bu and Zequn Wu}, doi = {10.1109/TASE.2025.3575804}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16157--16167}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-estimation and fault-tolerant control for T-S fuzzy descriptor systems with disturbances}, volume = {22}, year = {2025}, }
• (2025). Machine learning-based real-time walking activity and posture estimation in construction with a single wearable inertial measurement unit. TASE, 22, 16144--16156. (www)

  @article{ , abstract = {Construction workers regularly perform walking locomotion on level and inclined surfaces. It is critical to detect walking activity and estimate body postures in real time for monitoring workers’ safety and health conditions. This article presents a machine learning-based framework for real-time activity detection and posture estimation during human walking on level and sloped terrains using a single wearable inertial measurement unit (IMU). The framework integrates recurrent neural networks with Gaussian process dynamical models to achieve accurate predictions of walking activity, floor slope angles, and workers’ turning angles and full-body limb joint angles estimation in real time. The proposed design offers a streamlined, cost-effective solution with significant advantages over multi-sensor systems. Extensive experiments of different walking activities on level and sloped surfaces are conducted to validate and demonstrate the design. The proposed algorithm detects gait activities with 96% accuracy, the estimated human limb joint angle errors are within 11 deg, the predicted turning angles have an error less than 16 deg and the end-to-end detection latency is within 21 ms using only one single IMU attached to the human shank. Note to Practitioners—Construction workers exert intense physical effort, and they experience serious safety and health risk in hazardous, dynamic working environments. As a result, the construction industry is one of the highest-risk industrial sectors in most countries. Real-time monitoring of workers’ walking gait plays a critical role in construction safety. Wearable IMUs are particularly attractive for walking activity recognition because they are small, inexpensive, and non-intrusive. This study aims to develop machine learning-enabled, real-time IMU-based walking activity recognition and full-body posture and floor slope angle estimation. The main approach is built on long short-term memory (LSTM)-based recurrent neural networks and a manifold learning method to process time-series data from the IMU to predict human motion. The LSTM model achieves over 96% accuracy to classify various walking activities on different floor slopes. Additionally, the design also estimates human limb joint angles, floor slope angle and the worker’s body turning angle. A noteworthy aspect of the detection system is the minimal detection latency of 18 ms, ensuring the reliability and effectiveness of real-time monitoring and evaluation. This feature is particularly beneficial for immediate feedback and intervention systems that help protect workers from work-related injuries. The simplicity and efficiency of using a single IMU are attractive for a practical solution for real-time automation applications in dynamic environments.}, archive = {J_TASE}, author = {Siyu Chen and Chunchu Zhu and Xunjie Chen and Jingang Yi}, doi = {10.1109/TASE.2025.3573966}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16144--16156}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Machine learning-based real-time walking activity and posture estimation in construction with a single wearable inertial measurement unit}, volume = {22}, year = {2025}, }
• (2025). Conscious frontier analysis model: Learning-based path planning for PointGoal navigation. TASE, 22, 16131--16143. (www)

  @article{ , abstract = {The objective of PointGoal navigation is to guide mobile robot to their destination in the shortest distance possible without a prioir map. However, due to the environment uncertainty, finding a solution is complicated. To address this issue, we propose a conscious frontier analysis model based on a partially observable Markov decision process. Our model uses synthetic maps for training and selects frontiers (points between known and unknown areas). By leveraging a reinforcement learning framework for path planning, the model efficiently navigates without prior mapping, enhancing exploration and pathfinding capabilities to reach the PointGoal in unknown environments. We utilize a modified belief-based version of the Bellman equation to assess the cost of failing to reach the goal, enabling the selection of frontiers with the least associated cost. Additionally, we train a modified ResNet18 model to identify the cost and valuable properties of each frontier. We evaluated our model’s performance using the large-scale RGB-D indoor dataset Matterport3D on the Habitat simulator, achieving a 90.7% successful completion of the path in the environment, outperforming traditional frontier-based model by 27.9%, active neural SLAM by 14.3%, and learning augmented model by2.8%. In certain paths, our model outperformed the Habitat simulator’s state-of-the-art baseline normalized path length by 3.6%. Also, we reckon the efficacy of the model on the real robot, and it proved to be effective. This model supports 3D Lidar, 2D Lidar, and RGB-D perception, enabling exploration in complex environments before pinpointing the goal frontier, while synthetic maps reduce sampling biases in new environments. Note to Practitioners—This paper tackles the problem of navigating mobile robot in environments where there are no pre-existing maps, which is common in fields like surveillance, search-rescue operations, and mining. Traditional methods often struggle with efficiency and reliability in these settings. Our proposed conscious frontier analysis model solves this by using synthetic maps for training and a reinforcement learning framework. This allows the robot to explore and navigate by selecting optimal frontiers, which are transitional points between known and unknown areas. This method improves pathfinding efficiency and helps the robot handle the uncertainties in the environments. Key practical benefits include more efficient navigation, leading to faster and more reliable operations. The model is versatile, working with various perception systems like 3D LIDAR, 2D LIDAR, and RGB-D, making it adaptable to different robots and environments. It also improves exploration capabilities, enabling the robot to gather information efficiently before deciding on a path, ensuring better decision-making. While the model shows promising results, it depends heavily on the quality of synthetic maps and needs further optimization for specific environments. Future research could refine training methods and adapt the model for various practical applications. By integrating this model into existing robotic systems, practitioners can significantly improve the navigation capabilities of autonomous robots, boosting productivity and operational efficiency in dynamic and uncertain environments.}, archive = {J_TASE}, author = {Shri Harish Manoharan and Wei-Yu Chiu and Chao-Tsung Huang}, doi = {10.1109/TASE.2025.3573814}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16131--16143}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Conscious frontier analysis model: Learning-based path planning for PointGoal navigation}, volume = {22}, year = {2025}, }
• (2025). A self-triggered adaptive dynamic programming scheme for unknown nonlinear systems by iterative model predictive process. TASE, 22, 16121--16130. (www)

  @article{ , abstract = {This article designs a self-triggered adaptive dynamic programming (STADP) algorithm combined with the model predictive control (MPC) mechanism to address the approximate optimal control of a nonlinear system with terminal state constraints. The MPC mechanism transforms the problem into a series of subproblems and then reduces the number of subproblems through the self-triggered mechanism (STM), thereby reducing the amount of computation. The MPC-based STADP (MSTADP) algorithm employs neural networks (NNs) to construct three key modules: the model module, the critic module, and the action module. These modules enable the identification of the unknown system, approximation of the cost function, determination of the terminal penalty, and estimation of the optimal control strategy set. Furthermore, a relaxation factor is incorporated to modulate the convergence rate of the system states online. Finally, experimental validation on two distinct systems demonstrates the algorithm’s effectiveness, and comparative experiments against the traditional HDP algorithm are performed to highlight the superiority of the proposed approach. Note to Practitioners—Industrial processes often exhibit nonlinear nonaffine behavior, posing challenges in solving Hamilton-Jacobi-Bellman (HJB) equations for nonlinear MPC. Hence, leveraging ADP to address the nonlinear HJB equations presents significant value for solving the nonlinear MPC problem. This paper presents the MSTADP algorithm to enhance predictive control in addressing unknown system dynamics within real-world industrial processes. Given the complexity of establishing mathematical models for such systems, we construct a model network using a batch of data, serving as the prediction model. Then, MSTADP mechanism is employed to determine the predictive control input under STM. Simulation experiments validate the proposed algorithm’s effectiveness, supported by a flow chart and relevant experimental parameters. The results of the experiments reveal the scheme’s strong performance.}, archive = {J_TASE}, author = {Kun Zhang and Qi Zhu and Xiang-Peng Xie and Rong Su}, doi = {10.1109/TASE.2025.3571535}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16121--16130}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A self-triggered adaptive dynamic programming scheme for unknown nonlinear systems by iterative model predictive process}, volume = {22}, year = {2025}, }
• (2025). Estimation and prediction of CoM with terrain feature embedding during walking. TASE, 22, 16108--16120. (www)

  @article{ , abstract = {Wearable devices are currently being used to reduce metabolism and assist people with disabilities for daily walking. For the elderly and disabled people, improving the walking stability of wearable devices is a crucial and unsolved research. In particular, the center of mass (CoM) trajectory can reflect the walking state as well as the stability of a person. For this reason, realizing the prediction of CoM trajectory under daily walking is promising to improve the wearable devices. In this article, a method for estimation and prediction of CoM during daily walking was proposed. A visual-inertial-odometry algorithm was used to obtain history CoM trajectories during walking, and the depth information from camera data was extracted by sequential distance embedding method and encode the information into terrain vectors. The trajectory vectors were patched with the corresponding terrain vectors and then realized the fusion of multi-modal data, which were then fed into a well-trained temporal convolution network to output the prediction results. With data from two outdoor datasets, the method in this paper was verified to be able to be used for CoM trajectory prediction for a variety of walking tasks and consume low computational cost. This method has the potential to be extensible in improving the stability of wearable devices, such as exoskeleton, powered prosthetic, and so on. Note to Practitioners—It is necessary to obtain the future state of human center of mass (CoM) to improve the stability of user with wearable devices. However, current methods for CoM trajectory prediction, especially with multi-modal data fusion are difficult to apply to this prediction task due to the high computational cost. This article addressed the problem of estimation and prediction of CoM under daily walking conditions and proposed a method to solve the above problems, which enable to deploy on wearable robots with limited computation resources. In our study, a visual-inertial-odometry algorithm was utilized to estimate the history CoM and ran on our edge device in 30 Hz. And a terrain feature embedding method was proposed, which was able to extract the terrain depth information from the camera and reduce the dimensionality of the camera data. With the help of the embedding technique, terrain feature could be efficiently fused with history CoM trajectories in a low-cost way and handle the terrain transition situation in the daily walking environment. Finally, a temporal convolution network was designed to generate the predicted results based on the fusion of history CoM trajectories and terrain vectors in average of 8ms. This is the first study to enable CoM trajectory prediction under multiple walking conditions using cameras and IMU sensors via deep learning. However, the VIO will be affected by the camera data quality under low-light conditions and our prediction only takes the height as the main factor affecting the CoM trajectory. We hope that this research will provide some insights to wearable devices. For example, it may serve as a high-level controller for exoskeleton or prostheses to generate natural motion by means of kinematic model.}, archive = {J_TASE}, author = {Haolan Xian and Jingfeng Xiong and Yuanwen Zhang and Xinxing Chen and Chenglong Fu and Yuquan Leng}, doi = {10.1109/TASE.2025.3570875}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16108--16120}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Estimation and prediction of CoM with terrain feature embedding during walking}, volume = {22}, year = {2025}, }
• (2025). A nonlinear control allocation strategy for dual half bridge power converters. TASE, 22, 16091--16107. (www)

  @article{ , abstract = {This work focuses on designing nonlinear control algorithms for dual half-bridge converters (DHBs). We propose a two-layer controller to regulate the current and voltage of the DHB. The first layer utilizes a change in the control variable to obtain a quasi-linear representation of the DHB, allowing for the application of simple linear controllers to regulate current and power flow. The second layer employs a nonlinear control allocation algorithm to select control actions that fulfill (pseudo) power setpoints specified by the first control layer; it also minimizes peak-to-peak currents in the DHB and enforces voltage balance constraints. We apply the DHB and this new control strategy to manage power flow in a hybrid energy storage system comprising of a battery and supercapacitors. Numerical simulation results demonstrate that, in comparison with state-of-the-art approaches, our control algorithm is capable of maintaining good transient behavior over a wide operating range, while reducing peak-to-peak current by up to 80%. Note to Practitioners—Dual half bridges (DHB) are an important class of converters that is used in a wide range of practical applications, including automotive and renewable energies. Despite their benefits—such as galvanic isolation, zero-voltage switching, and bidirectional power flow—the design of control algorithms for this converter is challenging due to its highly nonlinear response. In this work, we provide a new control framework that allows the designer to obtain a linear input-to-state representation of the converter and to “cancel” the dominant nonlinearities. The paper also presents the application of the parameter space tuning technique, which allows the designer to synthesize DHB controllers that fulfill transient, disturbance rejection, and robustness requirements. The tuning approach provides a set of possible controller gains that fulfill the control specifications (and not just a single value). This enables the application engineer to perform fine-tuning of the controller gains within a pre-certified parameter space that satisfies the control requirements, providing another important practical benefit.}, archive = {J_TASE}, author = {Ricardo de Castro and Rui Esteves Araujo and Jonathan Brembeck}, doi = {10.1109/TASE.2024.3442078}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16091--16107}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A nonlinear control allocation strategy for dual half bridge power converters}, volume = {22}, year = {2025}, }
• (2025). Dynamic approaches for finding least squares solution of time-varying multi-linear systems. TASE, 22, 16079--16090. (www)

  @article{ , abstract = {The primary focus of this study is to estimate least squares solutions for time-varying multi-linear systems (TVMLS) that involve time-varying (TV) perturbations on the right-hand side. To address this issue, we introduce two neural network models: the hybrid neural network (HNN) and a nonlinear enhancement of the HNN design that uses a modified weighted sign-bi-power (Mwsbp) activation function, referred to as MwsbpHNN. Theoretical findings demonstrate the convergence of both the HNN and MwsbpHNN dynamic models to the exact solutions of underlying TVMLS. Furthermore, we rigorously prove a fixed-time convergence of MwsbpHNN and provide upper bounds for the convergence time. Computational complexity of our theoretical framework is compared with current state-of-the-art methods for solving TVMLS, highlighting its efficiency. Numerical simulation further confirms the robustness of the proposed HNN and MwsbpHNN models. A model arising from discretized high-order Bellman equation is considered. Note to Practitioners—This research develops neural network models designed to compute least squares solutions of time-varying multi-linear systems (TVMLS) with time-varying perturbations on the right-hand side. A comprehensive convergence analysis reveals a predefined time convergence to a least squares solution of the TVMLS by the hybrid neural network (HNN) that is activated by the modified weighted sign-bi-power (Mwsbp) function. Additionally, a range of numerical simulations validate theoretical findings. We also provide rigorous theoretical and comparative simulation results to showcase the performance of the proposed models. Additionally, we analyze a model derived from a discretized high-order Bellman equation, serving as a fundamental theoretical framework in areas such as economics, engineering control theory, and reinforcement learning.}, archive = {J_TASE}, author = {Xuezhong Wang and Predrag S. Stanimirović and Yimin Wei}, doi = {10.1109/TASE.2025.3575246}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16079--16090}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic approaches for finding least squares solution of time-varying multi-linear systems}, volume = {22}, year = {2025}, }
• (2025). MEF-explore: Communication-constrained multi-robot entropy-field-based exploration. TASE, 22, 16062--16078. (www)

  @article{ , abstract = {Collaborative multiple robots for unknown environment exploration have become mainstream due to their remarkable performance and efficiency. However, most existing methods assume perfect robots’ communication during exploration, which is unattainable in real-world settings. Though there have been recent works aiming to tackle communication-constrained situations, substantial room for advancement remains for both information-sharing and exploration strategy aspects. In this paper, we propose a Communication-Constrained Multi-Robot Entropy-Field-Based Exploration (MEF-Explore). The first module of the proposed method is the two-layer inter-robot communication-aware information-sharing strategy. A dynamic graph is used to represent a multi-robot network and to determine communication based on whether it is low-speed or high-speed. Specifically, low-speed communication, which is always accessible between every robot, can only be used to share their current positions. If robots are within a certain range, high-speed communication will be available for inter-robot map merging. The second module is the entropy-field-based exploration strategy. Particularly, robots explore the unknown area distributedly according to the novel forms constructed to evaluate the entropies of frontiers and robots. These entropies can also trigger implicit robot rendezvous to enhance inter-robot map merging if feasible. In addition, we include the duration-adaptive goal-assigning module to manage robots’ goal assignment. The simulation results demonstrate that our MEF-Explore surpasses the existing ones regarding exploration time and success rate in all scenarios. For real-world experiments, our method leads to a 21.32% faster exploration time and a 16.67% higher success rate compared to the baseline.Note to Practitioners—This paper is motivated by the problem of multi-robot exploration in unknown environments with communication constraints regarding information sharing between robots. The majority of existing methods have yet to consider that robots cannot always fully communicate when working in actual situations. To fill this gap, in this paper, we propose a new exploration method in which robots can share necessary information based on their communication status and explore the area efficiently based on the entropies of frontiers and robots. Experimental evaluation shows that the proposed approach leads to fast and successful exploration, which will be practical for real-world exploration tasks, such as in large buildings, warehouses, urban areas, and complex environments.}, archive = {J_TASE}, author = {Khattiya Pongsirijinda and Zhiqiang Cao and Billy Pik Lik Lau and Ran Liu and Chau Yuen and U-Xuan Tan}, doi = {10.1109/TASE.2025.3575237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16062--16078}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MEF-explore: Communication-constrained multi-robot entropy-field-based exploration}, volume = {22}, year = {2025}, }
• (2025). Robust whole-body safety-critical control for sampled-data robotic manipulators via control barrier functions. TASE, 22, 16050--16061. (www)

  @article{ , abstract = {In this paper, a novel robust safety-critical control method is proposed to ensure whole-body safety for the robotic manipulator, which is implemented as a sampled-data system with measurement errors. The manipulator and obstacles are approximated as several spherical enclosures, and a whole-body safety constraint with relative degree two is formulated based on the distance function. Robust control barrier function (CBF) constraints are first designed to handle the predefined joint velocity constraints. Building upon this, a robust high-order CBF constraint is derived to enforce the whole-body safety constraint. Each stage of the derivation incorporates the sample-and-hold error and measurement error. These robust CBF constraints are then unified with a nominal controller to form an optimization problem, ensuring that the velocity constraints and the safety constraint are satisfied. The effectiveness of the proposed algorithm is demonstrated through simulations and experiments on a 7-degree-of-freedom (DOF) Franka Emika Panda robot. Note to Practitioners—This paper presents a robust control approach aimed at addressing whole-body safety for robotic manipulators, ensuring that all parts of the robot operate safely without violating safety constraints. The manipulator and obstacles are modeled as spherical enclosures, which simplifies the formulation of the whole-body safety constraint and enhances computational efficiency. Furthermore, by introducing robust CBFs, this work effectively addresses errors caused by sampling and measurement to maintain safe operation even under real-world conditions. The robust design reduces the reliance on specific parameters, thereby offering greater flexibility in handling diverse safety constraints and improving adaptability to various safety-critical situations. This approach is expected to significantly improve whole-body safety in practical deployment, especially in dynamic or unstructured environments.}, archive = {J_TASE}, author = {Yuhan Xiong and Di-Hua Zhai and Yuanqing Xia}, doi = {10.1109/TASE.2025.3574342}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16050--16061}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust whole-body safety-critical control for sampled-data robotic manipulators via control barrier functions}, volume = {22}, year = {2025}, }
• (2025). Explicit-implicit subgoal planning for long-horizon tasks with sparse rewards. TASE, 22, 16038--16049. (www)

  @article{ , abstract = {The challenges inherent in long-horizon tasks in robotics persist due to the typical inefficient exploration and sparse rewards in traditional reinforcement learning approaches. To address these challenges, we have developed a novel algorithm, termed hlexplicit-implicit subgoal planning (EISP), designed to tackle long-horizon tasks through a divide-and-conquer approach. We utilize two primary criteria, feasibility and optimality, to ensure the quality of the generated subgoals. EISP consists of three components: a hybrid subgoal generator, a hindsight sampler, and a value selector. The hybrid subgoal generator uses an explicit model to infer subgoals and an implicit model to predict the final goal, inspired by way of human thinking that infers subgoals by using the current state and final goal as well as reason about the final goal conditioned on the current state and given subgoals. Additionally, the hindsight sampler selects valid subgoals from an offline dataset to enhance the feasibility of the generated subgoals. While the value selector utilizes the value function in reinforcement learning to filter the optimal subgoals from subgoal candidates. To validate our method, we conduct four long-horizon tasks in both simulation and the real world. The obtained quantitative and qualitative data indicate that our approach achieves promising performance compared to other baseline methods. These experimental results can be seen on the website https://sites.google.com/view/vaesi Note to Practitioners—This paper addresses the persistent challenges in executing long-horizon tasks in robotics, which are hindered by inefficient exploration and sparse rewards in traditional reinforcement learning methods. EISP overcomes these obstacles through the integration of a hybrid subgoal generator, a hindsight sampler, and a value selector, ensuring that the generated subgoals are both feasible and optimal, thereby facilitating the completion of the entire task by addressing each subgoal. Although preliminary results are promising, several challenges must be resolved in future research. Firstly, the automatic determination of the number of subgoals during training presents a challenge that may impact the low-level policy’s efficacy in achieving each subgoal. Secondly, the proposed method relies on an offline dataset that lacks successful trajectories from the initial state to the final state, complicating the exploration of subgoals towards the final goal. Finally, the scalability across multiple tasks is relatively limited due to the time-intensive nature of data collection for each task. Future research will focus on optimizing the allocation of timesteps to subgoals, enhancing exploration efficiency, and improving scalability across a broader range of tasks without extensive data collection.}, archive = {J_TASE}, author = {Fangyuan Wang and Anqing Duan and Peng Zhou and Shengzeng Huo and Guodong Guo and Chenguang Yang and David Navarro-Alarcon}, doi = {10.1109/TASE.2025.3574162}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16038--16049}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Explicit-implicit subgoal planning for long-horizon tasks with sparse rewards}, volume = {22}, year = {2025}, }
• (2025). MGTC: Multi-granularity temporal aware time series classification. TASE, 22, 16026--16037. (www)

  @article{ , abstract = {Self-supervised learning (SSL) can extract useful temporal representations for time series classification (TSC) tasks. However, existing methods with subsequence and instance-level augmentation lead to the loss of global information and inductive bias. Meanwhile, neglecting multi-granularity temporal representations also poses challenges for modeling complex temporal structures and relationships. Therefore, in this work, we propose a multi-granularity temporal-aware time series classification method, called MGTC, that learns multi-granularity temporal representations, enhancing the ability to perceive class discrepancies. Specifically, we propose a multi-density masking strategy to adapt the dynamic time-varying patterns for learning comprehensive temporal representations. Next, we employ cluster-wise constraint to hierarchically aggregate these representations at the instance level. Finally, we design two self-supervised tasks: i) granularity-aware contrastive learning, to extract intra-instance fine-grained temporal structures and inter-instance coarse-grained class relationships, and ii) cross-view prediction pretext task, to capture global contextual temporal dependencies. We conducted comprehensive experimental evaluations on various types of dataset, and the experimental results validated the effectiveness of our method. The code is publicly available at https://github.com/mrxiliang/MGTC Note to Practitioners—Time series classification (TSC) is widely used in many real-world applications, such as industrial fault diagnosis and healthcare monitoring. To guarantee classification accuracy, the TSC models must have high representation learning ability. However, many existing methods fail to capture global information and different granularities of temporal relations, hindering accurate classification. To address this challenge, this paper proposes a multi-granularity temporal-aware time series classification method (MGTC), capturing multi-granularity temporal representations from unlabeled data, enhancing the ability to perceive class discrepancies. The proposed multi-density masking strategy could adapt to the time-varying patterns for fully capturing temporal representations. The cluster-wise constraint could make the learned representations more compact in the embedding space. The granularity-aware contrastive learning, enable MGTC to extract fine-grained temporal structures and coarse-grained class relationships. The experimental results demonstrate the excellent performance of MGTC in TSC tasks.}, archive = {J_TASE}, author = {Liang Xi and Wei Wang and Han Liu and Jingqi Pan}, doi = {10.1109/TASE.2025.3574034}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16026--16037}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MGTC: Multi-granularity temporal aware time series classification}, volume = {22}, year = {2025}, }
• (2025). A decentralized adaptive architecture for resilient current sharing in DC microgrids under current sensor integrity attacks. TASE, 22, 16014--16025. (www)

  @article{ , abstract = {Effective current sharing among distributed generation (DG) units is crucial for microgrid stability, efficiency, and reliability. However, current sensors, essential for monitoring and control, might be subject to cyberattacks. This paper develops a novel resilience-enhancing framework based on a decentralized adaptive architecture for current sharing in DC microgrids under integrity cyberattacks in current sensor measurements. The proposed framework employs two key components for each DG unit; a resilient unknown input observer (UIO) and a dynamic virtual current sensor. The proposed resilient UIO and virtual sensors estimate the correct current from inaccurate current sensor measurements and baseline control inputs in the presence of uncertain and unknown load conditions and power lines. Our proposed approach consists of a mathematically rigorous input-output stability analysis. It is shown that the interconnection of the proposed framework with microgrids’ baseline controllers enables DC microgrids to adapt to and compensate for sensor uncertainties without compromising overall performance. Specifically, the augmented framework with the baseline microgrid controller proposed guarantees asymptotic stability of the closed-loop dynamical system under time-invariant current sensor uncertainties and uniform ultimate boundedness under time-varying current sensor uncertainties. Several case studies on a DC microgrid illustrate the efficiency and performance of the proposed resilient current sharing scheme. Note to Practitioners—This work addresses a critical cybersecurity challenge in DC microgrids: attacks on current sensors that can disrupt proper load sharing among distributed generation units and affect system stability. The proposed solution introduces a resilience-enhancing framework that works alongside existing microgrid controllers, requiring minimal modification to the installed setup. For each distributed generator unit, the framework adds two key components: a resilient observer that detects sensor anomalies and a virtual sensor that provides reliable current measurements even when physical current sensors are compromised. The main features of the proposed framework are: (i) it functions without requiring communication among distributed generators unit (decentralized design), it adapts automatically to changing load conditions (load-independent design), (iii) it does not require detailed knowledge of line parameters (graph-independent design), (iv) it maintains stable operation even under current sensor attacks (resilience), and (v) it works with existing microgrid secondary control systems (retrofit control). The implementation of the proposed framework requires only local measurements and control inputs, making it cost-effective and straightforward to deploy in both new and existing DC microgrids. The framework’s ability to handle both constant and time-varying sensor attacks provides robust protection against various cybersecurity threats while ensuring resilient power sharing.}, archive = {J_TASE}, author = {Mahdieh S. Sadabadi and Subham Sahoo}, doi = {10.1109/TASE.2025.3573525}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16014--16025}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A decentralized adaptive architecture for resilient current sharing in DC microgrids under current sensor integrity attacks}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time event-triggered N-coalition nash equilibrium seeking for disturbed second-order players and its application. TASE, 22, 16001--16013. (www)

  @article{ , abstract = {In this article, Nash equilibrium seeking in the N-coalition noncooperative game is studied for the second-order disturbed players. In this formulation, players are divided into different coalitions. Each coalition acts as a virtual player in the noncooperative game, but the real decision-maker is the individual player. Players within the same coalition work together to minimize the cost function of the coalition they belong to, while each coalition competitively minimizes its own cost function. A prescribed-time event-triggered N-coalition Nash equilibrium seeking strategy is proposed based on the gradient descent method and the dynamic average consensus protocol. The proposed algorithm ensures that the players’ actions converge to the Nash equilibrium of the N-coalition game within a prescribed time, which can be assigned in advance, without any knowledge of the initial states and the system parameters. Additionally, information exchanges between players only happen when the designed event-triggering condition is met, thereby reducing the communication burden. The prescribed-time stability of the N-coalition Nash equilibrium is rigorously proven by Lyapunov stability analysis. The Zeno behavior is shown to be prevented until the Nash equilibrium is reached. Finally, the simulation experiments on maneuvering automated ground vehicles demonstrate the effectiveness of the proposed algorithm. Note to Practitioners—This article was motivated by the formation control of multiple ground vehicles but can also be applied to other cooperative and competitive tasks of second-order multi-agent systems. While some existing works designed control laws for specific tasks with limited applicability, this article provides a new game-theoretical method to accomplish a wide range of control tasks for multi-agent systems. By building a game-theoretic model and designing appropriate cost functions, the proposed N-coalition Nash equilibrium seeking strategy can serve as the control law to drive the multi-agent system to desired states. Notably, the proposed algorithm ensures the multi-agent system achieves its objective within a prescribed time under event-triggered communication, thereby reducing the communication load. This article outlines the conditions for convergence and how to apply the proposed algorithm. The simulation experiments verify the feasibility of this algorithm, but it has not yet been tested on a real system. Future research will involve testing on real-world applications and considering unmodeled nonlinearities and switching topologies.}, archive = {J_TASE}, author = {Mengwei Sun and Shandan Wang and Jian Liu and Changyin Sun}, doi = {10.1109/TASE.2025.3573127}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {16001--16013}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time event-triggered N-coalition nash equilibrium seeking for disturbed second-order players and its application}, volume = {22}, year = {2025}, }
• (2025). A safety-enhanced multi-modal objectives motion fusion method for autonomous retraction in robotic surgery. TASE, 22, 15990--16000. (www)

  @article{ , abstract = {In minimally invasive surgery (MIS), tissue retraction is critical for exposing the surgical site and facilitating pathological tissue excision. However, tissue retraction in MIS is subject to multiple constraints, including restricted field of view, non-damaging tissue retraction force, and limited instrument operating range, which hinder the safe and continuous retraction of tissue during robotic-assisted surgery. This paper introduces a novel autonomous retraction method for MIS, capable of safely retracting tissue within the aforementioned multiple constraints, assisting the surgeon in tissue excision. The method takes into account information from three different modalities: vision, force, and position. Based on these, three distinct control objectives are defined: retraction angle, retraction force, and safety space constraints, with an individual controller and model-predictive evaluation function designed for each objective. We propose a novel multi-objective motion fusion strategy designed to balance three distinct control objectives. This strategy evaluates the sensitivity of each objective to changes in the system state by comparing the gradients of their respective prediction evaluation functions and fuses the control inputs of the individual controllers. The proposed method allows for rapid addition or removal of objectives without altering the algorithmic framework. Experiments with phantoms and ex vivo animal tissue are conducted on robotic platform to validate the effectiveness of the proposed method in various configurations. Note to Practitioners—The motivation of this study is to address the challenge of autonomous tissue retraction in robotic-assisted laparoscopic surgery under multiple constraints. Existing methods primarily focus on single objectives and overlook the conflicts and coordination required among multi-modal objectives such as retraction angle, force, and safety space constraints. These limitations reduce the safety of robotic systems in complex surgical scenarios. A safety-enhanced multi-modal objective fusion method was proposed, integrating vision, force, and position information through dynamic weight allocation based on model-predictive mechanism. The method allows flexible addition or removal of control objectives and constraints, ensuring robust performance in varying conditions. This work provides a practical solution to improve the safety of autonomous robotic retraction, offering valuable insights for practitioners aiming to enhance robotic systems in constrained environments.}, archive = {J_TASE}, author = {Yudong Shi and Hangjie Mo and Xilin Xiao and Jin Fang and Ling Li and Kang Min and Xiaojian Li}, doi = {10.1109/TASE.2025.3573037}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15990--16000}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A safety-enhanced multi-modal objectives motion fusion method for autonomous retraction in robotic surgery}, volume = {22}, year = {2025}, }
• (2025). Information fusion for positioning system with location constraint of camera and UWB. TASE, 22, 15978--15989. (www)

  @article{ , abstract = {In this paper, a fusion estimator is investigated for positioning systems with location constraint of a camera and an ultra-wideband (UWB). Local camera and UWB estimators are designed independently for positioning with missed detections and local UWB estimates are aligned to local camera estimates for fusion estimation. Next the location constraint of the camera and UWB is transformed to a state constraint by state extension, and a constrained estimation projection method is applied to construct a fusion estimator. To measure the accuracy of the designed constrained fusion estimator, it is demonstrated that the estimation error covariance of the proposed method is equal to that of the matrix-weighted fusion method. Moreover, it is found that the proposed method has significant advantages over the matrix-weighted fusion method in computational cost, which is crucial for the real-time requirements of perception systems. Finally, the effectiveness of the proposed fusion estimator is displayed by simulation results and experimental results on the positioning system for an Ackerman vehicle. Note to Practitioners—The proposed method aims to solve the challenges of missed detections and different location coordinates from two sensors in the positioning systems. The paper introduces independently designed local estimators for both the camera and UWB sensors, which are capable of handling missed detections. The local UWB estimates are aligned with the local camera estimates to ensure that local estimates from two sensors are time-aligned in the fusion process. The location constraint of the sensors is transformed into a state constraint through state extension for constructing the fusion estimator with the constrained estimation projection method. The paper demonstrates that the estimation error covariance of the proposed fusion estimator is equivalent to that of the traditional matrix-weighted fusion method. Moreover, the proposed method offers significant advantages in terms of computational cost, which is critical for real-time performance in perception systems. Practitioners can use the proposed fusion estimator to enhance the accuracy and reliability of positioning systems that incorporate two different sensors with missed detections and location constraint. The method is particularly well-suited for applications where real-time performance is a requirement.}, archive = {J_TASE}, author = {Xiaohui Hao and Hongjiu Yang and Yang Xu}, doi = {10.1109/TASE.2025.3572916}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15978--15989}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Information fusion for positioning system with location constraint of camera and UWB}, volume = {22}, year = {2025}, }
• (2025). Design and control of a lever-bridge differential displacement reducer with sub-nanometer resolution. TASE, 22, 15966--15977. (www)

  @article{ , abstract = {Piezoelectric actuators and giant magnetostrictive actuators are widely used in micropositioning and micromanipulation devices. To achieve sub-nanometer resolution with these actuators, a compact flexure-based displacement reducer, which shows the capability of obtaining a very large reduction ratio so as to achieve motion resolution of sub-nanometer, is proposed. It incorporates both bridge-type and lever-type mechanisms, arranged such that the reducer’s output equals the differential displacement between the two mechanisms. Additionally, a kinetostatic model for the reducer is developed. The parameters of the reducer are optimized to minimize the variation of the reduction ratio. The optimization results are validated by those of a finite element model, proving the effectiveness and correctness of the proposed reducer and the kinetostatic model. A prototype is fabricated and valuated by open-loop control and closed-loop control. The average reduction ratio can reach 206.4. By adopting a neural network-based H $\infty $ robust controller, the reducer achieves satisfactory tracking performance. This comprehensive study affirms the reducer’s potential in enhancing the precision of micropositioning and micromanipulation devices. Note to Practitioners—This paper was motivated by the need for higher resolution in actuators to meet the growing demand for precise motion control in applications such as semiconductor manufacturing and biomedical devices. Currently, most compliant mechanisms focus on improving the motion range of actuators, often at the expense of motion resolution. The innovative design presented in this paper combines bridge-type and lever-type mechanisms to achieve a high reduction ratio, which can enhance the motion resolution of actuators, including piezoelectric and giant magnetostrictive systems. Furthermore, integrating a neural network-based H $\infty $ robust controller significantly improves tracking accuracy in dynamic environments. Practitioners in precision-driven fields can leverage this advanced displacement reducer to enhance system performance and explore new technological advancements, ultimately contributing to increased efficiency and capability in micro/nano manufacturing and related sectors.}, archive = {J_TASE}, author = {Houqi Wu and Yintian Zhang and Lingling Li and Ruiyu Bai and Haitao Wang and Guimin Chen}, doi = {10.1109/TASE.2025.3572151}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15966--15977}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and control of a lever-bridge differential displacement reducer with sub-nanometer resolution}, volume = {22}, year = {2025}, }
• (2025). A telerobotic shared control architecture for learning and generalizing skills in unstructured environments. TASE, 22, 15953--15965. (www)

  @article{ , abstract = {Direct teleoperation of robots in unstructured environments by non-experts often leads to low efficiency and increased risk. To this end, this paper proposes a shared control architecture where the robot can generalize demonstrations based on variable environments (start, obstacles, and goal positions) and infer user intention online to assist tasks safely and efficiently. First, the complex task is decomposed into unit actions, where the cubic-quintic-cubic Bezier curve is utilized to resolve the limitation of the classical dynamic movement primitives (DMP) algorithm in generalizing via-points. Gaussian process regression (GPR) was implemented to generalize expert demonstrations to different environments, ensuring full generalizable trajectory in the subtask space. GPR further extrapolate simple demonstrations from a subspace to the entire task space, avoiding the demand for numerous original demonstrations. Then, the online evolution of DMP is optimized: 1) an adaptive temporal scaling system is developed to synchronize evolution with human operations; 2) an intention prediction and expected evolution selection method based on operational input is proposed, achieving human-led operation guidance. Experiments validate the effectiveness of the generalization, obstacle avoidance, and operation assistance. Tests with non-experts reported the designed architecture enhances safety and efficiency, and reduces collisions. Note to Practitioners—This paper aims to improve the safety and efficiency of teleoperated robots in unstructured environments. To tackle complex missions with defined subtasks, this article adds auxiliary operation functions to the traditional direct control mode. The novel system can offline plan obstacle avoidance trajectories in unstructured environments and then guide human operations online. It can provide expected paths in advance, judge human intentions, and adapt to user execution speed. In practical applications, GPR can be trained from little number of demonstrations performed in a simple space. The guidance trajectories in complex task can be obtained by configuring multiple subspace units. With the proposed tactile feedback method, human operations can be effectively assisted.}, archive = {J_TASE}, author = {Xingmao Shao and Lun Xie}, doi = {10.1109/TASE.2025.3572103}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15953--15965}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A telerobotic shared control architecture for learning and generalizing skills in unstructured environments}, volume = {22}, year = {2025}, }
• (2025). PierGuard: A planning framework for underwater robotic inspection of coastal piers. TASE, 22, 15941--15952. (www)

  @article{ , abstract = {Using underwater robots instead of humans for the inspection of coastal piers can enhance efficiency while reducing risks. A key challenge in performing these tasks lies in achieving efficient and rapid path planning within complex environments. Sampling-based path planning methods, such as Rapidly-exploring Random Tree* (RRT*), have demonstrated notable performance in high-dimensional spaces. In recent years, researchers have begun designing various geometry-inspired heuristics and neural network-driven heuristics to further enhance the effectiveness of RRT*. However, the performance of these general path planning methods still requires improvement when applied to highly cluttered underwater environments. In this paper, we propose PierGuard, which combines the strengths of bidirectional search and neural network-driven heuristic regions. We design a specialized neural network to generate high-quality heuristic regions in cluttered maps, thereby improving the performance of the path planning. Through extensive simulation and real-world ocean field experiments, we demonstrate the effectiveness and efficiency of our proposed method compared with previous research. Our method achieves approximately 2.6 times the performance of the state-of-the-art geometric-based sampling method and nearly 4.9 times that of the state-of-the-art learning-based sampling method. Our results provide valuable insights for the automation of pier inspection and the enhancement of maritime safety (https://youtu.be/A54AJ4bbX98). Note to Practitioners—The research presented in this paper focuses on enhancing the efficiency and safety of underwater inspections of coastal piers by utilizing underwater robots instead of human divers. One of the primary challenges in this context is the development of efficient path planning strategies within complex, dynamic environments. Our study investigates the use of sampling-based path planning methods, specifically RRT*, to improve the efficiency of these inspections. The key takeaway for practitioners is that integrating bidirectional search with neural heuristic regions in the path planning of underwater robots can significantly enhance inspection efficiency, reducing both the time required and the complexity of navigating hazardous environments. The method is particularly useful in environments where the terrain is difficult to model or constantly changing. By adopting these algorithms, practitioners can improve operational efficiency, enhance safety protocols, and reduce the overall cost of inspections. We recommend that practitioners looking to implement this approach invest in suitable robotic platforms capable of handling the complexities of underwater navigation. Additionally, integrating these path planning methods with real-time data acquisition systems will further enhance the effectiveness of inspections, allowing for more accurate and faster assessments of coastal infrastructure.}, archive = {J_TASE}, author = {Pengyu Wang and Hin Wang Lin and Jialu Li and Jiankun Wang and Ling Shi and Max Q.-H. Meng}, doi = {10.1109/TASE.2025.3570694}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15941--15952}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PierGuard: A planning framework for underwater robotic inspection of coastal piers}, volume = {22}, year = {2025}, }
• (2025). Event-driven robust model predictive control for electromagnetic levitation systems with voltage fluctuation constraints. TASE, 22, 15930--15940. (www)

  @article{ , abstract = {In this article, the levitation control problem of electromagnetic levitation systems subject to track irregularities, air gap and vertical velocity and voltage fluctuation constraints are taken into account. An equivalent tracking model is established, which incorporates the flux density as a state variable to simplified the derivation of the robust constraints. With these robust constraints in place, a novel robust hierarchical control strategy is developed, which combines feedback linearization techniques with event-triggered robust model predictive control (ERMPC). Moreover, the theoretical properties are investigated, including recursive feasibility and closed-loop stability. Finally, to verify the superiority of the proposed control, simulations are conducted and compared with the conventional ERMPC, RMPC algorithms, and the proposed strategy under different parameter environments. The results demonstrate that the proposed algorithm can effectively reduce the frequency of controller updates and improve levitation performance while addressing various constraints, thereby enhancing system stability and safety. Note to Practitioners—The main purpose of this paper is to improve levitation performance while considering various constraints and reducing the frequency of the controller updates. Most existing levitation control strategies have overlooked the hazards posed by sharp voltage fluctuations, which may result in control failure, increased energy consumption, and even security risks. To address these control challenges in electromagnetic levitation systems with consideration of security and physical constraints, a novel robust hierarchical control strategy is proposed. This strategy combines feedback linearization techniques with event-triggered robust model predictive control (ERMPC). It aims to mitigate the impact of the track irregularities and achieve a satisfactory tradeoff between levitation performance and controller computing resources. The recursive feasibility and closed-loop stability of the proposed scheme are analyzed, and its superiorities are confirmed through the numerical simulations. However, further verification through experiments is necessary.}, archive = {J_TASE}, author = {Xiuming Yao and Yirui Han and Hongze Xu and Yu Yang}, doi = {10.1109/TASE.2025.3568118}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15930--15940}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-driven robust model predictive control for electromagnetic levitation systems with voltage fluctuation constraints}, volume = {22}, year = {2025}, }
• (2025). Trust-based sequential monotonic concession protocol for two-player task negotiation. TASE, 22, 15916--15929. (www)

  @article{ , abstract = {We propose a trust-based bilateral negotiation mechanism for task allocation problems under incomplete information, where a deceptive agent may disseminate false information to manipulate the outcome. Although numerous techniques have been explored to identify and reject such malicious agents, these approaches are often impractical due to insufficient prior knowledge, distributed implementation, or dependency on external entities. In contrast, the proposed mechanism enables agents to autonomously mitigate deception through trust evolution and distributed decision-making while acknowledging possible deception. The proposed mechanism integrates trust and cooperative utility into the Monotonic Concession Protocol, where the extended Zeuthen strategy forms a Nash equilibrium. A trust-weighted utility function is introduced, allowing agents to adaptively adjust their negotiation behavior based on the opponent’s behavior. We provide theoretical analysis of the mechanism, including Nash equilibrium, possible deception, and reachability. During negotiation, agents iteratively update trust, which governs both their degree of cooperation and information disclosure. A higher trust level encourages collaboration and gradual information sharing, while lower trust promotes selfish behavior and limited information exposure. Extensive Monte Carlo simulations demonstrate that the proposed mechanism effectively mitigates deception and enhances negotiation outcomes across various deception scenarios. Note to Practitioners—This research is motivated by the task allocation problem, specifically focusing on the cost functions that can be directly mapped into the traveling salesman problem. All simulation examples presented in this paper employ dynamic programming to compute exact cost functions and identify the optimal solution. However, given the combinatorial nature of the problem, this technique may prove impractical within feasible time constraints, especially when dealing with a large number of tasks or a complex environment with multiple obstacles. In such cases, a heuristic approach could serve as an alternative that alleviates computational burden.}, archive = {J_TASE}, author = {Donghae Kim and Maruthi R. Akella}, doi = {10.1109/TASE.2025.3574289}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15916--15929}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trust-based sequential monotonic concession protocol for two-player task negotiation}, volume = {22}, year = {2025}, }
• (2025). Receding horizon control for UAV formation reconfiguration: An enhanced marine predators strategy. TASE, 22, 15904--15915. (www)

  @article{ , abstract = {The uncrewed aerial vehicle (UAV) formation reconfiguration problem, due to its various limitations and high nonlinearity, has become a significant challenge in UAV formation. The existing methods mainly convert the problem into a constrained global optimization problem using control parameterization and time discretization (CPTD). This method, however, overlooks possible constraints on global optimization and the fact that solving the optimization problem itself can be difficult and time-consuming. In this paper, we firstly comprehensively considered the constraints in the UAV formation reconfiguration and proposed a problem model closer to real-world conditions. Then, we propose a new heuristic algorithm, the enhanced marine predators algorithm (EMPA) to address the constrained optimization problem, which has stronger optimization capability and greater stability when solving high-dimensional nonlinear problems. Finally, to decrease the computation cost, a more accurate and real-time receding horizon controller is designed based on the proposed EMPA, where the global optimal control problem of UAV formation reconfiguration is decomposed into a series of online local optimization problems. Simulation experiments verify the feasibility and effectiveness of the proposed control method. Note to Practitioners—This paper was motivated by the formation reconfiguration problem of uncrewed aerial vehicles (UAV). Existing approaches convert the problem into a constrained global optimization one using control parameterization and time discretization, which not only overlooks possible constraints on global optimization, but also has great difficulty in finding a feasible solution. This paper comprehensively takes into account such optimization constraints to build a problem model closer to real-world conditions, resulting in an enhanced marine predators algorithm of stronger optimization capability, greater stability as well as lower complexity. In this paper, a new heuristic algorithm called is developed, based on which a new receding horizon controller is established with its effectiveness validated through comparative experiments with other well-performing heuristic-algorithm-based controllers on the CEC-2017 benchmark. In future research, we will address the formation reconfiguration problem with leader-follower communication constraints, e.g. cyber-attacks, time-delays and topology failures.}, archive = {J_TASE}, author = {Zixiang Zhou and Xiaoxue Feng and Feng Pan and Hang Geng}, doi = {10.1109/TASE.2025.3573813}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15904--15915}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Receding horizon control for UAV formation reconfiguration: An enhanced marine predators strategy}, volume = {22}, year = {2025}, }
• (2025). SDS-SLAM: VSLAM fusing static and dynamic semantic information for driving scenarios. TASE, 22, 15889--15903. (www)

  @article{ , abstract = {Visual semantic SLAM integrates geometric measurements with semantic perception, making it widely applicable in autonomous driving and robotics. Semantic-assisted localization and dynamic object perception are two critical tasks in visual semantic SLAM. However, many existing state-of-the-art methods address only one of these tasks in isolation. To address issues of functional limitations and insufficient information utilization in a single framework, we propose a unified visual semantic SLAM framework, SDS-SLAM, which tightly couples static and dynamic semantic information to handle the motion estimation of both the camera and observed objects in driving scenarios. A multi-task network for driving perception is employed to extract semantic information, including drivable areas, lanes, and vehicles. Based on various information obtained, we propose semantic local ground manifolds (SLGMs) to represent the geometric structure and semantic features, enabling the online generation of a lightweight semantic map. Subsequently, we integrate SLGM-based constraints such as lane alignment and planar motion to promote camera and object pose estimation. We evaluated our method on the public KITTI dataset and self-collected real-world data. The results demonstrate that our method effectively perceives both dynamic and static semantic elements in driving scenarios, achieving high accuracy in estimating the poses of the camera and objects. Note to Practitioners—This work aims to utilize semantic information to assist ego-localization and enhance dynamic object motion perception in a unified visual SLAM framework, which mainly serves decision-making and planning for ground-based intelligent robots such as autonomous-driving vehicles. Existing methods typically deal with semantic-assisted localization or dynamic object tracking in isolation, thus limiting their ability to fully leverage the semantic richness of the environment and ignoring the mutual constraints between semantic information. By introducing semantic local ground manifolds, which couple road planes and lane lines for traffic scene representation, this work improves camera pose estimation accuracy through joint optimization with additional geometric constraints, including lane alignment, coplanarity, and planar support. Moreover, the planar motion constraints provided by SLGMs effectively reduce cumulative drift in object pose estimation, enabling more robust and accurate object tracking and motion perception over the long term.}, archive = {J_TASE}, author = {Yang Liu and Chi Guo and Jiao Zhan and Xiaoyu Wu}, doi = {10.1109/TASE.2025.3573739}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15889--15903}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SDS-SLAM: VSLAM fusing static and dynamic semantic information for driving scenarios}, volume = {22}, year = {2025}, }
• (2025). Observer-based fixed-time attitude tracking control of rigid spacecraft with output constraints. TASE, 22, 15878--15888. (www)

  @article{ , abstract = {This paper investigates the fixed-time attitude tracking control problem for rigid spacecraft subject to external disturbance and output constraints. First, the state transformed function (STF) technique is employed to convert the constrained spacecraft error dynamics into an unconstrained one. Subsequently, a fixed-time disturbance observer (FXTDO) is designed to estimate and reconstruct the lumped disturbance of unconstrained system. Combined with the developed STF, FXTDO and fixed-time integral terminal sliding mode (FITSM) surface techniques, the proposed fixed-time control law provides zero-error attitude tracking with high control precision and chattering avoidance, while the output constraints are never transgressed. The fixed-time stability of the closed-loop system is conducted via the Lyapunov technique and bi-limit homogeneity theory, and the expression of convergence time is also presented. Simulations illustrate the efficiency of the investigated controller. Note to Practitioners—There are ineluctable external disturbances and state constraints in spacecraft attitude systems, which degrades system performance and operation safety. Besides, for the practical engineering, the spacecraft is anticipated to achieve accurate attitude pointing and re-orientation maneuvers in a specified time interval, thus it is vital for spacecraft to improve convergence rate and tracking accuracy. Until now, these problems have not been uniformly addressed in the existing reports. To handle these issues, the paper concerns on developing an observer-based fixed-time control scheme to achieve accurate disturbance estimation and zero-error attitude tracking, while satisfying the constraints requirements. Such features make this controller extremely suitable for practical attitude tracking applications. Future research will devote to spacecraft cooperative control under state constraints and limited communication.}, archive = {J_TASE}, author = {Ganghui Shen and Bing Cui and Leonard Felicetti and Yuanqing Xia and Panfeng Huang}, doi = {10.1109/TASE.2025.3572693}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15878--15888}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based fixed-time attitude tracking control of rigid spacecraft with output constraints}, volume = {22}, year = {2025}, }
• (2025). BPSO-AHDL-IDS: Binary particle swarm optimization-based automated hybrid deep learning model for intrusion detection of internet of things. TASE, 22, 15859--15877. (www)

  @article{ , abstract = {The pervasive adoption of Internet-of-Things (IoT) systems has exposed critical vulnerabilities in cyber-security frameworks due to their decentralized deployment in unattended environments. While deep learning-based intrusion detection systems (IDSs) offer promising solutions, the design of hyper-parameters and neural architectures in existing models imposes prohibitive computational costs and expert dependency. To address these limitations, this work proposes an innovative automated hybrid deep learning method for IDS by employing a binary particle swarm optimization (BPSO) algorithm called BPSO-AHDL-IDS to effectively address the intrusion detection tasks of IoT. In BPSO-AHDL-IDS, the combination of convolutional neural network and recurrent neural network is considered as the hybrid deep learning model to extract features of the IoT dataset for accurate detection of intrusions. First, an efficient binary encoding mechanism is developed to describe the hyper-parameters and neural architectures of the hybrid deep learning model. Then, an efficient BPSO-based evolutionary operation is introduced to evolve the hyper-parameters and neural architectures to discover the optimized hybrid deep learning model. The performance of the proposed BPSO-AHDL-IDS method is testified by employing four datasets gathered from different IoT scenarios. It achieves accuracy of 0.9832, 0.9959, and 0.9897, precision of 0.9700, 0.9788, and 0.9843, recall of 0.9724, 0.9822, and 0.965, and $F_{1}$ -score of 0.9712, 0.9804, and 0.9744 on Bot-IoT, ToN-IoT, and Gas Pipeline datasets, respectively. On SWaT dataset, it achieves a precision of 0.9962, a recall of 0.9969, and an $F_{1}$ -score of 0.9965, respectively. The experimental results show the superiority of the proposed BPSO-AHDL-IDS method to machine learning methods, state-of-the-art manually designed and automated deep learning-based IDSs in terms of accuracy, precision, recall, and $F_{1}$ -score. Note to Practitioners—The open deployment of Internet-of-Things (IoT) devices in unattended environments makes them highly vulnerable to malicious cyber-attacks. This paper proposes a novel approach to enhance the security of IoT systems through the use of automated deep learning-based intrusion detection systems (IDSs). Traditional deep learning models used in IoT security require manual setting of hyper-parameters and neural architectures, which is a time-consuming and tedious trial-and-error process that relies on experts. To overcome this limitation, we propose an automated hybrid deep learning-based IDS, called BPSO-AHDL-IDS. The BPSO-AHDL-IDS method combines the strengths of a convolutional neural network and a recurrent neural network to extract features from IoT datasets to enhance the performance of intrusion detection. An efficient binary encoding mechanism is developed to describe the hyper-parameters and neural architectures of the hybrid deep learning model. Additionally, an evolutionary operation based on a binary particle swarm optimization (BPSO) algorithm is introduced to evolve these parameters and discover the optimized hybrid deep learning model. The performance of the proposed BPSO-AHDL-IDS method is evaluated using four datasets collected from various IoT scenarios. Overall, this paper presents a promising approach to improve IoT security through the integration of automated deep learning techniques. The proposed BPSO-AHDL-IDS method offers an automated and effective solution for addressing the challenges of intrusion detection in IoT systems. Practitioners in the IoT domain can significantly benefit from adopting this approach.}, archive = {J_TASE}, author = {Kang-Di Lu and Yao-Wei Yang and Guo-Qiang Zeng and Chen Peng and Guang-Gang Geng and Jian Weng}, doi = {10.1109/TASE.2025.3572510}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15859--15877}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {BPSO-AHDL-IDS: Binary particle swarm optimization-based automated hybrid deep learning model for intrusion detection of internet of things}, volume = {22}, year = {2025}, }
• (2025). HCNV: Hand-eye calibration based on surface normal optimization and view selection. TASE, 22, 15846--15858. (www)

  @article{ , abstract = {Accurate hand-eye calibration is crucial for robotics and computer vision tasks, such as robotic manipulation and 3D object recognition, as it directly affects operational precision and visual perception. However, existing methods rely solely on point clouds and neglect geometric features like plane normals, leading to suboptimal spatial alignment. Additionally, the lack of an intelligent view selection strategy worsens initial alignment errors, causing cumulative inaccuracies and degraded performance. We propose HCNV, a novel framework that combines optimized point cloud alignment with intelligent view selection to overcome these limitations. Our method uses plane normals and a hybrid genetic algorithm to refine the point cloud transformation matrix, significantly improving spatial alignment accuracy. Furthermore, the intelligent view selection strategy enhances point cloud matching and optimizes view coverage, increasing robustness across various conditions. Experimental results show that HCNV improves calibration accuracy by 20% and reduces computational time by 30% compared to state-of-the-art methods, demonstrating its effectiveness and practicality in real-world applications. Note to Practitioners—This paper addresses the robotic hand-eye calibration problem, essential for accurate sensor alignment. Existing methods face two main challenges: 1) the lack of geometric model information, leading to insufficient constraints during point cloud alignment and resulting in cumulative errors, and 2) the absence of a mechanism to filter out views with significant point cloud errors during calibration. To address these issues, we propose a novel hand-eye calibration framework, HCNV, which optimizes point cloud alignment by fitting point clouds to planes, improving relative pose estimation accuracy. Additionally, the framework incorporates an error-based view selection strategy to reduce the impact of reprojection errors during calibration. We demonstrate the effectiveness and robustness of the proposed method through extensive comparisons with state-of-the-art techniques on both synthetic and real-world datasets.}, archive = {J_TASE}, author = {Yu Liu and Hui Ma and Penglei Liu and Jun Cheng}, doi = {10.1109/TASE.2025.3568141}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15846--15858}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {HCNV: Hand-eye calibration based on surface normal optimization and view selection}, volume = {22}, year = {2025}, }
• (2025). Robust adaptive fixed-time consensus for nonlinear high-order multi-agent systems. TASE, 22, 15832--15845. (www)

  @article{ , abstract = {This paper investigates the consensus problem of leader-follower nonlinear high-order multi-agent systems through a fully distributed approach. A distributed consensus tracking control scheme is proposed, embedding a fixed-time convergent command filter with adaptive gain to ensure that the filter error ultimately reaches the origin. To tackle the challenges posed by unknown bounds of system nonlinearities and external disturbances, a robust control protocol is developed, incorporating an adaptive mechanism to implement compensatory action instead of directly estimating the lumped uncertainty. A notable feature of the proposed approach lies in its capability to determine the settling time bound without any global information, including the number of nodes, the eigenvalue of the Laplacian matrix, and other unknown parameters. The efficacy of the developed methodology is validated through numerical simulations and physical experiment conducted on a multi-motor system. Note to Practitioners—In contemporary engineering systems, the consensus tracking control of high-order multi-agent systems has emerged as a significant research topic. This paper proposes a fully distributed control scheme to achieve fixed-time consensus tracking. Unlike conventional finite-time and fixed-time approaches, the upper bound of the convergence time in this paper is independent of any global information and system’s initial states. Practical operational environments frequently introduce complex nonlinearities and disturbances into physical systems. To address this challenge, an adaptive mechanism is integrated into the robust control protocol. Importantly, the number of adaptive parameters remains fewer, irrespective of the number of uncertain parameters or the dimension of disturbances and nonlinearities. The feasibility of the proposed scheme is demonstrated via a physical experiment executed on a multi-motor system. In future research, we will simultaneously consider input saturation constraints.}, archive = {J_TASE}, author = {Jiale Li and Guofei Li and Zheng Guo and Zongyu Zuo and Xiaojing Zheng}, doi = {10.1109/TASE.2025.3574060}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15832--15845}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust adaptive fixed-time consensus for nonlinear high-order multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). 3D vision-guided robotic grinding framework for repairing random defects. TASE, 22, 15819--15831. (www)

  @article{ , abstract = {Robotic grinding is utilized for machining large complex components due to its exceptional flexibility and expansive workspace. However, the existing research primarily focuses on the global machining and lacks solutions for repair and remanufacturing of components with local defects. In order to perform the local repair efficiently, this paper proposes a 3D vision-guided robotic grinding framework for repairing random defects, by taking the removal of welding slags on the automotive body as an example. Based on the measured point cloud, a robust function weighted variance minimization (RFWVM) registration algorithm is utilized to position the automotive body with high precision. Meanwhile, the datum plane is reconstructed to extract the defect point clouds. On this basis, the welding slags are divided into regions in accordance with their height characteristics. For these regions, a robotic grinding depth prediction model is then constructed to determine the mapping between welding slag height and process parameters, and the genetic algorithm is further improved to solve the shortest path decision-making problem in robotic grinding. The experimental results conducted in a typical region of an automotive body confirm the practicality and effectiveness of the proposed framework. This study provides a valuable reference for local defects repair of complex components. Note to Practitioners—This work is motivated by local repair and remanufacturing in robotic machining of large complex components. Due to the random positions and sizes of local defects, it is highly challenging for robots to perform efficient and high-quality machining of these localized defects. The framework proposed in this paper is implemented based on high-precision 3D vision,which can realize the positioning of the workpiece and defect extraction simultaneously. The shortest machining path is determined using intelligent optimization algorithms, while the developed grinding depth prediction model assigns specific machining parameters to each defect area, ensuring both high-quality and efficiency machining. This framework can also be adapted for the removal of similar random defects, such as putty defects and paint defects on high-speed train bodies. However, it has certain limitations. For example, the approach is not suitable for machining small components using robots, does not offer advantages in conventional fixed-point global machining scenarios, and may face measurement challenges when machining workpieces that surface with numerous bends, which requires further improvement.}, archive = {J_TASE}, author = {Tao Ding and Peng Ouyang and Hongyou Zhang and Xiaozhi Feng and Lin Hua and Dahu Zhu}, doi = {10.1109/TASE.2025.3570709}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15819--15831}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {3D vision-guided robotic grinding framework for repairing random defects}, volume = {22}, year = {2025}, }
• (2025). Stochastic stability of the continuous-discrete maximum correntropy cubature kalman filter. TASE, 22, 15805--15818. (www)

  @article{ , abstract = {This paper investigates the stochastic stability of the continuous-discrete maximum correntropy cubature Kalman filter (CD-MCCKF), which is developed by embedding the maximum correntropy criterion as an optimization strategy into the continuous-discrete filtering architecture. By assuming that the relevant functions satisfy certain properties, we first establish the convergence of the employed numerical method and fixed-point iteration technique. Subsequently, combined with the findings on the error covariance boundary, it is proved that the estimation error of CD-MCCKF keeps mean-square bounded under specific conditions. To validate this stability analysis finding and evaluate the filtering performance, we conduct a classical case study. Simulation results confirm that CD-MCCKF not only maintains bounded estimation errors, but also surpasses the conventional continuous-discrete filters in terms of accuracy and convergence. Note to Practitioners—In the field of industrial automation, accurate state estimation is crucial for optimal control. Traditional Gaussian-based filters struggle with non-Gaussian noise prevalent in real-world applications such as aerospace tracking and smart manufacturing. This noise degrades estimation accuracy and system performance. Our work addresses this challenge by proposing the CD-MCCKF. By adaptively adjusting the kernel width, CD-MCCKF maintains high estimation accuracy without the need for trial-and-error tuning. In practical applications, this filter can significantly improve the reliability of the system in non-Gaussian environments by providing more accurate state estimation. Key insights include the proposal of an adaptive kernel width adjustment strategy and the theoretical proof of stochastic stability. However, the boundedness assumption and non-Gaussian noise in our work do not hold in all practical scenarios. Future research could explore the filter’s robustness in more complex environments and extend its applicability to high-dimensional systems, while also investigating its potential applications in various domains such as robotics, navigation systems, and intelligent manufacturing systems.}, archive = {J_TASE}, author = {Yanhui Wang and Hongbin Zhang and Dongmei Liu and Haibo Du}, doi = {10.1109/TASE.2025.3570422}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15805--15818}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stochastic stability of the continuous-discrete maximum correntropy cubature kalman filter}, volume = {22}, year = {2025}, }
• (2025). A novel deep reinforcement learning-based Path/Force cooperative regulation framework for dual-arm object transportation. TASE, 22, 15792--15804. (www)

  @article{ , abstract = {Dual-arm robots, with their high flexibility and broad operational range, are widely used in industrial and household transportation tasks. However, their high degrees of motion freedom and the closed-chain constraints formed during transportation pose challenges for path planning and force control. This study proposes a dual-agent control framework based on the Soft-Actor-Critic (SAC) algorithm of Deep Reinforcement Learning (DRL), where one agent is responsible for path planning and the other handles force control. This framework enables dual-arm robots to achieve dynamic obstacle avoidance, avoid unsolvable configurations and singularities, and generate efficient and smooth paths, while also fulfilling internal force tracking requirements based on task demands. Additionally, it addresses the gap in transferring the force control agent from simulation to real-world applications through a state mapping network, and the force control agent does not require retraining for different objects. Finally, the effectiveness of the proposed framework is validated through two scenes, including multiple bookshelves stacking and dynamic obstacle avoidance during the box transportation. Validation was also carried out with objects of different geometries and weights. Note to Practitioners—The motivation of this study is the path planning and force tracking control problem for dual-arm transportation. Existing methods rarely provide a comprehensive framework to address the task of dual-arm transporting different objects in dynamic environments. A DRL-based dual-agent control framework is proposed in this study, which simultaneously achieves task-compliant path planning and force tracking control. The trained agents exhibit strong generalization capabilities and stability, allowing them to handle objects of different geometries, weights, and stiffness without retraining. The limitation lies in the lack of methods for generating dual-arm initial grasping actions for complex-shaped objects. Future research will integrate image recognition techniques to enable grasping action generation for objects in diverse poses, further enhancing the applicability of the proposed framework.}, archive = {J_TASE}, author = {Yiyuan Hong and Huan Zhao and Xiangfei Li and Yanjia Chen and Guanxiao Xia and Han Ding}, doi = {10.1109/TASE.2025.3572722}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15792--15804}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel deep reinforcement learning-based Path/Force cooperative regulation framework for dual-arm object transportation}, volume = {22}, year = {2025}, }
• (2025). Bipartite synchronization of stochastic coupled systems with hybrid time-varying delays via asynchronous impulsive control. TASE, 22, 15778--15791. (www)

  @article{ , abstract = {This paper investigates the bipartite synchronization of stochastic coupled systems with hybrid time-varying delays and Markov jump via asynchronous impulsive control. Unlike existing studies, the asynchronous impulses considered in this paper are decentralized, which allows for the investigation of a wider range of dynamic properties. Additionally, to address asynchronous impulsive control, asynchronous average dwell-time condition is introduced for the first time, and new auxiliary functions are defined. Several sufficient criteria for attaining bipartite synchronization are derived through the application of the Lyapunov method and graph theory. In addition, the theoretical findings are applied to single-link robot arms with hybrid time-varying delays. To demonstrate the practicality of the findings, this paper includes two numerical examples. Note to Practitioners—This paper was motivated by existing results on asynchronous impulsive control for network systems. The asynchronous impulsive control is designed in this paper, which is decentralized and overcomes the drawbacks of existing asynchronous distributed impulsive control. In addition, asynchronous average dwell-time condition is established firstly, which broadens the application scope of existing average dwell-time condition. To achieve the bipartite synchronization of stochastic coupled systems, the graph theory and Lyapunov method as well as new auxiliary functions are utilized. Then, single-link robot arms systems are investigated as the application of theoretical results.}, archive = {J_TASE}, author = {Yan Liu and Yiliang Sun and Hui Zhou}, doi = {10.1109/TASE.2025.3572564}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15778--15791}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bipartite synchronization of stochastic coupled systems with hybrid time-varying delays via asynchronous impulsive control}, volume = {22}, year = {2025}, }
• (2025). Distributed secure balancing control for battery energy storage systems under privacy-preserving mechanisms. TASE, 22, 15768--15777. (www)

  @article{ , abstract = {This paper deals with the privacy-preserving-based distributed secure balancing control problem for battery energy storage systems (BESSs) in a microgrid. A novel distributed consensus algorithm is proposed based on the state decomposition strategy and the homomorphic cryptography technique, under which the state-of-charge (SOC) balancing issue is addressed amid potential privacy leak threats. By decomposing the state of the modified SOC for BESSs based on the number of neighboring BESSs, the privacy of each BESS’s state is safeguarded, provided that the BESS has at least one neutral neighboring BESS. Each homologous substate exchanges information directly, and non-homologous substates exchange information using a homomorphic cryptography algorithm. A weighted approach is used to ensure that the sum of the homologous substates aligns with the group decision value, thereby maintaining consistency among these homologous substates. The effectiveness of the proposed method is validated through a series of illustrative simulation experiments. Note to Practitioners—The integration of BESSs into microgrid poses challenges in maintaining SOC consistency, attributed to inherent parameter differences among BESS units. To meet these challenges, a distributed control algorithm is proposed, aiming to stabilize charging and discharging states across these systems. The use of public internet infrastructure for BESS communication raises privacy concerns, as it is vulnerable to security breaches. To mitigate the risk of eavesdropping and enhance data security, robust privacy protection mechanisms are integrated into the proposed methodology, ensuring the secure integration of BESS into microgrid systems. The results of this study provide a simple but useful algorithms achieve SOC consistency between BESSs, while performing privacy preservation.}, archive = {J_TASE}, author = {Zhuoyang Zhao and Engang Tian and Licheng Wang}, doi = {10.1109/TASE.2025.3571457}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15768--15777}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed secure balancing control for battery energy storage systems under privacy-preserving mechanisms}, volume = {22}, year = {2025}, }
• (2025). Contrastive conditional adversarial autoencoder with class-specific forces for imbalanced open-set fault detection. TASE, 22, 15757--15767. (www)

  @article{ , abstract = {In real-world industrial scenarios, fault detection faces the widely recognized challenge of data imbalance, which not only refers to the scarcity of fault data but also includes the imbalance in healthy data. This article is concerned with imbalanced open-set fault detection (IOSFD), a practical yet challenging scenario in industrial applications where multiple healthy operating conditions and multiple fault types are imbalanced. In this article, we propose a new contrastive conditional adversarial autoencoder for IOSFD. It constructs an end-to-end unified model based on multi-class known healthy and faulty data to address the reliance of traditional methods on fault samples, while optimizing with class-specific weighted forces to ensure equal attention to imbalanced known classes. Input and feature reconstruction conditioned on operating modes are utilized to learn a compact decision plane and achieve both unknown fault detection and known data classification. Significantly, we formulate the optimization objective of conditional reconstruction based on contrastive learning and introduce adversarial training to further enhance the model’s performance. The effectiveness of the proposed method is validated through real-world pipeline leak detection and Tennessee-Eastman multi-fault detection. Note to Practitioners—Existing data-driven fault detection methods often assume a balanced training dataset, where the quantities of data for various fault types and healthy conditions are roughly equal. However, in real-world industrial settings, fault data is frequently scarce, and the amount of data for different healthy operating conditions can vary significantly. This imbalance can lead to biased models that do not perform well in practice. To address this challenge and leverage the available data effectively, this paper develops a contrastive conditional adversarial autoencoder. The proposed method directly models multiple classes of known fault and healthy data to identify unknown faults, eliminating the need for unknown fault samples in model training. Additionally, the model employs class-specific weighted forces to ensure that all categories, regardless of their frequency in the training set, are given equal attention during learning. This helps to mitigate the effects of data imbalance, ensuring that the model learns from both the abundant healthy data and the sparse known fault data. The proposed approach has been validated through real-world applications, including pipeline leak detection and the Tennessee-Eastman multi-fault detection task. These case studies demonstrate the effectiveness of the method in practical settings. This technique offers a robust solution for industries where fault data is scarce and operating conditions are varied, making it a valuable tool for practitioners working with imbalanced datasets.}, archive = {J_TASE}, author = {Zhonglin Zuo and Tao Meng and Zheng Li and Hao Zhang and Tong Liu and Zhansheng Chen and Zhibo Pang}, doi = {10.1109/TASE.2025.3570077}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15757--15767}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Contrastive conditional adversarial autoencoder with class-specific forces for imbalanced open-set fault detection}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered model predictive control for nonlinear discrete cyber-physical systems with hybrid attacks. TASE, 22, 15745--15756. (www)

  @article{ , abstract = {This study presents a novel approach to dynamic event-triggered model predictive control (MPC) for nonlinear discrete-time cyber-physical systems (CPSs) subject to simultaneous hybrid denial of service (DoS) and deception attacks on the sensor-to-controller and controller-to-actuator communication channels. The proposed method introduces a dynamic event-triggered mechanism (DETM) with a predetermined offset to optimize communication efficiency in bandwidth-limited systems. Furthermore, a mechanism for detecting attacks has been developed to ensure the integrity of transmitted information, thereby mitigating the impact of hybrid attacks. Theoretical conditions are established to ensure the input-to-state practical stability of the closed-loop system using the proposed method. Finally, simulation cases are presented to validate the superiority of the proposed method in comparison to existing approaches in scenarios where CPSs encounter similar hybrid attacks. Note to Practitioners—This paper addresses the challenges of security control in nonlinear discrete information physical systems, such as robotic arms, mobile robots, and automotive systems. In CPSs, sensor and actuator components communicate over network channels, which can result in limited resources and vulnerabilities to tampering or delays due to network attacks, ultimately affecting system performance. Although MPC is effective, intermittent network attacks and inaccuracies in state estimation hinder the adjustment of control strategies. To tackle these issues, we propose a predictive control strategy that employs a dynamic event-triggering model with a detection mechanism. The method is validated through experiments conducted on a spring-damping system and a single-link flexible manipulator, demonstrating its potential for broader engineering applications. Future work will extend this strategy to multi-agent systems.}, archive = {J_TASE}, author = {Hongchao Song and Zhenlei Wang}, doi = {10.1109/TASE.2025.3572469}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15745--15756}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered model predictive control for nonlinear discrete cyber-physical systems with hybrid attacks}, volume = {22}, year = {2025}, }
• (2025). Multi-condition fault diagnosis of dynamic systems: A survey, insights, and prospects. TASE, 22, 15728--15744. (www)

  @article{ , abstract = {With the increasing complexity of industrial production systems, accurate fault diagnosis is essential to ensure safe and efficient system operation. However, due to changes in production demands, dynamic process adjustments, and complex external environmental disturbances, multiple operating conditions frequently arise during production. The multi-condition characteristics pose significant challenges to traditional fault diagnosis methods. In this context, multi-condition fault diagnosis has gradually become a key area of research, attracting extensive attention from both academia and industry. This paper aims to provide a systematic and comprehensive review of existing research in the field. Firstly, the mathematical definition of the problem is presented, followed by an overview of the current research status. Subsequently, the existing literature is reviewed and categorized from the perspectives of single-model and multi-model approaches. In addition, typical real-world application scenarios are then summarized and analyzed. Finally, the key challenges and prospects in the field are thoroughly discussed. Note to Practitioners—As industrial production systems grow more complex, traditional fault diagnosis methods face considerable challenges in addressing multi-condition issues arising from changing production demands, dynamic process adjustments, and disturbances from complex external environments. As a result, multi-condition fault diagnosis has become a key area of research and has garnered extensive attention in recent years. This paper presents a systematic review of multi-condition fault diagnosis for dynamic systems, ranging from developmental overview to key frameworks to typical application scenarios. To assist researchers in quickly grasping the current trends in the industry, this paper further summarizes the challenges and future prospects within the field.}, archive = {J_TASE}, author = {Pengyu Han and Zeyi Liu and Xiao He and Steven X. Ding and Donghua Zhou}, doi = {10.1109/TASE.2025.3571516}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15728--15744}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-condition fault diagnosis of dynamic systems: A survey, insights, and prospects}, volume = {22}, year = {2025}, }
• (2025). A novel rolling-window and production-maintenance-based machine learning approach to virtual metrology in semiconductor manufacturing. TASE, 22, 15713--15727. (www)

  @article{ , abstract = {Virtual metrology enables semiconductor manufacturers to continuously monitor and predict the quality and performance of semiconductor production during the manufacturing process. This kind of real-time monitoring tool is essentially important for maintaining process stability and ensuring consistent product quality. In this paper, a new rolling-window and production-maintenance-based approach is proposed on a weekly basis to establish the virtual metrology system. The proposed thickening and dewatering framework serves as the central core of the virtual metrology model, enabling real-time information and feedback to aid in making timely adjustments to the processing tool within machinery limitations. A meticulous approach to addressing each component within this innovative framework has been instrumental in achieving exceptional VM performance. The training dataset is routinely updated as the production maintenance is taking place to facilitate model retraining. A practical challenge to the domain engineer under certain circumstances where the process parameter considered in the virtual metrology model is realistically run out of specification is also rigorously investigated in this research. For the evaluation purposes, a real-world industrial case of the chemical vapor deposition process in semiconductor manufacturing is presented to illustrate the proposed approach. The proposed virtual metrology approach shows a substantial improvement in mean absolute error with 39.99%, 26.26%, and 30.75% for three different recipes, respectively, as compared to the base model currently implemented on the operation site. Note to Practitioners—Virtual Metrology relies on advanced data analytics and machine learning algorithms to establish predictive models for production measurements. This kind of models leverage historical raw trace data, sensor readings, and other relevant information to estimate critical quality characteristics, such as film thickness, wafer properties, and critical dimensions. To respond to dynamic changes of processing specification and production conditions, the dataset is collected online on a weekly basis to train the predictive model. Practically, an occurrence of production maintenance to trigger the model re-training and feature selection between rolling windows are particularly designed in the proposed approach as well.}, archive = {J_TASE}, author = {Shu-Kai S. Fan and An-Ting Zheng}, doi = {10.1109/TASE.2025.3570863}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15713--15727}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel rolling-window and production-maintenance-based machine learning approach to virtual metrology in semiconductor manufacturing}, volume = {22}, year = {2025}, }
• (2025). Unknown-input-proportional-differential observer-based event-triggered intrusion-tolerant control for human-in-the-loop multi-agent systems against unconstrained actuator and sensor FDIAs. TASE, 22, 15701--15712. (www)

  @article{ , abstract = {This paper investigates an intrusion-tolerant control problem for human-in-the-loop multi-agent systems (HILMASs) subjected to external disturbances and unconstrained actuator and sensor false data injection attacks (FDIAs) under directed graph. It is critical to emphasize that once a hacker gets into the control loop of the HILMAS, he can do whatever damage he wants, which means that the attack signal should be free of any constraints. Two main unconstrained attacks, i.e., unbounded FDIAs and variable-frequency FDIAs, are hard to accurately estimate and defend against since the widely adopted constraints on the existing FDIAs such as the bounded or/and the bounded first-order derivative have been removed. To tackle this challenging obstacle, a novel unknown-input-proportional-differential observer (UIPDO) is developed to not only reconstruct the follower agents’ states as well as unconstrained actuator and sensor FDIAs simultaneously, but also avoid the decrease of estimation accuracy caused by measurement deviation. It should be noted that this measurement deviation may be extremely large as it is caused by unconstrained sensor FDIAs, which renders the traditional observer ineffective in providing reliable and accurate estimates of the system states and unconstrained actuator and sensor FDIAs. Then, a novel UIPDO-based intrusion-tolerant control strategy without requiring boundedness of the first-order derivatives of the FDIAs as in existing literature is proposed. Furthermore, an adaptive Zeno-free event-triggered mechanism (ETM) solely relying local state information is developed to reduce the communication burden. Finally, the numerical simulation is provided to verify the merits and effectiveness of the developed methodology. Note to Practitioners—In HILMASs, security is of paramount importance. However, these systems face increasing vulnerability to external disturbances and unconstrained FDIAs under directed graph. Unconstrained attacks, such as unbounded and variable-frequency FDIAs, pose significant challenges due to their lack of traditional signal constraints, complicating their estimation and defense. This paper introduces a novel unknown-input-proportional-differential observer (UIPDO) based on an augmented descriptor system to reconstruct agent states and estimate these attacks, maintaining accuracy even with large measurement deviations. Furthermore, we propose an UIPDO-based intrusion-tolerant control strategy that does not assume boundedness of FDIA first-order derivatives, allowing tolerance of rapid and significant attack changes. Additionally, an adaptive Zeno-free ETM using local real-time state information optimizes communication resources. Our research provides a robust solution to enhance the security, resilience, and practicality of HILMASs against unconstrained FDIAs and external disturbances.}, archive = {J_TASE}, author = {Bo-Qun Wang and Xiang-Gui Guo and Jian-Liang Wang and Daniel Coutinho and Ju H. Park}, doi = {10.1109/TASE.2025.3570756}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15701--15712}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Unknown-input-proportional-differential observer-based event-triggered intrusion-tolerant control for human-in-the-loop multi-agent systems against unconstrained actuator and sensor FDIAs}, volume = {22}, year = {2025}, }
• (2025). Transfer multi-agent deep meta reinforcement learning method for load frequency control of performance market-based multi-area microgrid with prosumers. TASE, 22, 15687--15700. (www)

  @article{ , abstract = {In a performance-based frequency regulation market-based multi-area islanded microgrid, there are frequent load disturbances and tie-line power fluctuations caused by prosumers, which intensified the conflict of interests among the regulation service providers (units) in different areas. In order to address this issue, a fog computing-based cooperative load frequency control (FCC-LFC) method is proposed in this paper. This method draws on the idea of fog computing, with the units treated as independent agents to output their commands. In online application, each unit only needs to collect the state in its own area for decision-making, with no more communication required. In addition, to support the method, a transferred multi-agent deep meta policy gradient (TMA-DMPG) algorithm is also proposed in this paper that introduces transfer learning and meta-reinforcement learning techniques into a multi-agent actor critic learning framework. The meta-learner first acquires meta-knowledge through high-value demonstration trajectories, while the base learner applies transfer learning techniques to achieve fast adaptation to tasks and improve the robustness of FCC-LFC strategy. As revealed by the experiments conducted on the four-area LFC model in Sansha Island in the China Southern Grid (CSG), the proposed method is effective in reducing frequency error, generation cost and regulation mileage charge. Note to Practitioners—This paper introduces a FCC-LFC method designed for isolated multi-area microgrids. The FCC-LFC method enables each regulation service provider to operate as an independent agent, reducing communication requirements and improving response times. Using transfer learning and meta-reinforcement learning techniques, the proposed solution enhances frequency control efficiency, lowers generation costs, and increases system reliability. Practical applications include improved frequency regulation in microgrids and better integration of distributed energy resources, with the potential to extend to other decentralized energy systems.}, archive = {J_TASE}, author = {Jiawen Li and Jichao Dai}, doi = {10.1109/TASE.2025.3570620}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15687--15700}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transfer multi-agent deep meta reinforcement learning method for load frequency control of performance market-based multi-area microgrid with prosumers}, volume = {22}, year = {2025}, }
• (2025). A continual learning method for generalized grasping manipulation in a musculoskeletal robot. TASE, 22, 15671--15686. (www)

  @article{ , abstract = {Musculoskeletal robotic systems offer structural advantages while presenting significant control challenges. Current research on their manipulation capabilities, particularly multi-object grasping scenarios, remains insufficient. Furthermore, as robots operate in dynamic environments with evolving task requirements, developing their ability to grasp novel objects while maintaining existing manipulation skills becomes crucial. To address these challenges, we propose a novel continual learning method for generalized grasping manipulation in a musculoskeletal upper limb robot. First, we propose an end-to-end learning method for multi-object grasping that leverages object-specific latent features and multi-level states of a musculoskeletal robot. Additionally, a novel continual learning method for grasping tasks is proposed with a bio-inspired object-preference-based experience replay selector, which optimizes learning efficiency while mitigating catastrophic forgetting. Our approach successfully masters the grasping manipulation task of 10 objects in the first phase and continually learns 23 additional objects in the second phase, outperforming existing methods in both multi-object grasping and continual learning. Furthermore, our analyses of muscle synergy and methodological robustness demonstrate that the proposed approach generates biologically plausible muscle synergies and exhibits strong robustness against object observation bias and neural excitation noise. Our experiments, conducted both in simulation and on a hardware system, demonstrate the practical transferability of our method to an articulated dexterous hand grasping task. Note to Practitioners—Musculoskeletal robots, which incorporate human-like joints, muscles, and actuation mechanisms, exhibit exceptional robustness, dexterity, and compliance, offering promising applications and significant control challenges. While current research predominantly addresses lower limb locomotion and upper limb movements, studies on multi-object grasping and continual learning remain limited. This article presents a novel continual learning method for generalized grasping manipulation in a musculoskeletal upper limb, enabling end-to-end multi-object grasping without the dependency on predefined or generated grasping postures. The proposed methodology encompasses two key components: 1) the extraction and processing of object point cloud features as partial observations for the controller and 2) the implementation of a bio-inspired object-preference-based experience replay selector that balances new object learning and existing skill retention. Experimental evaluations reveal that the proposed approach exhibits effective muscle synergy patterns and demonstrates robust performance under diverse noise conditions and magnitudes, indicating significant potential for real-world deployment and cross-domain adaptation. Nevertheless, future work should address current limitations, including the extension from fixed-point to arbitrary position grasping, and the advancement from single-step grasping to complex multi-stage manipulation and assembly tasks.}, archive = {J_TASE}, author = {Bo Jiang and Ci Song and Siyuan Liu and Shuai Gan and Jiahao Chen}, doi = {10.1109/TASE.2025.3569402}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15671--15686}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A continual learning method for generalized grasping manipulation in a musculoskeletal robot}, volume = {22}, year = {2025}, }
• (2025). Stabilization of boolean networks: A model-free constrained state-flipped control approach. TASE, 22, 15662--15670. (www)

  @article{ , abstract = {In this paper, we explore the stabilization of Boolean Networks (BNs) under state-flipped control, incorporating three key constraints to address the challenges of disease treatment response times and the practical limitations of therapeutic interventions. Our approach includes: (i) a feasible flipped nodes constraint to ensure a subset of nodes can be realistically manipulated; (ii) a transient period constraint to align treatment with time-sensitive windows of therapeutic opportunity; and (iii) a constraint on the number of flip transitions to minimize resource consumption and potential side effects. Furthermore, an evolution selection rule is set for the BNs. Then, a necessary and sufficient condition is given for the stabilization under the above constraints. We present a modified Q-learning (QL) algorithm to design a policy for stabilization under constraints in the model-free case. Finally, BN models of metastatic melanoma network and T-cell large granular lymphocyte (T-LGL) network are considered to show the applicability of our results. Note to Practitioners—In this paper, we investigate the stabilization problem under three constraints for Boolean networks with state-flipped control. In gene regulatory networks, drug intervention usually only controls the activation and silencing of some or specific genes. Although the effect of drug intervention is significant in the short term, long-term application may cause cells to develop resistance. To solve these challenges, we consider the feasible flipped node constraint, transient period constraint and the number of flip transitions constraint. Current research on Boolean networks often overlooks various real-world constraints when considering state-flipped control, and our study addresses this oversight. We employ a model-free reinforcement learning approach to find the optimal state-flipped control policy under the given constraints, including timing and duration of drug use. The goal is to change the cell state from unstable to stable, which has important implications for improving the therapeutic effect of gene regulation. The solution proposed in this paper not only considers the dynamic characteristics of drug intervention, but also fully considers the biological feasibility and clinical application safety, which provides a new perspective and method for the state control of gene regulatory network.}, archive = {J_TASE}, author = {Zejiao Liu and Amol Yerudkar and Yang Liu and Jinde Cao}, doi = {10.1109/TASE.2025.3569241}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15662--15670}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stabilization of boolean networks: A model-free constrained state-flipped control approach}, volume = {22}, year = {2025}, }
• (2025). Exponential tracking control with guaranteed performance for strict-feedback systems under deferred full-state asymmetric constraints. TASE, 22, 15650--15661. (www)

  @article{ , abstract = {This work provides an exponential tracking control solution for a class of nonlinear systems characterized by unmatched, non-vanishing uncertainties and deferred full-state constraints. The presence of both non-vanishing and unmatched uncertainties complicates the task of achieving exponential tracking rather than regulation, particularly in scenarios involving deferred full-state asymmetric constraints alongside steady-state and transient performance requirements. To address these challenges, several key techniques are employed. Firstly, a time-varying feedback gain technique is utilized to ensure exponential tracking of the strict-feedback system. Secondly, we develop an asymmetric constraint mapping function that integrates the system state, tracking error, a finite-time adjustment function (AF), and an exponential AF to tackle performance issues without violating the deferred full-state asymmetric constraints, even when the initial conditions are unknown. Thirdly, an important lemma is derived to guarantee the boundedness of virtual controller derivatives, even as the exponential AF approaches infinity. Additionally, all signals in the closed-loop system are ultimately uniformly bounded. The effectiveness of the proposed scheme is validated through two examples. Note to Practitioners—In most existing studies on constrained systems, the results primarily focus on ensuring ultimately uniformly bounded behavior or asymptotic stability, with exponential tracking being rarely achieved. This paper aims to address this gap by proposing a robust adaptive control solution for uncertain strict-feedback systems under deferred full-state asymmetric constraints, achieving state-of-the-art exponential tracking performance. Furthermore, the steady-state and transient performance can be tailored by selecting an appropriate exponential adjustment function. Notably, the proposed approach ensures that system states remain within the asymmetric constraint boundaries after a finite time T, even with unknown initial conditions. This scenario is particularly relevant to physical systems transitioning from an unconstrained to a constrained state. The feasibility of the proposed method is demonstrated through a simulation example and a physical experiment on the Turtlebot 3-Burger robot. Future work will explore extending this approach to multi-agent systems for tasks such as formation control and consensus achievement.}, archive = {J_TASE}, author = {Yunfei Dai and Yujuan Wang and Zhuwu Shao and Yongduan Song}, doi = {10.1109/TASE.2025.3568098}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15650--15661}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exponential tracking control with guaranteed performance for strict-feedback systems under deferred full-state asymmetric constraints}, volume = {22}, year = {2025}, }
• (2025). Learn-gen-plan: Bridging the gap between vision language models and real-world long-horizon dexterous manipulations. TASE, 22, 15638--15649. (www)

  @article{ , abstract = {Long-horizon dexterous tasks have been a long-standing problem in robotic manipulation. Previous studies have developed task and motion planning, imitation learning, and reinforcement learning methods for long-horizon manipulations. However, these methods are hard to achieve efficient planning for new tasks. Empowered with the Vision Language Model (VLM), recent studies significantly improve the generalization of robot systems. However, these works are only verified in simple pick-and-place tasks due to limited skills. To this end, we propose the Learn-Gen-Plan (LGP), which combines the VLM and learning-based primitives to endow robots with the ability to efficiently plan and complete various long-horizon dexterous tasks. LGP contains two key phases: skill generation and task planning. In skill generation, the Skill Generator is proposed to utilize the learned key primitives and hand-crafted trivial primitives to generate adaptive robot skills. In task planning, the Multimodal Planner generates the robot plan based on image observation, generated skills, and text prompts. We set up a series of dexterous tasks (e.g., cable routing, peg-in-hole assembly) in a real-world lighting circuit wiring scenario to evaluate LGP. The experimental results show that LGP efficiently generates robot plans with learned skills, controlling the robot to complete various multi-step cable wiring tasks. Note to Practitioners—The motivation of this paper is to tackle the challenge of multi-step robotic manipulation tasks, where robots must utilize various skills (e.g., grasping, inserting) to manipulate different objects (e.g., pegs, wires) multiple times to achieve the task goal. Existing approaches either generate robot plans inefficiently for new tasks or support only limited skills (e.g., pick-and-place), leading to poor performance in real-world scenarios. We present a novel approach LGP, which leverages a Vision Language Model (VLM, e.g., ChatGPT-4o) and data-driven robot skills to enable robots to plan and adaptively execute multi-step tasks efficiently. LGP has two key stages: skill generation and task planning. In skill generation, we develop a ChatGPT-based generator that utilizes data-driven key and hand-crafted trivial motions to create adaptive robot skills. In task planning, a ChatGPT-based Multimodal Planner generates the robot’s plan using image observations, generated skills, and human instructions. Real-world cable wiring experiments demonstrate the feasibility of this approach in robotic assembly. Our approach is particularly beneficial for reducing the deployment costs of robotic systems in small and medium-sized enterprises where production tasks frequently change. Future research will focus on developing on-device models to improve control frequency and productivity in more complex tasks.}, archive = {J_TASE}, author = {Peng Hao and Shaowei Cui and Junhang Wei and Tao Lu and Yinghao Cai and Shuo Wang}, doi = {10.1109/TASE.2025.3567609}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15638--15649}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learn-gen-plan: Bridging the gap between vision language models and real-world long-horizon dexterous manipulations}, volume = {22}, year = {2025}, }
• (2025). Adaptive bayesian optimization for high-precision motion systems. TASE, 22, 15627--15637. (www)

  @article{ , abstract = {Controller tuning and parameter optimization are crucial in system design to improve closed-loop system performance. Bayesian optimization has been established as an efficient model-free controller tuning and adaptation method. However, Bayesian optimization methods are computationally expensive and therefore difficult to use in real-time critical scenarios. In this work, we propose a real-time purely data-driven, model-free approach for adaptive control, by online tuning low-level controller parameters. We base our algorithm on GoOSE, an algorithm for safe and sample-efficient Bayesian optimization, for handling performance and stability criteria. We introduce multiple computational and algorithmic modifications for computational efficiency and parallelization of optimization steps. We further evaluate the algorithm’s performance on a real precision-motion system utilized in the semiconductor industry applications by modifying the payload and reference stepsize and comparing it to an interpolated constrained optimization-based baseline approach. Note to Practitioners—This work is motivated by developing a comprehensive control and optimization framework for high-precision motion systems. Precision motion is an integral application of advanced mechatronics and a cornerstone technology for high-value industrial processes such as semiconductor manufacturing. The proposed method framework relies on data-driven optimization methods that can be designed by prescribing desired system performance. By using a method based on Bayesian Optimization and safe exploration, our method optimizes desired parameters based on the prescribed system performance. A key benefit is the incorporation of input and output constraints, which are satisfied throughout the optimization procedure. Therefore, the method is suitable for use in practical systems where safety or operational constraints are of concern. Our method includes a variable to incorporate contextual information, which we name the task parameter. Using this variable, users can input external changes, such as changing step sizes for a motion system, or changing weight on top of the motion system. We provide two parallel implementation variants of our framework to make it suitable for run-time operation under context changes and applicable for continuous operation in industrial systems. We demonstrate the optimization framework on simulated examples and experiment on an industrial motion system to showcase its applicability in practice.}, archive = {J_TASE}, author = {Christopher König and Raamadaas Krishnadas and Efe C. Balta and Alisa Rupenyan}, doi = {10.1109/TASE.2025.3565776}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15627--15637}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive bayesian optimization for high-precision motion systems}, volume = {22}, year = {2025}, }
• (2025). Dynamic charging and path planning for UAV-powered rechargeable WSNs using multi-agent deep reinforcement learning. TASE, 22, 15610--15626. (www)

  @article{ , abstract = {Unmanned Aerial Vehicle (UAV)-powered 5G/6G networks integrated with rechargeable wireless sensor networks (RWSNs) offer promising solutions for extending system lifetime, collecting data, and providing computing services and power to sensor nodes (SNs). UAVs offer significant advantages, including exceptional mobility, cost-effective deployment, and the ability to be easily reprogrammed for a wide range of missions. However, the limited onboard power capacity of UAVs, coupled with the lack of dynamic and intelligent charging station (CS) management and inefficient path planning, can lead to SN failure in dynamic mobile environments. To address these challenges, we propose an energy-efficient laser-charged UAV (LCU)-enabled RWSN environment, wherein UAVs, powered by laser beams from ground-based stations, provide services, collect data, and transfer energy to SNs. We formulate a joint optimization problem involving power allocation, dynamic charging strategy (DCS), and path planning to minimize task completion time and sensor node death time. Given the NP-hard nature of the problem, we employ a stochastic game model based on a Markov decision process (MDP) for its solution. To solve this problem, we propose a deep reinforcement learning (DRL) based algorithm that enables real-time charging scheduling decisions while optimizing network performance. We introduce a multi-agent double deep Q-network (MA-DDQN) model to determine the optimal trajectories for all UAVs in large and complex environments. Simulation results demonstrate that the MA-DDQN approach outperforms state-of-the-art techniques, showing significant improvements in terms of average delay, energy consumption, and task completion time. Note to Practitioners—UAVs integrated with laser charging and deep reinforcement learning (DRL) enhance the performance and longevity of RWSNs in dynamic environments. Using energy-efficient LCUs for data collection and SN charging, this framework reduces task completion times, prevents SN failures, and optimizes energy use. The MA-DDQN algorithm, which optimizes UAV trajectories and charging schedules in real-time, offers a scalable and adaptive solution that outperforms traditional methods in energy consumption, data delivery efficiency, and system robustness. This approach benefits applications of smart agriculture by improving network reliability, extending sensor lifetimes, and enhancing system efficiency in challenging environments.}, archive = {J_TASE}, author = {Mesfin Leranso Betalo and Supeng Leng and Abegaz Mohammed Seid and Hayla Nahom Abishu and Aiman Erbad and Xiaoshan Bai}, doi = {10.1109/TASE.2025.3558945}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15610--15626}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic charging and path planning for UAV-powered rechargeable WSNs using multi-agent deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time consensus tracking for high-order MASs with privacy protection via fully actuated system approach. TASE, 22, 15599--15609. (www)

  @article{ , abstract = {This study addresses the adaptive prescribed-time consensus tracking control problem for high-order multi-agent nonlinear systems with an improved privacy protection mechanism. A distinctive feature of the control method lies in the utilization of the fully actuated system approach to study high-order multi-agent systems, enabling system control without the need to simplify the high-order systems into the first-order systems. For agents requiring state information protection, the flexible privacy protection treats the output mask function as the transformation function, thereby protecting the privacy of all signals in the systems and allowing users to define the protection time. In addition, a prescribed-time scale function suitable for high-order systems is proposed by incorporating a constant term to avoid the singularity issue. This prescribed-time tracking control strategy ensures that the synchronization error converges to the prescribed region within the prescribed time, and prescribed time aligns with user-defined time in the privacy protection mechanism. Finally, the superiority of the proposed control method is verified through a numerical simulation comparison with different privacy protection method. Note to Practitioners—This paper aims to develop prescribed-time control algorithm for high-order multi-agent systems with privacy protection via the fully actuated system approach. In the field of automation, the sensitivity of information exchange among agents is extremely high, and any carelessness may lead to the leakage of original data details, posing a serious threat to the overall performance and stability of the system. Therefore, the adoption of advanced privacy protection technologies is particularly crucial, as it not only effectively safeguards personal privacy rights but also maintains the security and stability of the system. The fully actuated system approach enables a more precise characterization of the cooperation relationships and dynamic behavioral characteristics among agents, laying a solid foundation for designing more efficient control strategies. Furthermore, the introduction of prescribed-time control technology allows users to achieve the required control accuracy within a prescribed time, thereby minimizing the unnecessary expenditure of industrial resources and boosting resource efficiency in automation.}, archive = {J_TASE}, author = {Yifan Wang and Wei Sun and Shun-Feng Su}, doi = {10.1109/TASE.2025.3570513}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15599--15609}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time consensus tracking for high-order MASs with privacy protection via fully actuated system approach}, volume = {22}, year = {2025}, }
• (2025). Adaptive power regulation control for floating wind turbines with guaranteed transient performance. TASE, 22, 15588--15598. (www)

  @article{ , abstract = {Floating offshore wind turbines (FOWTs) hold significant potential for developing wind energy in deep-sea areas. However, they are prone to additional motions, heavy workloads, and undesirable fluctuations under combined wind-wave loads. To tackle these issues, this paper presents a novel adaptive blade pitch control strategy with guaranteed transient performance for FOWTs operating in above-rated wind speed regions. This strategy effectively suppresses platform pitch motion to enhance system stability while maintaining the power output at its rated value. Specifically, this method initiates by defining a filtered regulation error to solve the non-affine nature of its model. An adaptive blade pitch controller is subsequently proposed, integrating an online learning approximator to cope with the unknown dynamics caused by unmeasurable external environmental inputs and model uncertainties. To further mitigate approximation errors, a nonlinear robust control law with adaptive gains is employed, enhancing the robustness and adaptive capabilities of the controller. Superior to the traditional adaptive controllers, a significant advantage of this strategy is its ability to quantify and ensure regulatory performance within predefined constraints during both transient and steady-state stages, thereby achieving high-performance control in power regulation. Finally, simulation studies are conducted using the OpenFAST software to show the capability of the presented blade pitch controller, which guarantees stable power generation with guaranteed transient performance. Note to Practitioners—This paper addresses the need for floating offshore wind turbines (FOWTs) to develop a control mechanism that simultaneously mitigates platform motion and regulates power production in the above-rated wind speed regions. In practice, neglecting the effects of floating platform motions on the controlled system can significantly degrade the performance of FOWTs, resulting in unexpected power fluctuations and increased mechanical loads. To this end, we propose an adaptive blade pitch control strategy that utilizes the function approximator technique and specifically incorporates platform pitch motion into the controller design. This approach effectively balances the dual demands of platform stabilization and power regulation, aligning the control strategy with practical requirements. Furthermore, the proposed controller ensures that power regulation errors remain within predefined limits, thereby enhancing both transient and steady-state performance and better matching the actual operational conditions of FOWTs.}, archive = {J_TASE}, author = {Yingjie Gong and Wenchao Meng and Hua Geng and Qinmin Yang}, doi = {10.1109/TASE.2025.3568515}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15588--15598}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive power regulation control for floating wind turbines with guaranteed transient performance}, volume = {22}, year = {2025}, }
• (2025). High-quality dataset-sharing and trade based on a performance-oriented directed graph neural network. TASE, 22, 15576--15587. (www)

  @article{ , abstract = {The advancement of Artificial Intelligence (AI) models heavily relies on large high-quality datasets. However, in advanced manufacturing, collecting such data is time-consuming and labor-intensive for a single enterprise. Hence, it is important to establish a context-aware and privacy-preserving data sharing system to share small-but-high-quality datasets between trusted stakeholders. Existing data sharing approaches have explored privacy-preserving data distillation methods and focused on valuating individual samples tied to a specific AI model, limiting their flexibility across data modalities, AI tasks, and dataset ownership. In this work, we propose a performance-oriented representation learning (PORL) framework in a Directed Graph Neural Network (DiGNN). PORL distills raw datasets into privacy-preserving proxy datasets for sharing and learns compact meta data representations for each stakeholder locally. The meta data will then be used in DiGNN to forecast the AI model performance and guide the sharing via graph-level supervised learning. The effectiveness of the PORL-DiGNN is validated by two case studies: data sharing in the semiconducting manufacturing network between similar processes to create similar quality defect models; and data sharing in the design and manufacturing network of Microbial Fuel Cell anodes between upstream (design) and downstream (Additive Manufacturing) stages to create distinct but related AI models. Note to Practitioners—To accelerate AI adoption in advanced manufacturing, there is an urgent need for data sharing among participants to prepare high-quality datasets and improve AI model performance. This work proposes a dataset-sharing framework that lays the foundation for future data exchange and trade. Current approaches lack the flexibility to support sharing across diverse context and may expose the value of the data prematurely. To address these challenges, we introduce a performance-oriented representation learning framework that generates data for sharing and valuation to secure the value and preserve the private information, and then utilizes graph-based supervised learning to guide the sharing decisions for data receivers. The framework’s effectiveness and generalizability are demonstrated through two real-world manufacturing dataset-sharing case studies. Industrial participants can use this framework to rank datasets from others based on their predicted utility for specific downstream AI tasks.}, archive = {J_TASE}, author = {Yingyan Zeng and Xiaona Zhou and Premith Kumar Chilukuri and Ismini Lourentzou and Ran Jin}, doi = {10.1109/TASE.2025.3561081}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15576--15587}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-quality dataset-sharing and trade based on a performance-oriented directed graph neural network}, volume = {22}, year = {2025}, }
• (2025). Curvature-based continuous steering of stiffness-dominant concentric tube robots. TASE, 22, 15565--15575. (www)

  @article{ , abstract = {Existing works on controlling a concentric tube robot (CTR) mostly focus on the trajectory of its tip position or pose. In order to safely send CTRs in a confined lumen space, we propose to continuously steer the CTRs so that its entire shape will always attempt to approximate target curves over time. We focus on stiffness-dominant CTRs. Considering the differential geometry of such CTR shapes, we propose to work on the curvature domain to reduce the computational cost in searching the configuration of the CTRs. With our formulation, we model the curvature control of the CTR to find the optimal translation of each tube and then search for the rotation of the tubes to fit the target shapes. We demonstrate our method using sets of different target paths. The computational time per frame, ranging between 0.1 to 0.3 seconds across all experiments, highlights the efficiency of our approach in aligning the complete shape of the CTR with specified paths. Notably, for time-varying trajectories that could be reproduced by the CTR with its maximum deployment length reaching 150 mm, the root mean square error and median error were 0.98mm and 0.46mm, respectively. Note to Practitioners—This paper presents a novel approach for controlling stiffness-dominant CTRs, such that their entire shape consistently attempts to approximate target curves over time. This is crucial for applications such as navigation in confined anatomical spaces. Considering the differential geometry of these CTR shapes, we propose operating in the curvature domain. We model the curvature control of the CTR to determine the optimal translation for each segment, followed by searching for the rotation of the segments to fit the target shape. Our method avoids costly and computationally intensive full morphology evaluations while also achieving superior control accuracy, making it feasible for practical use in robotic surgery. In fact, this solution could be implemented in robot-assisted minimally invasive surgical platforms. Future work could explore adaptive modeling to address more complex clinical conditions, such as torsional deformations or variable tissue stiffness, or incorporate shape sensing for closed-loop control.}, archive = {J_TASE}, author = {Luhao Xie and Lifeng Zhu and Xiaoliang Jin and Aiguo Song}, doi = {10.1109/TASE.2025.3570861}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15565--15575}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Curvature-based continuous steering of stiffness-dominant concentric tube robots}, volume = {22}, year = {2025}, }
• (2025). Dioxin emission detection based on dynamic pruning online ensemble binary tree algorithm. TASE, 22, 15549--15564. (www)

  @article{ , abstract = {As one of the by-products of the municipal solid waste incineration (MSWI) process, dioxin (DXN) is not only difficult to detect but also potential harm to humans and the environment. The article proposes a method for detecting DXN emissions. It addresses the challenge of poor generalization performance in detection models due to the dynamic nature of the MSWI process. Firstly, the method constructs a historical soft sensor model and a drift detection model based on historical samples. Secondly, it assesses online samples for drift detection. When drift is detected, it calculates a distance threshold to prune the ensemble model. Subsequently, it reconstructs new ensemble sub-models and integrates them with the historical model to form an preliminary online ensemble model. Finally, it conducts local pruning and reconstruction on each sub-model, refining the final online ensemble model based on weighted posterior information. The efficacy of this approach is validated using synthetic, benchmark, and real DXN datasets from an MSWI plant in Beijing. Note to Practitioners—As an effective method for solid waste treatment, the MSWI process efficiently converts chemical energy into thermal and electrical energy. However, it also produces DXN as a by-product, posing significant risks that lead to the “Not-In-My-Back-Yard” effect for MSWI facilities. Currently, the mechanisms behind DXN generation, adsorption and emission remain unclear, and challenges such as high detection costs and significant lag hinder real-time detection efforts. To reduce DXN emission concentrations, operators in actual MSWI plants often resort to injecting large quantities of activated carbon for DXN adsorption. However, while effective, this method holds great importance for optimizing MSWI process in terms of pollution reduction control. Our objective is to achieve online soft measurement of DXN emissions, considering the dynamic operational conditions (conceptual drift). This approach can be applied by practitioners in similar industrial processes to enhance control and monitoring capabilities.}, archive = {J_TASE}, author = {Chaofan Xu and Jian Tang and Heng Xia and Wen Yu and Junfei Qiao}, doi = {10.1109/TASE.2025.3563856}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15549--15564}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dioxin emission detection based on dynamic pruning online ensemble binary tree algorithm}, volume = {22}, year = {2025}, }
• (2025). Self-triggered scheme design for takagi-sugeno fuzzy model based on mismatch premise variable with time-varying delay. TASE, 22, 15536--15548. (www)

  @article{ , abstract = {Over the last 20 years, numerous experts in the manufacturing companies have become interested in controlling flexible joint robots. The control design process using a networked framework has played a significant role in the industrial sector. This research addresses the self-triggered control problem for Takagi-Sugeno (T-S) fuzzy systems based on time delay is established in this research. A new self-triggered mechanism determines when the plant state is transmitted. Instead of continuously tracking predefined triggering conditions, the self-triggered system determines the next triggering instant based on the most recently triggered state information. The self-triggered mechanism relying on T-S fuzzy controller is then created using the triggered state, allowing for flexible configuration of its fuzzy rules and membership operations (MOs) that are not dependent on the controlled fuzzy system. Furthermore, it precisely demonstrates the interaction between the triggered condition and the plant state under the constraint of mismatched premise variables. To ensure system stability and performance, delay-dependent requirements are derived using a novel Lyapunov-Krasovskii functional based on fuzzy concepts to develop the fuzzy control design. The presence of the required controller is then shown under LMI-based conditions. Lastly, a simulation of a flexible robotic arm example is provided along with an evaluation of other triggered methods. Note to Practitioners—Flexible joint robots are widely used in robotics, automation, medicine, and other areas where they are unable to do certain tasks. This project aims to develop innovative solutions to significant issues in industrial robotic manipulators by enhancing flexibility, improving control techniques, alleviating stability conditions, and encouraging human-robot collaboration. This research proposes a fuzzy control technique that can handle time-delays signal and mismatch premise variables to preserve system stability in networked control systems and achieve related control goals. Among the advantages of this method that facilitate engineering applications are simpler architectures, reduced tuning factors, and enhanced control behaviors. Finally, a simulation that employs a realistic approach is shown to support the theoretical claims.}, archive = {J_TASE}, author = {Muhammad Shamrooz Aslam and Hazrat Bilal and Athanasios V. Vasilakos}, doi = {10.1109/TASE.2025.3570089}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15536--15548}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-triggered scheme design for takagi-sugeno fuzzy model based on mismatch premise variable with time-varying delay}, volume = {22}, year = {2025}, }
• (2025). Learning-based asynchronous sliding mode control for switching systems with partly unknown probabilities. TASE, 22, 15526--15535. (www)

  @article{ , abstract = {This study focuses on the learning-based asynchronous sliding mode control for switching systems, operating under a general switching rule and partially unknown probability information. A novel switching rule is constructed, governed by a joint probability distribution dependent on the current mode and its duration time, thereby overcoming the limitations of traditional Markov/semi-Markov models in terms of the difficulty in obtaining transition probabilities and computational complexity. Departing from traditional geometric distribution assumption, the proposed method follows more general duration distribution. Acknowledging the challenge of obtaining complete probability information in practical scenarios, partially unknown probability information is considered. In addition, a learning-based asynchronous sliding mode control law is developed, aimed at guiding state signals onto preset sliding regions and effectively reducing chattering induced by mode switchings. Finally, the efficacy and superiority of the developed theories are verified through both numerical and practical examples. Note to Practitioners—With the rapid development of networked control technologies, designing switching rules for switching systems and suppressing chattering in sliding mode control remain key challenge. In practical engineering systems, such as power systems, robotic control, and networked systems, the duration of different modes is often influenced by external disturbances, equipment characteristics, and load variations, making it difficult to model simply as a geometric distribution. Meanwhile, traditional Markov/semi-Markov models rely on extensive statistical data, and their transition probabilities are challenging to obtain accurately in practice. To address these issues, a joint probability distribution function based on the current mode and its duration, is adopted to accurately reflect the duration characteristics of the system in different states while reducing computational burden. The duration distribution, which depends on the current mode, is characterized by a general distribution, making the model more applicable to complex engineering systems. Furthermore, in industrial applications, sliding mode control often suffers from chattering issues caused by system modes, asynchronous double switching, discontinuities in the sign function, and other factors. As considered in this paper, a novel learning-based asynchronous sliding mode control method, incorporating emission probability and iterative learning, is proposed to alleviate chattering in sliding mode control.}, archive = {J_TASE}, author = {Jun Cheng and Tianfeng Tang and Huaicheng Yan and Ju H. Park and Michael V. Basin}, doi = {10.1109/TASE.2025.3569748}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15526--15535}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based asynchronous sliding mode control for switching systems with partly unknown probabilities}, volume = {22}, year = {2025}, }
• (2025). Constrained predictive learning control for omnidirectional wheeled mobile robots via reducing iteration horizon. TASE, 22, 15513--15525. (www)

  @article{ , abstract = {This paper proposes a predictive learning controller for an omnidirectional wheeled mobile robot (OWMR) with the target of achieving high-precision tracking of the desired trajectory within repetitive tasks. A prediction mechanism, based on the concept of “learning from the future,” is presented to accelerate learning convergence. Moreover, a constrained mechanism is leveraged to guarantee learning safety by introducing constraints on the input difference. The stability of OWMR under the proposed predictive learning controller is realized by formulating an optimization problem that incorporates terminal costs and constraints, together with its performance across iterations being evaluated through an analysis method based on composite energy functions. Additionally, a reducing prediction horizon strategy is adopted to mitigate computational burden, which both enhances the suitability and ensures the initial feasibility of the predictive learning controller for real-world applications. Subsequently, our predictive learning controller is implemented on the OWMR, together with several practical discussions for its deployment. Comparative simulations and real-world experiments are conducted to validate the practicality and effectiveness, highlighting its advantages over existing approaches. Note to Practitioners—Precise trajectory tracking control is a challenging but essential task, particularly for mobile robots. This paper proposes a learning-based controller that enhances performance through practice in repetitive tasks. However, traditional iterative learning controller often requires numerous repetitive operations and does not guarantee dynamic safety during each trial. To address these issues, a novel constrained predictive learning controller is introduced, which not only learns from historical experiences but also predicts future actions while accounting for input constraints. Furthermore, a new prediction strategy is incorporated to reduce computational burden. The effectiveness of proposed predictive learning controller is demonstrated through real-world experiments on the omnidirectional mobile robot.}, archive = {J_TASE}, author = {Wenxian Wang and Deyuan Meng}, doi = {10.1109/TASE.2025.3569611}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15513--15525}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Constrained predictive learning control for omnidirectional wheeled mobile robots via reducing iteration horizon}, volume = {22}, year = {2025}, }
• (2025). Event-triggered intermittent control for IT2 T-S fuzzy interconnected system on time scales. TASE, 22, 15502--15512. (www)

  @article{ , abstract = {In this paper, event-triggered control synthesis of Interval Type-2 Takagi-Sugeno (IT2 T-S) fuzzy interconnected system is investigated via aperiodic intermittent method. Remarkably, time scale differential equation is introduced to construct a hybrid mode of discrete and continuous interconnections. An aperiodic intermittent control mindset is proposed inspired by the system energy attenuation characteristics. What’s more, a novel event-triggered mechanism embedding exponential decay function is developed to help reduce the sampling amount during the work cycle. By establishing time-scale type Lyapunov functions, sufficient criterion is obtained to guarantee stabilization of IT2 T-S fuzzy interconnected system with no Zeno behavior. Finally, some simulations on different time scales are given to verify the effectiveness of the proposed method. Note to Practitioners–Interconnected systems are widely utilized in practical scenarios, such as electrical power systems, intelligent transport systems and multi-agents. The interconnections between subsystems often exhibit a hybrid mode of continuous and discrete cases at arbitrary moments. Unlike existing approaches by differential and difference equations, a unified framework that can solve model and control synthesis in such hybrid models is urgently needed. This paper introduces time-scale theory to unify the continuous and discrete interconnected systems. Furthermore, an energy-based intermittent control strategy is proposed, which can design appropriate energy functions tailored to specific requirements of different types of systems. Considering the heterogeneous nature of interconnected subsystems, the working and resting intervals of each subsystem controller are repartitioned by decomposing the overall energy function as $V(t)=\sum _{i=1}^{N}V_{i}(t)$ . The optimized approach results in a significant reduction in communication resource consumption, making it well-suited for practical applications in interconnected systems. In particular, obtained results can be applied directly to communication scheduling of vehicle formations. The energy-based intermittent control strategy can allow the smart grid commercial vehicle fleet scheduling and control system to respond quickly to tasks. Thereby signal collisions and bandwidth congestion can be avoided, and more scheduling errors and delays can be reduced.}, archive = {J_TASE}, author = {Nannan Rong and Dongjie Li and Sanbo Ding and Lei Liu}, doi = {10.1109/TASE.2025.3567999}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15502--15512}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered intermittent control for IT2 T-S fuzzy interconnected system on time scales}, volume = {22}, year = {2025}, }
• (2025). Incorporating commonsense knowledge to enhance robot perception. TASE, 22, 15488--15501. (www)

  @article{ , abstract = {Robots have been significantly employed across various avenues in recent years. However, their application has been largely limited to controlled environments where variables are few and predictable. To address this challenge, we propose Robo-CSK-Organizer, a novel system that enhances robotic perception by integrating commonsense knowledge (CSK) for improved object organization, classification, and decision-making. By combining ConceptNet for semantic reasoning, DETIC for object identification, and BLIP for contextual analysis, Robo-CSK-Organizer achieves superior ambiguity resolution, task adaptation, and explainability compared to models without CSK. Testing in a real-world robotics setting demonstrates notable gains in transparency, user trust, and error handling, making this approach valuable for advancing AI transparency and the development of versatile robotic applications in automation and engineering. Future directions of this work are also comprehensively discussed. Note to Practitioners—This paper introduces the Robo-CSK-Organizer, a system designed to enhance robotic decision-making through the integration of commonsense knowledge (CSK). Developed with the goal of improving both the efficiency and transparency of robots in task execution, this work addresses the critical need for advanced object organization and classification capabilities in multipurpose robotics, especially in domestic household settings. By leveraging a classical commonsense knowledge base alongside cutting-edge object detection and context discernment technologies, Robo-CSK-Organizer demonstrates significant improvements in ambiguity resolution, placement consistency, and explainable AI (XAI), highlighted by its superior performance over a baseline model using ChatGPT for comparison. Practitioners engaged in design and implementation of robotic systems will find Robo-CSK-Organizer truly valuable for its adequate use of commonsense reasoning (inherent in humans) due to which it is adaptable to numerous real-world applications.}, archive = {J_TASE}, author = {Rafael Hidalgo and Aparna S. Varde and Jesse Parron and Weitian Wang}, doi = {10.1109/TASE.2025.3565191}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15488--15501}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Incorporating commonsense knowledge to enhance robot perception}, volume = {22}, year = {2025}, }
• (2025). Flue gas oxygen content control based on interval type-2 fuzzy broad learning system and PID for MSWI process. TASE, 22, 15473--15487. (www)

  @article{ , abstract = {The flue gas oxygen content is critical for ensuring the efficient and stable operation of the municipal solid waste incineration (MSWI) process. This article proposes a novel control method based on an interval type-2 fuzzy broad learning system and proportional-integral-derivative (IT2FBLS-PID). First, a parameter self-learning complex controller based on IT2FBLS is designed, alongside a simple adaptive PID controller. To reduce computational complexity and resource consumption while maintaining control accuracy, a double event-triggering (DET) mechanism is introduced. Additionally, to enhance the controller’s adaptability and robustness, structural self-organization adjustments are made to the IT2FBLS. Finally, a stability analysis of the parameter learning and structural self-organization algorithms of the complex controller is conducted. The superiority of the proposed control method is demonstrated through actual data from a MSWI power plant in Beijing. Note to Practitioners—The MSWI process is characterized by strong nonlinearity and inherent uncertainty. In actual MSWI plants, operators typically rely on accumulated experience for manual control, which can be inconsistent and unstable. Traditional PID control and neural network control (NNC) methods often suffer from poor robustness, while fuzzy NNC struggles to maintain operational stability and converge quickly to the setpoint value during the initial control phase. In contrast, the method proposed leverages the wide applicability of PID and the fast response and uncertainty-handling capabilities of the IT2FBLS controller. It introduces a dual event-driven mechanism for both algorithms, offering a solution that practitioners can apply to similar complex industrial processes.}, archive = {J_TASE}, author = {Weiwei Yang and Jian Tang and Hao Tian and Wen Yu and Junfei Qiao}, doi = {10.1109/TASE.2025.3563863}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15473--15487}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Flue gas oxygen content control based on interval type-2 fuzzy broad learning system and PID for MSWI process}, volume = {22}, year = {2025}, }
• (2025). Adaptive formation control with limited information for uncertain nonlinear multi-agent systems. TASE, 22, 15461--15472. (www)

  @article{ , abstract = {The paper studies the distributed adaptive formation control in the setting of limited information transmitted between agents. The limited information, due to nonzero-kernel communication weight matrices between agents, has benefits in information security and privacy preserving, although as incomplete information, it challenges the formation control. Particularly, the multi-agent systems allow heterogeneous nonlinearities and serious uncertainties, making the formation control more difficult to achieve. In the paper, a fully-distributed adaptive control scheme is presented under appropriate conditions on the communication network, by incorporating two dynamic compensators and an auxiliary dynamic variable. The two compensators are employed to separately counteract the serious uncertainties and unavailable global information of the network resulting from unmatched states and weights. The auxiliary dynamic variable is introduced in the coordinate transformation to ensure the boundedness of the control, which cannot be derived by the traditional coordinate transformation. Two numerical examples are provided to show the effectiveness of the developed approach. Note to Practitioners—In real-world applications such as smart grids and smart transportation, communication networks play an essential role in cooperative operation, by which information is transmitted between agents. However, information transmitted in the communication network might be vulnerable to malicious eavesdropping due to the openness of networks. This could result in the leakage of sensitive information of agents, potentially disrupting the cooperative behavior and even the overall system security. In this paper, we consider adaptive formation control with limited information, instead of complete one. The limited information stems from nonzero-kernel weight matrices. In this case, even if the transmitted limited information is obtained by eavesdroppers, they still cannot reconstruct the full information of agents owning to the irreversibility of matrix weights, thus achieving information security and privacy preserving. In addition, the proposed distributed adaptive control strategy also can effectively counteract heterogeneous nonlinearities and serious uncertainties, which extends its application to a wide range of physical situations, such as cooperative engagement of multiple unmanned aerial and ground vehicles.}, archive = {J_TASE}, author = {Xiaoyu Xu and Yungang Liu}, doi = {10.1109/TASE.2025.3569646}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15461--15472}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive formation control with limited information for uncertain nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Single-period optimal inventory control with substitution: An end-to-end framework. TASE, 22, 15449--15460. (www)

  @article{ , abstract = {The paper presents a framework for addressing the single-period inventory control problem with substitutable resources. It introduces an end-to-end approach to determine inventory levels of multiple resource grades to meet stochastic demand at reduced costs. This framework is applicable to various scenarios such as multi-grade resource order decisions, used product procurement, and cloud computing resource allocation. In contrast to traditional Predict-then-Optimize approaches, which suffer from decoupling between demand forecasting and inventory decision-making, the paper leverages neural networks to directly yield inventory control decisions. A re-engineered loss function is proposed to tackle unsupervised learning challenges, demonstrating improved solution quality and faster inference speed compared to conventional Predict-then-Optimize methods through numerical experiments on three datasets with different data distributions. The research enhances decision efficiency and quality in high-frequency decision-making scenarios. Future directions for model improvement are also discussed. Note to Practitioners—This research offers an end-to-end inventory control model using neural networks to manage stock levels for goods with uncertain demand. The model determines inventory decisions directly from observable data, skipping traditional forecasting steps. It is designed for real-time use and can improve efficiency in sectors like remanufacturing and cloud services. Consider integrating this model into your operations for faster, data-driven inventory management. It could help reduce costs and enhance responsiveness to fluctuating demands. Future work will explore extending this approach to more complex scenarios.}, archive = {J_TASE}, author = {Yuewei Zhang and Hailei Gong and Zhi-Hai Zhang}, doi = {10.1109/TASE.2025.3568642}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15449--15460}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Single-period optimal inventory control with substitution: An end-to-end framework}, volume = {22}, year = {2025}, }
• (2025). Time-varying formation-containment control for heterogeneous multi-agent systems with communication and output delays via observer-based feedback protocol. TASE, 22, 15435--15448. (www)

  @article{ , abstract = {In this article, the time-varying formation-containment control problem of heterogeneous multi-agent systems (MASs) with an observer-based state feedback protocol is studied, where both communication delays and output delays are taken into consideration. By introducing an augmented Lyapunov-Krasovskii functionals (LKFs) candidate and employing less conservative methods to estimate its derivative, an improved quasi-formation-containment criterion is derived. On this basis, some observers and controllers are designed to allow for a larger maximum upper bound (MUB) of time delays on the premise of guaranteeing the fulfillment of the time-varying formation of leaders and the bounded convergence of followers. The estimation of the MUB determines the allowable delay range for system stability under the designed observers and controllers. Moreover, a distributed algorithm is developed to calculate the MUB of time delays, the observer gains, and the controller gains. Finally, two simulation experiments are implemented to demonstrate the obtained results, which verify the superiority and practicability of the criterion and algorithm. Note to Practitioners—This paper aims to solve the problem of formation-containment for MASs. It has been extensively implemented in various practical engineering fields, such as the exploration of the ocean through autonomous underwater vehicles and the teamwork carried out by mobile robots. In real-world engineering scenarios, the interaction between agents often has the following issues: 1) The state of agents may not be directly measurable in certain environments. 2) The information communication among agents is usually affected by time delays. 3) The existing formation-containment controllers are only suitable for dealing with predetermined small time delays. Therefore, a distributed algorithm with great effectiveness is designed to handle these problems.}, archive = {J_TASE}, author = {Yi Yang and Yong He}, doi = {10.1109/TASE.2025.3567992}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15435--15448}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time-varying formation-containment control for heterogeneous multi-agent systems with communication and output delays via observer-based feedback protocol}, volume = {22}, year = {2025}, }
• (2025). TRDM: A two-stage real-time discrimination method for spiral weld defects under dynamic distorted imaging. TASE, 22, 15420--15434. (www)

  @article{ , abstract = {Non-destructive testing of weld based on radiographic images is crucial for quality control of spiral steel pipes. However, accurately discriminating defect images from high-resolution and low-contrast weld images remains a challenging task. Moreover, the real-time imaging of radiography in industrial settings introduces dynamic distorted interferences to weld images, which has received limited attention in existing research. To address these issues, we propose a two-stage real-time discrimination method (TRDM) consisting of two carefully designed networks: an extremely lightweight weld region segmentation network and a semantic reconstruction-based discrimination network. Our method aims to adaptively reduce interference from invalid regions in the weld images to better exploit defect features from a semantic perspective and reconstruct them as defect-free regions as much as possible. To overcome the scarcity of weld defect image data, we construct two publicly available industrial weld datasets, including an image dataset and a video dataset. TRDM achieves competitive performance with state-of-the-art techniques, demonstrating its effectiveness in discriminating weld images. Furthermore, we integrate the TRDM into developed detection software for defect assessment, validating its efficacy in real welding scenarios. The source code and dataset are publicly available at https://github.com/Weld-det/TRDM Note to Practitioners—The motivation of this work is to address the detection challenges posed by dynamic distorted imaging of spiral welds in real-time operating conditions. Currently, manual judgment by professionals remains the predominant approach in spiral welding nondestructive testing processes, but it suffers from various limitations. This study aims to assist professionals in making real-time decisions for weld defect identification, thereby reducing missed detections and false alarms. To the best of our knowledge, there has been limited research attention given to this area, which represents a significant challenge. Therefore, this paper constructs a two-stage real-time discrimination network for screening spiral weld defect images. TRDM only requires training with normal defect-free weld images, which is particularly applicable to spiral welding scenarios with extremely scarce defect images. The trained model demonstrates outstanding real-time detection performance on both static weld images and dynamic distorted weld images, showcasing significant potential for practical applications in spiral welding inspection systems.}, archive = {J_TASE}, author = {Wenqi Cui and Kechen Song and Yu Zhang and Xiujian Jia and Xiaozheng Liu and Yunhui Yan}, doi = {10.1109/TASE.2025.3570251}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15420--15434}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TRDM: A two-stage real-time discrimination method for spiral weld defects under dynamic distorted imaging}, volume = {22}, year = {2025}, }
• (2025). Singularity-free low-complexity fault-tolerant prescribed performance control for spacecraft attitude stabilization. TASE, 22, 15408--15419. (www)

  @article{ , abstract = {This paper is concerned with the problem of fault-tolerant prescribed performance attitude stabilization for spacecraft under model uncertainties and actuator failures. In most of the existing control designs for spacecraft described by the unit quaternion, the possible singularity issue of the virtual control coefficient matrix is neglected such that the controllability of the attitude subsystem cannot be warranted throughout. On the other hand, the related works depend on complex algorithms of approximation, estimation or diagnosis to deal with unknown system dynamics. In this paper, a singularity-free low-complexity fault-tolerant prescribed performance control (PPC) strategy is put forward. To exclude the singularity issue, an initialization principle of the performance envelope is devised. On this basis, a static PPC law is developed, without parameter identification, function approximation, disturbance estimation, failure detection, failure isolation, failure estimation. In place of the classical Lyapunov stability theory, a unified performance analysis framework based on proof by contradiction and the barrier Lyapunov function is constructed. It not only turns out attitude stabilization with the preassigned settling time and accuracy whenever the actuator failures happen, but also discloses the intrinsic robustness of the control system versus actuator failures and model uncertainties. A comparative study on a rigid spacecraft is performed to demonstrate the validity and advantage of our approach. Note to Practitioners—Attitude stabilization is a fundamental element for the spacecraft to accomplish various important missions. A mainstream method to model the spacecraft attitude is the unit quaternion. In this way, the control distribution matrix of the attitude subsystem becomes singular under certain attitude, which causes the loss of controllability. This problem is neglected by the majority of the existing attitude stabilization methods. On the other side, the harsh space environment with high-intensity radiation, extremely low temperature and dust and fine particles gives rise to actuator failures in the spacecraft control system. Whilst the existing fault-tolerant control methods are capable of accommodating the actuator failures ultimately, the control performance during the post-failure transient stage is inevitably degraded. This problem is solved by the fault-tolerant prescribed performance control method, whereas the control algorithm shows a high complexity due to the dependence of function approximation or parameter identification. Motivated by the above observation, this paper presents a singularity-free low-complexity fault-tolerant prescribed performance control approach for the spacecraft with unknown dynamics and actuator failure. It achieves fast accurate attitude stabilization even in the faulty case.}, archive = {J_TASE}, author = {Jin-Xi Zhang and Yun-Qi Liu and Tianyou Chai}, doi = {10.1109/TASE.2025.3569566}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15408--15419}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Singularity-free low-complexity fault-tolerant prescribed performance control for spacecraft attitude stabilization}, volume = {22}, year = {2025}, }
• (2025). Enhanced adaptive prescribed performance constrained control with safe dynamic envelope for input-saturated nonlinear system. TASE, 22, 15396--15407. (www)

  @article{ , abstract = {When prescribed performance control (PPC) is implemented for state-constrained control, compact error boundary can easily lead to input saturation, which will trigger singularities. Existing methods for adjusting boundary through input saturation are mostly based on experience, often resulting in improper adjustments that either violate state constraints or offer limited improvement in singularity issue. To address this issue, an enhanced adaptive prescribed performance constrained control (EAPPCC) strategy with a safe dynamic envelope (SDE) is proposed. Firstly, an enhanced adaptive performance function with the SDE is designed, which offers the following advantages: 1) the safety-adjustable range of performance boundary is determined; 2) a saturation detection system (SDS) is designed to adaptively adjust the performance boundary within the adjustable safety range. A balance among performance constraint, input saturation, and safety condition is achieved by the SDE. Secondly, unknown dynamics are approximated by fuzzy logic system (FLS) independently of the control loop but integrated within a fuzzy state observer (FSOB). Then, a set of disturbance observers (DOBs) is designed to estimate the approximation errors of the FLS and unknown disturbances. Furthermore, a saturation-tolerant constrained controller (STCC) which can satisfy performance constraint is designed. Finally, the closed-loop stability is demonstrated based on the Lyapunov method, and the effectiveness and advantages of the proposed strategy are verified through simulations and experiment verifications. Note to Practitioners—The singularity issue caused by tracking error hitting performance boundary is addressed by adjusting performance boundary to balance input saturation and performance constraint in the existing PPC methods considering input saturation. However, practical systems such as autonomous vehicles, aircrafts, spacecrafts, and robots often impose state constraints for safety. Experience is relied upon in the aforementioned methods for adjusting performance boundary, which can easily lead to violation of state constraint and cause safety risks, or only provide limited improvements in addressing singularity issue. To address this issue, the EAPPCC strategy with the SDE is proposed. The maximum safe adjustable range of the dynamic envelope is calculated in real-time by this strategy, and the performance boundary is adjusted within the safety range based on the SDS. Additionally, a hybrid observer framework combined with the FSOB and DOBs is designed to address unmodeled dynamics and disturbances. Finally, the STCC based on dynamic surface control (DSC) method is designed to ensure performance constraint can be met. This method maximizes the improvement in addressing singularity issue within the PPC while ensuring system safety, thereby satisfying critical safety requirements in practical applications.}, archive = {J_TASE}, author = {Yuanwen Wang and Baojia Zhang and Jiuxiang Dong}, doi = {10.1109/TASE.2025.3569539}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15396--15407}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhanced adaptive prescribed performance constrained control with safe dynamic envelope for input-saturated nonlinear system}, volume = {22}, year = {2025}, }
• (2025). Network-based rail running band anomaly recognition via recurrent attention graphs. TASE, 22, 15381--15395. (www)

  @article{ , abstract = {Anomaly detection for rail running bands, the pattern of wheel-rail contact area, is crucial to analyze composite rail irregularities. This paper presents an all-weather vision-based solution for running-band inspection. However, two major algorithmic challenges restrict inspection effectiveness: 1) the identification of high-quality features for diversified and imbalanced data subject to noise and outliers, and 2) inferring implicit anomaly co-occurrence patterns. We regard overall running-band anomaly detection as a multi-label classification problem and develop a novel deep multi-anomaly recognition network via recurrent attention graphs (RAGRN). Note to Practitioners—The proposed RAGRN consists of two fundamental components, each directly addressing the two major challenges of this paper: 1) Class-specific features are extracted for fine-grained discrimination via split-channel and gradient-guided class-specific attention mechanisms; 2) We develop a multi-anomaly classifier, which effectively captures long-distance correlation features via a recurrent attention graph with visual and statistical guidance for graph propagation, containing prior statistical and image-specific information. The experiments and statistical analyses demonstrate that RAGRN outperforms all related state-of-the-art frameworks and has the potential to be applied to practical inspection.}, archive = {J_TASE}, author = {Xuefeng Ni and Paul W. Fieguth and Ziji Ma and Bo Shi and Hongli Liu}, doi = {10.1109/TASE.2025.3569178}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15381--15395}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Network-based rail running band anomaly recognition via recurrent attention graphs}, volume = {22}, year = {2025}, }
• (2025). Multi-modal self-supervised learning algorithm-based product recommendation. TASE, 22, 15371--15380. (www)

  @article{ , abstract = {With the rise of e-commerce, personalized recommendation algorithms have received much attention in recent years. Meanwhile, multimodal recommendation algorithms have become the next competitive track of personalized recommendation. The current mainstream recommendation algorithms (e.g., NeuMF, NGCF) rely too much on empirical data and do not effectively utilize multimodal information to model user preferences, while there is the problem of sparsity of user preference data. Secondly, they are still limited to using the same weights to fuse the item features of different modalities, and have the problem of inaccurate recommendation. For this reason, this paper proposes Multi-Modal Self-Supervised Learning Algorithm Based Product Recommendation (MMSLRec). 1) Data augmentation by self-supervised adversarial perturbation generation learning is performed to supplement the missing interaction information of tailed products to solve the common data sparsity problem in product recommendation. 2) A multimodal representation learning module is introduced to capture deep features in image and text modalities using AlexNet and Bert, respectively, in order to avoid the recommendation accuracy degradation problem caused by insufficient data feature learning. 3) Learning the user’s perceived strength in different modalities, using the user co-occurrence matrix to fuse the features and make recommendations. Experiments on Amazon-Electronics, Amazon-Clothing, and Amazon-Baby, with different degrees of data sparsity verify the effectiveness of our MMSLRec in solving the data sparsity problem and the multimodal user preference modeling problem, and compared to the state-of-the-art baseline recommendation model on the three datasets, the Recall@20 improved by 9.26%, 3.11% and 2.92%, respectively. Note to Practitioners—The motivation for this article arises from the requirement of efficient recommendation algorithms for e-commerce platforms. Mainstream recommendation algorithms are mainly based on sparse user item history interaction data for recommendation, which receives the limitation of data sparsity problem to a larger extent. Secondly, it is still limited to using the same weights to fuse the features of different modalities in the multimodal recommendation process. In this paper, we propose MMSLRec, which solves the data sparsity problem to a certain extent by using adversarial generative networks and fuse the features from different users using the user co-occurrence matrix. This method achieves better results in different applications.}, archive = {J_TASE}, author = {Li Gao and Haiping Fan and Qingkui Chen and Heyu Yang}, doi = {10.1109/TASE.2025.3568286}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15371--15380}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-modal self-supervised learning algorithm-based product recommendation}, volume = {22}, year = {2025}, }
• (2025). Large pre-trained models and few-shot fine-tuning for virtual metrology: A framework for uncertainty-driven adaptive process control in semiconductor manufacturing. TASE, 22, 15351--15370. (www)

  @article{ , abstract = {High-precision wafer metrology poses significant cost and throughput challenges in modern semiconductor manufacturing, where frequent process changes and recipe variations demand highly adaptive and scalable solutions. In this paper, we present a Generative-FewShot-Active Virtual Metrology (GFA-VM) framework that unifies large-scale generative modeling, few-shot fine-tuning, and uncertainty-driven active sampling into a single, data-centric system. A foundational generative model, built on a hybrid architecture of Transformer networks and Variational Autoencoders (VAEs), learns diverse sensor characteristics in an offline stage without relying on extensive labeled data. During online inference, the model produces both wafer quality predictions and predictive uncertainties; samples exceeding a dynamic uncertainty threshold are selected for physical measurement and few-shot model recalibration. This selective sampling both reduces measurement costs and adapts rapidly to new process conditions (e.g., novel recipes or equipment upgrades), requiring only a handful of freshly labeled wafers. The paper further addresses the long-term stability of the system through a self-updating mechanism that adjusts the uncertainty threshold when distributional shifts occur. Empirical evaluations confirm that our GFA-VM approach achieves state-of-the-art accuracy while significantly reducing metrology overhead compared to conventional virtual metrology methods. Additionally, rigorous theoretical analyses—including proofs of convergence and label cost bounds—demonstrate the reliability of using a generative foundation plus meta-learning technique. By fostering on-demand adaptation within a closed-loop framework, GFA-VM offers a comprehensive, scalable strategy for next-generation semiconductor process control. Note to Practitioners—Today’s semiconductor fabrication processes involve numerous machine types and recipe variations, making frequent measurements both expensive and time-consuming. The approach described in this paper, Generative-FewShot-Active Virtual Metrology (GFA-VM), aims to help practitioners maintain consistent product quality while significantly reducing the number of physical wafer measurements. Traditional virtual metrology solutions often require extensive retraining when new recipes or equipment changes occur; our method mitigates these burdens by using a “foundation model” that learns from large volumes of historical (both labeled and unlabeled) sensor data and can be quickly adapted with only a few newly measured wafers. Practically, engineers can integrate GFA-VM into existing Manufacturing Execution Systems (MES) or Advanced Process Control (APC) platforms. The system continuously estimates each wafer’s quality and flags only those with high uncertainty for actual measurement. This targeted approach optimizes the use of metrology resources and speeds up decision-making. The key benefit is flexibility: each time a new recipe or tool is introduced, the model can be recalibrated using as few as one to five measurements. However, initial setup requires careful data gathering to train the generative model, and ongoing tuning depends on reliable sensor signals. In addition, practitioners should note that sudden, large-scale process changes still need more measurements for model stability. Looking forward, the same strategy can be applied to other manufacturing contexts—where high-dimensional sensor data and limited measurement capacities challenge real-time quality control.}, archive = {J_TASE}, author = {Chin-Yi Lin and Tzu-Liang Tseng and Solayman Hossain Emon and Tsung-Han Tsai}, doi = {10.1109/TASE.2025.3568010}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15351--15370}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Large pre-trained models and few-shot fine-tuning for virtual metrology: A framework for uncertainty-driven adaptive process control in semiconductor manufacturing}, volume = {22}, year = {2025}, }
• (2025). SegNet4D: Efficient instance-aware 4D semantic segmentation for LiDAR point cloud. TASE, 22, 15339--15350. (www)

  @article{ , abstract = {4D LiDAR semantic segmentation classifies the semantic category of each LiDAR point and detects whether it is dynamic, a critical ability for tasks like obstacle avoidance and autonomous navigation. Existing approaches often rely on computationally heavy 4D convolutions or recursive networks, which result in poor real-time performance. In this paper, we introduce SegNet4D, a novel real-time 4D semantic segmentation network, offering both efficiency and strong semantic understanding. SegNet4D addresses 4D segmentation as two tasks: single-scan semantic segmentation and moving object segmentation, each tackled by a separate network head. Both results are combined in a motion-semantic fusion module to achieve comprehensive 4D segmentation. Additionally, instance information is extracted from the current scan and exploited for instance-wise segmentation consistency. Extensive experiments on the SemanticKITTI and nuScenes datasets demonstrate that our method outperforms the state-of-the-art in both 4D semantic segmentation and moving object segmentation. Through detailed runtime analysis, our method shows greater efficiency, enabling real-time operation. Besides, its effectiveness and efficiency have also been validated on a real-world robotic platform. The implementation of our method has been released at https://github.com/nubot-nudt/SegNet4D Note to Practitioners—This paper focuses on enhancing the efficiency of 4D semantic segmentation for applications in autonomous driving or navigation. Existing LiDAR-based 4D semantic segmentation methods fall short of real-time processing capabilities, thereby severely limiting their practicality for autonomous vehicles and robotic systems. To tackle these issues, we design a high-efficiency 4D semantic segmentation network that not only performs real-time operations on real-world robotic systems but also delivers superior performance, validating its practical utility. Future work can further leverage the instance information we introduced to improve the network’s functionality by achieving panoptic segmentation or 4D panoptic segmentation. Furthermore, some studies may use the 4D semantic labels predicted by our approach to reinforce tasks associated with robotic autonomy.}, archive = {J_TASE}, author = {Neng Wang and Ruibin Guo and Chenghao Shi and Ziyue Wang and Hui Zhang and Huimin Lu and Zhiqiang Zheng and Xieyuanli Chen}, doi = {10.1109/TASE.2025.3568001}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15339--15350}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SegNet4D: Efficient instance-aware 4D semantic segmentation for LiDAR point cloud}, volume = {22}, year = {2025}, }
• (2025). Secure stabilization of networked systems without DoS attack detection via novel aperiodic sampled-data control method with switched gains and its applications. TASE, 22, 15326--15338. (www)

  @article{ , abstract = {This article investigates the secure stabilization of networked systems against denial-of-service (DoS) attacks, by developing a new aperiodic sampled-data transmission scheme without attack detection and correspondingly proposing a novel resilient sampled-data control method with switched gains. For this problem, the nonlinear controlled systems are approximated by Takagi-Sugeno (T-S) fuzzy systems. In the new transmission scheme, the sampling of the system state is aperiodic. Additionally, based on each moment when the sampled data is successfully transmitted, the scheme determines in real time whether the aperiodic sampled-data pattern is impacted by DoS attacks, thereby categorizing it into various scenarios. Throughout the transmission scheme, there is no need to detect the start and end of each attack. In line with this transmission scheme, the resilient sampled-data control signal with switched gains is designed, adopting various control gains to cope with the different scenarios. This also means that the control continues to act upon the system even when transmission is impacted by attacks. Meanwhile, a new secure control standard against DoS attacks is introduced, which only requires an assumption about the cumulative duration impacted by attacks. In light of the above aspects, and according to the newly constructed discontinuous piecewise functional, the improved stability and stabilization results are derived, which are more advantageous than those in the literature. Finally, the results are applied to the permanent magnet synchronous motor (PMSM) system and the Rösser’s system, demonstrating the effectiveness and advantages of our new methods and results. Note to Practitioners—For networked systems, sampled-data control is an essential control approach for conserving bandwidth resources, but it is also susceptible to DoS attacks. Furthermore, aperiodic sampled-data control is a more realistic form. However, its irregular sampling makes analysis difficult, especially when attack detection is not feasible, resulting in a lack of existing research on the topic. Moreover, when under the impact of DoS attacks, traditional sampled-data control methods fail to respond flexibly, often resorting to either maintaining previous signals or taking no action. To address these challenges, we suggest an innovative aperiodic sampled-data transmission scheme, which does not rely on detecting each DoS attack and has greater practical significance. In addition, we introduce a novel resilient sampled-data control method with switched gains, thereby responding flexibly to attacks and improving control performance. The above designs embody both theoretical innovation and feasibility, leading to advanced secure control results. Furthermore, the effectiveness of our novel designs and results is validated through their applications to examples.}, archive = {J_TASE}, author = {Zhaoliang Sheng and Shengyuan Xu}, doi = {10.1109/TASE.2025.3567998}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15326--15338}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure stabilization of networked systems without DoS attack detection via novel aperiodic sampled-data control method with switched gains and its applications}, volume = {22}, year = {2025}, }
• (2025). Automated parts segmentation of sperm via a contrastive learning-based part matching network. TASE, 22, 15314--15325. (www)

  @article{ , abstract = {Sperm morphology measurement is vital for diagnosing male infertility, which involves quantification of multiple subcellular parts for each sperm. Instance-aware part segmentation networks have been introduced to address this task by automatically identifying individual sperm and segmenting their subcellular parts. However, major limitations of state-of-the-art instance-aware part segmentation networks include: 1) they are time-consuming and computational expensive due to sequential processing and multi-stage frameworks; 2) they perform poorly for densely packed sperm that overlap or cross over one another. To overcome these challenges, this paper proposes 1) integrating instance identification and subcellular part segmentation within a single-stage framework to save inference time and memory usage; 2) dividing a sperm target into simpler components (head and tail) to improve prediction accuracy, followed by a contrastive learning-based matching method to pair the head and tail. Experimental results on our clinically collected human sperm dataset demonstrated that the proposed network not only outperformed state-of-the-art CP-Net (by 3.5% $\mathrm{AP}^{p}_{vol}$ ) but also achieved real-time inference (48.0 frames per second), effectively meeting the clinical requirements for automated parts segmentation of sperm. Note to Practitioners—This work aims to automate the tedious and highly subjective process of manual sperm’s parts measurement. State-of-the-art methods employ instance-aware part segmentation; however their speed and accuracy cannot satisfy clinical requirements. To address theses issues, this paper introduces two key innovations. 1) Inspired by single-stage instance segmentation network, we develop an instance-aware part prediction branch. The proposed branch directly predicts part-encoded coefficients from the feature map, which are subsequently combined with prototype masks to produce final part segmentation results. 2) Since the sperm head and tail have simpler shapes, they are detected separately to improve segmentation accuracy. A contrastive learning-based method is then designed to pair head and tail based on similarity of feature embeddings extracted from the proposed instance prediction branch. The proposed method significantly outperformed existing networks, particularly in handling densely packed sperm. The presented method has applicability to analyzing sperm and more broadly other cell types.}, archive = {J_TASE}, author = {Wenyuan Chen and Haocong Song and Guanqiao Shan and Changsheng Dai and Hang Liu and Aojun Jiang and Chen Sun and Changhai Ru and Clifford Librach and Zhuoran Zhang and Yu Sun}, doi = {10.1109/TASE.2025.3567946}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15314--15325}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated parts segmentation of sperm via a contrastive learning-based part matching network}, volume = {22}, year = {2025}, }
• (2025). Meta-learning-based safety-critical control in multi-obstacles environments. TASE, 22, 15299--15313. (www)

  @article{ , abstract = {Autonomous robots operating in diverse scenarios are expected to safely and efficiently adapt to new, unknown, and cluttered environments. In this paper, we introduce a real-time goal-seeking and exploration framework incorporating novel meta-signed distance functions (MetaSDFs) and meta-buffer robust control barrier functions (Meta-BRCBFs). To adapt to environmental changes in real time, we employ Bayesian meta-learning to construct MetaSDFs. Deep neural network weights are initially trained offline, followed by efficient online adaptation at the last Bayesian layer, allowing for online updates at linear time complexity. Each MetaSDF is individually trained for its corresponding obstacle class, enhancing online distance estimation accuracy. Subsequently, buffer zones are constructed around the MetaSDFs to establish corresponding Meta-BRCBFs. These Meta-BRCBFs are activated only when the robot enters these zones, substantially reducing the number of CBFs required. Outside these specified buffer zones, the robot remains in goal-seeking mode, focusing on task completion. After entering a buffer zone, it transitions to exploration mode, prioritizing safety and exploring safe pathways, effectively balancing task execution with environmental adaptability. We demonstrate that, under this framework, the system achieves both safety and asymptotic stabilization. Extensive simulations and experiments are conducted to demonstrate our framework’s effectiveness in both simulated scenarios and real-world environments. These tests confirm our framework’s real-time capabilities and safety assurances in dynamic settings where state-of-the-art methods fail. The video is available at: https://www.youtube.com/watch?v=C6eshldAMxA Note to Practitioners—This paper is motivated by enhancing safety for autonomous robots operating in diverse and changing environments. Existing methods, which utilize traditional or incremental neural networks, fail to perceive environmental changes and obtain the nearest distances to obstacles for safety in real time. Additionally, with the increased number of obstacles, the real-time performance of optimization solvers for such approaches is further reduced, which quickly causes error motions. To overcome these challenges, this paper introduces a novel approach enabling online updates of the trained neural network so that it adapts to unknown environments rapidly. Buffer zones are established around obstacles, allowing us to focus solely on those that the robot enters into these zones. This eliminates the need to consider all obstacles simultaneously, significantly reducing the time required for the optimization solver. The safety and asymptotic stability have been rigorously proven through mathematical validations. Extensive simulations and physical experiments demonstrate the framework’s real-time and safety performance in complex and variable scenarios. Future research will extend this approach to robots with higher-dimensional control systems, such as robotic manipulators.}, archive = {J_TASE}, author = {Yu Zhang and Long Wen and Yuhong Huang and Siming Sun and Zhenshan Bing and Wei He and Alois Knoll}, doi = {10.1109/TASE.2025.3565524}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15299--15313}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Meta-learning-based safety-critical control in multi-obstacles environments}, volume = {22}, year = {2025}, }
• (2025). Hybrid model and data-driven optimal PID control with enhanced disturbance rejection. TASE, 22, 15286--15298. (www)

  @article{ , abstract = {This paper proposes an optimal proportional-integral-derivative (PID) control method with improved disturbance rejection for nonlinear uncertain systems. The controller achieves optimal closed-loop dynamics by integrating constant and time-varying control gains. Initially, the system is transformed into a linear model with a nonlinear uncertainty term that lumps unmodeled dynamics and disturbances. Constant PID gains are subsequently derived using an optimal tuning rule based on the frequency and gain of the linear model at $-180^{o}$ and $-120^{o}$ phase lag points for systems with and without ultimate point, respectively, by employing a fractional-order integrator and Tchebyshev polynomials to simplify computation. Then an extended state observer (ESO)-assisted recursive optimization approach is introduced to adaptively adjust the time-varying PID gains in real time based on the uncertainty estimation from the ESO and operational data, improving adaptation under nonlinear uncertainties. This hybrid tuning enhances the robustness of the closed-loop system to uncertain dynamics and mitigates the instability caused by poor initial parameters, outperforming traditional model-driven and data-driven methods. The stability of the proposed approach is proven, and its effectiveness is validated through simulations and an industrial case study on the pressure control of the hematite regrinding process. Note to Practitioners—This paper addresses practical challenges in controlling nonlinear uncertain systems and proposes an optimal proportional-integral-derivative (PID) control strategy by integrating the advantages of model- and data-driven methods. The hybrid tuning strategy demonstrates superior performance compared to traditional model-based tuning methods, which struggle with adapting to nonlinear uncertain dynamics, and data-driven methods, which are prone to instability due to inappropriate initial values. Particularly, the designed extended state observer (ESO)-assisted recursive optimization method can improve the disturbance rejection performance of the closed-loop system through the continual learning of PID parameters, which enhances the robustness of the closed-loop system. This paper provides practical guidance for applying the proposed method to pressure control in hematite regrinding processes. By following a similar framework, the approach can be extended to most systems in the process industry with a PID controller. First- or second-order linear models are recommended in practice when considering the complexity of implementation. For systems of greater complexity, additional auxiliary strategies may be required, such as decoupling techniques for multivariable systems and Smith predictors for systems with large time delays.}, archive = {J_TASE}, author = {Ning Liu and Yajun Zhang and Zhuqing Wang}, doi = {10.1109/TASE.2025.3562398}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15286--15298}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hybrid model and data-driven optimal PID control with enhanced disturbance rejection}, volume = {22}, year = {2025}, }
• (2025). Resilient time-varying formation optimal tracking for heterogeneous multi-agent systems with coupled constraints. TASE, 22, 15274--15285. (www)

  @article{ , abstract = {Formation constrained optimal tracking problems for heterogeneous multi-agent systems under Byzantine attacks are studied. The objective of the honest agents, unaffected by Byzantine attacks, is to find optimal trajectories that minimize the cumulative local cost functions of all honest agents, while simultaneously satisfying the cumulative local inequality constraints and maintaining the formation configuration. A distributed resilient formation tracking controller utilizing preview control and primal-dual strategy is proposed without requiring each agent to know which agents are affected by Byzantine attacks. The performance of the proposed algorithm is analyzed in terms of the upper bounds of dynamic regret and constraint violations. Numerical simulations and experiments, involving one uncrewed aerial vehicle and four uncrewed ground vehicles, are performed to verify the effectiveness of the obtained results. Note to Practitioners—This paper is motivated by the formation tracking problem of multi-agent systems, which has potential applications in collaborative reconnaissance, logistics transportation, search and rescue. Existing formation tracking methods typically address the issues under the premise that formation references are known in advance to one or multiple leaders. However, in practical applications, multi-agent systems may be required to converge to a predefined formation configuration while ensuring globally optimal performance even under cyber attacks, given that each agent possesses a local cost function and local constraint functions that may vary in time. Therefore, this paper proposes a distributed resilient formation constrained optimal tracking protocol for heterogeneous multi-agent systems to withstand Byzantine attacks, utilizing neighboring predictable information. The performance is mathematically analyzed, and practical experiments involving one uncrewed aerial vehicle and four uncrewed ground vehicles yield promising results.}, archive = {J_TASE}, author = {Lingfei Su and Zhi Lin and Yongzhao Hua and Zhexin Shi and Xiwang Dong and Jinhu Lü and Danwei Wang}, doi = {10.1109/TASE.2025.3568081}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15274--15285}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient time-varying formation optimal tracking for heterogeneous multi-agent systems with coupled constraints}, volume = {22}, year = {2025}, }
• (2025). Improving information collection environment of magnetocardiography device through adaptive interference suppression strategy. TASE, 22, 15264--15273. (www)

  @article{ , abstract = {Multiple frequencies high-amplitude magnetic interference in the environment can adversely affect the detection of cardiac magnetic field signals inside magnetocardiography device. Adaptive multi-resonance control (AMR) is suitable for suppressing such interference, but the stability of active magnetic field compensation system (AMC) will be destroyed by the influence of the time delay of the sensor. To address this problem, this paper proposes a interference suppression method that combines an improved adaptive multi-resonance controller (IAMR) with proportional differential (PD) and linear extended state observer (LESO). Experimental results show that PD-LESO-IAMR can effectively suppress multiple frequencies magnetic interference under the premise of ensuring system stability. Compared with conventional PD-LESO, the average magnetic field interference in the MSC can be reduced from 3.73 pT to 1.80 pT, and the maximum anti-interference ability was increased by 50.25%. Note to Practitioners—The motivation of this paper is to control the magnetic field generated by the coil to reduce the influence of environment magnetic field fluctuations on the measurement of weak magnetic field signals. Existing magnetic field compensation methods mainly focus on the design of coils, and rarely consider optimization and improvements in control methods. Therefore, the suppression effect of magnetic field interference at specific frequencies in the environment is insufficient. This paper proposes an adaptive interference suppression strategy that can effectively suppress high-amplitude magnetic interference without being affected by system time delays. By applying the proposed control strategy, the magnetic field fluctuations are kept within a smaller range, improving the accuracy of magnetic field measurement. Future work is to extend the proposed method to more application scenarios that require extremely weak magnetic field environments. The proposed control strategy is also valuable in solving the high-frequency unbalanced magnetic pull of the magnetic levitation rotor system, improving the high-frequency current suppression of the permanent magnet synchronous motor, and improving the measurement accuracy of the atomic magnetometer for periodic magnetic field signals.}, archive = {J_TASE}, author = {Pengtao Tian and Haifeng Zhang and Zhihui Hong and Shiqiang Zheng}, doi = {10.1109/TASE.2025.3567988}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15264--15273}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Improving information collection environment of magnetocardiography device through adaptive interference suppression strategy}, volume = {22}, year = {2025}, }
• (2025). Reinforced prescribed performance control for virtually-coupled trains with saltatory targets and switching constraints. TASE, 22, 15251--15263. (www)

  @article{ , abstract = {Virtual Coupling (VC) has emerged as a promising strategy for railway system control, offering significant enhancements to rail-line capacity and addressing uneven transport demand by dynamically minimizing the headway distance between trains when necessary. Consequently, the operating targets of VC rear trains, which can be influenced by operational environments and railway line conditions, must be promptly adjusted based on the spacing from the adjacent leading train. This adjustment ensures operational safety within the train movement authority. Different from traditional continuous prescribed performance control (PPC), this paper proposes a discontinuous reinforced prescribed performance control (RPPC) strategy, which takes into account both actual and converted errors and introduces innovative segmented dynamic control strategies designed for continuous operation time periods and discontinuous impulse points. To prevent unnecessary disruptive switching during transient moments, switching constraints (SCs), based on the speed and position of the rear train, are established. Additionally, leveraging Lyapunov theory, the stability of the closed-loop system for train operation is rigorously proven. Finally, a simulation scenario for a 3-train formation, along with a further data-based semi-physical simulation of a 2-train system, are presented for demonstrating the feasibility of the main results. Note to Practitioners—For railway system operators, the evolution of VC marks a significant advancement in enhancing railway capacity and addressing uneven transport demand. By minimizing headway distances when needed, it optimizes overall operational efficiency. Implementing VC, however, requires precise adjustments to the operating targets of trailing trains based on their spacing from the leading train, while considering diverse operational environments and railway line conditions. Ensuring safety demands advanced control strategies capable of dynamically adapting to these variations. This study introduces a discontinuous RPPC strategy that integrates actual and converted train tracking errors, employing segmented dynamic control for both continuous and discontinuous operation periods. This approach enhances control responsiveness and accuracy while avoiding unnecessary disruptions caused by SC techniques during transient moments through speed-position-based operation. In practice, the RPPC strategy enables railway operators to optimize scheduling, increase capacity, and effectively manage uneven transport demand. It contributes to improved operational efficiency, safety, and the overall quality of railway services.}, archive = {J_TASE}, author = {Wenxiao Si and Shigen Gao and Miao Zhang and Hang Zhang and Hongwei Wang}, doi = {10.1109/TASE.2025.3567964}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15251--15263}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforced prescribed performance control for virtually-coupled trains with saltatory targets and switching constraints}, volume = {22}, year = {2025}, }
• (2025). Online digital twin adaptive critic design with long short-term memory for wastewater treatment plants. TASE, 22, 15241--15250. (www)

  @article{ , abstract = {The wastewater treatment system is a complex unknown system with nonlinear and uncertain characteristics. It is necessary to control the concentration of the dissolved oxygen and the nitrate nitrogen at the set value in the wastewater treatment process. However, traditional control methods are difficult to meet the accuracy requirements for the required concentration value. At the same time, the digital twin (DT) technology has been widely applied to practical industrial systems like the wastewater treatment system in recent years. Based on the above backgrounds, an online DT adaptive critic design (DTACD) is developed by combining the long short-term memory (LSTM) neural network with the action-critic structure. First, a set of historical data is collected, which is used to build a digital model using LSTM. Then, the digital model is used to guide the real wastewater treatment system to control the nitrate nitrogen concentration and the dissolved oxygen concentration at set values. Finally, we select the Benchmark Simulation Model No.1 as the model for the simulation experiment. Compared with other methods, DTACD shows a better performance. Note to Practitioners—Since the concentration control of the dissolved oxygen and the nitrate nitrogen is involved in the process of the wastewater treatment, this problem can be regarded as the tracking control problem of nonlinear systems, which is an important research direction in the field of the optimal control. Adaptive dynamic programming has shown good performance in this field, which is adopted to the tracking control. In the actual wastewater treatment process, the control variables that do not meet the input requirements may affect the control performance. The DT technology can test whether the control strategy satisfies the requirements in advance by constructing a digital model. Therefore, we are committed to combining adaptive dynamic programming and DT to realize optimal tracking control for industrial systems such as the wastewater treatment system, integrating neural networks to improve the tracking control accuracy and ensure the rationality of control variables. Finally, simulation experiments are carried out on the internationally recognized Benchmark Simulation Model No.1 to verify the effectiveness of the control design.}, archive = {J_TASE}, author = {Ding Wang and Hongyu Ma and Jin Ren and Honggui Han and Junfei Qiao}, doi = {10.1109/TASE.2025.3567285}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15241--15250}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online digital twin adaptive critic design with long short-term memory for wastewater treatment plants}, volume = {22}, year = {2025}, }
• (2025). Weakly supervised deep learning for monitoring sleep apnea severity using coarse-grained labels. TASE, 22, 15227--15240. (www)

  @article{ , abstract = {Sleep apnea, a prevalent sleep-related breathing disorder, often remains undiagnosed and untreated in a large patient population due to the need of extensive manual annotations on various physiological signals for clinical diagnosis. Despite the surge of interest in applying machine learning to automate apnea detection, the effectiveness of existing techniques highly relies on strongly supervised learning that requires massive finely labeled training data for sufficiently short time intervals – a requirement often unmet due to the prohibitively high cost of manual labeling in clinical practice. In this article, we incorporate clinical knowledge to establish a weakly supervised deep learning framework for automatically estimating the latent fine-grained apnea severity when only coarse-grained labels indicating apnea presence are available in the training data. Specifically, a novel knowledge-enhanced dual-granularity consistency loss, which simultaneously considers the consistency between coarse- and fine-granularity and the integration of clinical knowledge on apnea diagnosis, is designed to boost the model’s learning of apnea severity at the fine granularity. A mathematical encoding of clinical knowledge is proposed to calibrate fine-grained estimation accuracy through ordinal alignment functions, which quantitatively relates the severity of apnea to the prominence of key diagnosis-informed physiological symptoms. The proposed method is able to accurately estimate fine-grained apnea severity in real time with significantly reduced labeling costs, extending the reach of sleep apnea diagnostics to larger population both in lab and at home. An experiment is conducted to demonstrate the superior estimation performance of the proposed method for monitoring apnea severity at high temporal resolution. Note to Practitioners—The effectiveness of existing automatic methods for sleep apnea detection is significantly constrained by the need of extensive finely labeled training datasets. Such datasets are expensive and time-consuming to acquire in practice due to the intensive manual labeling efforts involved. This article proposes a weakly supervised deep learning framework for automatically estimating accurate fine-grained apnea severity at high temporal resolution by only using less expensive, coarse-grained annotations indicating apnea presence in the training dataset. The integration of clinical knowledge related to apnea diagnosis into the proposed framework further enhances the severity estimation accuracy despite the limited availability of finely labeled data. This ability is highly desirable for practitioners by significantly reducing the burden of labor-intensive, costly fine-grained manual annotation efforts, providing a cost-effective solution to extend the reach of real-time sleep apnea diagnostics and monitoring to larger population.}, archive = {J_TASE}, author = {Xin Zan and Di Wang and Changyue Song and Feng Liu and Xiaochen Xian and Richard Berry}, doi = {10.1109/TASE.2025.3566682}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15227--15240}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Weakly supervised deep learning for monitoring sleep apnea severity using coarse-grained labels}, volume = {22}, year = {2025}, }
• (2025). Duality-based optimization of occlusion avoidance for active optical navigation system in robotic orthopedic surgeries. TASE, 22, 15215--15226. (www)

  @article{ , abstract = {In robotic orthopedic surgery, the optical tracking system (OTS) is typically placed in a fixed location. In surgery, the OTS’s line of sight is likely to be blocked. This will interrupt the navigation and affect surgical safety. To solve this occlusion problem, an RGB-D camera is used to detect possible occluders and a navigation robot is utilized to actively adjust the OTS viewpoint before occlusion occurs. To guarantee the applicability, the occluder is enveloped using a convex polytope, and a two-phase optimization method is proposed based on duality of convex optimization. The effectiveness of the proposed method is verified via simulations and experiments. Experimental results show thattheproposed method can avoid occlusion between the OTS and the occluder, and the targets are located near the center of the measurement volume. This active navigation guarantees the continuity of intraoperative navigation, and thus helps to improve the safety in robotic orthopedic surgery. Note to Practitioners—OTS is recognized as the gold standard for intraoperative navigation in orthopedic surgery. Fixed navigation is typically adopted in the operating room, where the surgeon manually adjusts the viewpoint of the OTS when necessary. However, fixed navigation has some limitations, i.e., measurement volume is limited, and the line of sight is easily blocked. To address this occlusion problem, the OTS and RGB-D camera are mounted on a 7-DOF navigation robot, and the OTS viewpoint can be actively adjusted before occlusion occurs. In this paper, complex occluders in the operating room are enveloped using convex polytopes. Occlusion avoidance is formulated as two-phase optimization based on duality. Simulations and experiments validate the effectiveness of the proposed method with Control Lyapunov Function constraints. The proposed method can maintain occlusion avoidance between the OTS and occluder. In future work, efforts will be directed towards the occlusion-free strategies of multiple occluders.}, archive = {J_TASE}, author = {Pengxiu Geng and Mengde Luo and Tianyao Li and Hongpeng Wang and Yanding Qin and Jianda Han}, doi = {10.1109/TASE.2025.3566512}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15215--15226}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Duality-based optimization of occlusion avoidance for active optical navigation system in robotic orthopedic surgeries}, volume = {22}, year = {2025}, }
• (2025). Learning 6-DoF fine-grained grasp detection based on part affordance grounding. TASE, 22, 15200--15214. (www)

  @article{ , abstract = {Robotic grasping is a fundamental ability for a robot to interact with the environment. Current methods focus on how to obtain a stable and reliable grasping pose in object level, while little work has been studied on part (shape)-wise grasping which is related to fine-grained grasping and robotic affordance. Parts can be seen as atomic elements to compose an object, which contains rich semantic knowledge and a strong correlation with affordance. However, lacking a large part-wise 3D robotic dataset limits the development of part representation learning and downstream applications. In this paper, we propose a new large Language-guided SHape grAsPing datasEt (named LangSHAPE) to promote 3D part-level affordance and grasping ability learning. From the perspective of robotic cognition, we design a two-stage fine-grained robotic grasping framework (named LangPartGPD), including a novel 3D part language grounding model and a part-aware grasp pose detection model, in which explicit language input from human or large language models (LLMs) could guide a robot to generate part-level 6-DoF grasping pose with textual explanation. Our method combines the advantages of human-robot collaboration and LLMs’ planning ability using explicit language as a symbolic intermediate. To evaluate the effectiveness of our proposed method, we perform 3D part grounding and fine-grained grasp detection experiments on both simulation and physical robot settings, following language instructions across different degrees of textual complexity. Results show our method achieves competitive performance in 3D geometry fine-grained grounding, object affordance inference, and 3D part-aware grasping tasks. Our dataset and code are available on our project website https://sites.google.com/view/lang-shape. Note to Practitioners—This paper was motivated by the problem of pragmatic 6 DoF robotic grasping for a robot in a real-world scenario. Different from existing work focusing on task-oriented solutions, our research attempts to consider fine-grained grasping as a general and essential modeling problem (i.e. Point 1: which part to grasp; Point 2: how to grasp the specific part). For that, our method fully considers the scalability and compatibility with subsequently developed functional modules. A two-stage loosely-coupled fine-grained grasping framework (LangPartGPD) is designed, including a 3D part language grounding model and a part-aware 6 DoF grasping pose detection model (Point 1). The former is used to link the symbolic-level grasped part representation to the specific region of an object. The latter is to realize kinematic-level grasping point generation based on the constraint of parts (Point 2). To realize the designed model, we propose a newly large-scale language-pointcloud-grasp dataset (LangSHAPE) to train fine-grained 3D part-level grounding and part-constrained grasping detection models. It is noted that our method is purely concerned with the target object, grasped part, and functionality to realize to model the essential fine-grained graspability, which is an insufficiently explored direction currently and could promote task-level grasping performance. The symbolic-level description of grasping in the form of natural language provides an explainable decision and a user-friendly interface for human-robot collaboration. Experiments in simulations and real-robot settings show our method can effectively enable a robot to make a cognitive decision for fine-grained manipulation. Summarily, our work is a preliminary exploration to link symbolic-level cognitive space to kinematic-level control space directly. In future research, we will further combine more complicated cognitive knowledge from public multimodal knowledge graphs and multimodal LLMs, to build a general robotic cognitive decision system using LangPartGPD as the basic components.}, archive = {J_TASE}, author = {Yaoxian Song and Penglei Sun and Piaopiao Jin and Yi Ren and Yu Zheng and Zhixu Li and Xiaowen Chu and Yue Zhang and Tiefeng Li and Jason Gu}, doi = {10.1109/TASE.2025.3566461}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15200--15214}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning 6-DoF fine-grained grasp detection based on part affordance grounding}, volume = {22}, year = {2025}, }
• (2025). Hierarchical loss constraint filter for low-visibility tiny crack detection. TASE, 22, 15188--15199. (www)

  @article{ , abstract = {Tiny crack detection plays a vital role in ensuring the safety of critical industrial components and infrastructure, providing early intervention to prevent crack propagation and mitigate potential damage. Although machine vision-based defect detection has been greatly utilized in safety inspection and maintenance, tiny crack detection remains a big issue due to the low visibility and weak features with large background noise. To address this challenge, this paper presents an end-to-end trainable hierarchical constrained neural network for tiny crack detection. Firstly, we present a hierarchical loss constraint Filter (HLCF) module based on a region-level attention mechanism to capture the holistic vein structure of cracks and enhance the faint feature extraction of tiny cracks. In order to balance local high contrast and holistic crack features, a feature fusion loss constraint is designed to reduce noise interference and refine the boundary details of cracks by learning different receptive fields in each layer. Besides that, we created a dataset consisting of tiny crack samples collected through image processing and manual labeling from industrial production, named TinyCrack. The proposed HLCF model is evaluated on TinyCrack and seven public databases. The experimental indices show that the model achieves Precision over 0.62%, Recall over 0.12%, F-score over 0.53%, and IOU over 1.12% in TinyCrack, demonstrating the best accuracy compared with other public crack detection models. The results indicate that the HLCF detection model has better performance in identifying low-visibility tiny cracks. Note to Practitioners—The motivation for this article stems from the need to develop a high-performance method of detecting tiny cracks for the safety of critical industrial components and infrastructure. Recognition Methods based on image deep learning are widely used in detecting surface cracks due to their good performances of automatically learning features and nonlinear maps. However, the detection performance of tiny cracks, as the original of coarse cracks, is not guaranteed and the attention discrepancies among coarse, thin, and tiny cracks are neglected. To enhance the detection performance of tiny cracks, we propose the deep learning structure-based HLCF model. It concentrates on exploring the selectivity between different crack regions and non-crack regions through dynamically querying in a coarse-to-fine manner. By balancing the crack classes, the model achieves better detection of tiny cracks without the loss of features of coarse cracks. The model provides a feasible way of automatic quality inspection for the safety of important infrastructures such as high-pressure vessels, bridges, and die-cast components.}, archive = {J_TASE}, author = {Lijing Zheng and Zhengkun Yi and Tiantian Xu and Can Wang and Wanfeng Shang}, doi = {10.1109/TASE.2025.3565009}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15188--15199}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical loss constraint filter for low-visibility tiny crack detection}, volume = {22}, year = {2025}, }
• (2025). A hierarchical approach-based energy consumption optimization of industrial robots. TASE, 22, 15177--15187. (www)

  @article{ , abstract = {Due to the wide application and low energy efficiency of industrial robots (IRs), energy consumption (EC) optimization techniques for them have attracted increasing attention. At present, the methods for optimizing the EC of IRs are generally simple. Energy saving for IRs through conventional methods is always achieved through optimizing the joint trajectory, which generally requires pre-planning of the joint trajectory and has limited energy efficiency improvement. Besides, it is time and labor consuming. In this paper, a hierarchical EC optimization method for IRs based on inverse kinematics solution and dynamic time-scaling method based on modified particle swarm optimization (DTS-MPSO) is proposed, with the previously established EC prediction model. For the primary EC optimization, inverse kinematics for the joint angles is performed firstly under a given initial operating posture of the IRs, then the set of joint angles with the lowest EC is selected. In the secondary EC optimization process, DTS-MPSO optimization is applied to further minimize the EC of IRs. Verification experiments with an ABB robot were conducted according to the joint angles obtained through inverse kinematics and those after secondary optimization. It is proved that EC is saved by 10.68% and 39.27% by the inverse kinematics and DTS-MPSO method respectively. The single EC for the robot to move from the initial posture to the initial operating posture was reduced by 45.76% in total when it operated under the trajectory after hierarchical optimization. The EC is effectively saved, and the trajectory for minimum EC can be automatically planned with this method. Note to Practitioners—The EC cost of IRs accounts for a large part of the manufacturing cost. The primary objective of this paper focuses on the EC optimization of IRs, and a hierarchical EC optimization method is proposed. Firstly, the inverse kinematics solution is used as the primary EC optimization method to obtain the inverse kinematics solution of the IRs. After the inverse kinematics solution is optimized by the EC prediction model, the reference trajectory with the optimal EC is determined. Then, the DTS is used as the secondary EC optimization method to scale the reference trajectory time, and finally the robot’s motion EC is minimized. The case results shows that the proposed EC optimization method can reduce EC and conserve labor and other resources.}, archive = {J_TASE}, author = {Wei Xiao and Xubing Chen and Zhongtao Fu and Guirong Han}, doi = {10.1109/TASE.2025.3564131}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15177--15187}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical approach-based energy consumption optimization of industrial robots}, volume = {22}, year = {2025}, }
• (2025). A riccati matrix equation solver design based neurodynamics method and its application. TASE, 22, 15163--15176. (www)

  @article{ , abstract = {Riccati matrix equation (RME), a critical nonlinear matrix equation in autonomous driving and deep learning. However, memory-compute separation in traditional solving systems leads to latency and inefficiency when solving nonlinear equations, particularly under real-time requirements. Existing hardware lacks dedicated accelerators for RME, no specialized solvers quick addressing its nonlinear complexity. To address this issue, we propose a novel RME solver based on a memristive array, which leverages the parallel and fast computing advantages of analog circuits to quickly solve any order RME. Inspired by Neurodynamics for non-linear matrix equations, we first introduce an innovative Neurodynamics-based RME solving algorithm specifically designed from an analog circuit perspective. Based on this algorithm, we constructed a pioneering closed-loop analog circuit solver, overcoming bottlenecks in using circuits for such nonlinear matrix equations. Our evaluation demonstrates that the proposed solver achieves over 90% accuracy for a 128th-order parameter Riccati matrix equation. Compared to traditional digital processors, the solver offers significant energy efficiency advantages and is three orders of magnitude faster than CPU. Additionally, the solver successfully accelerates our proposed dung beetle optimizer-linear quadratic control algorithm for vehicle suspension control, achieving high precision while significantly reducing time and energy consumption compared to CPU and GPU. Note to Practitioners—The motivation for this study arises from the need for Riccati matrix equation in real world applications. Currently the main factors leading to the high time complexity of Riccati matrix equation (e.g., in the optimal algorithm solving, trajectory tracking and autonomous driving) originate from the large number of iterative computations, and the nonlinear structure of its essence. As such, the work in this paper propose a Neurodynamics-based algorithm for solving nonlinear matrix equation, and designs an efficient, non-iterative solver for Riccati matrix equations based on the proposed algorithm, essentially addressing the two problems mentioned above. Although the proposed circuit does not seem to be as perfect as numerical computation in terms of solution accuracy, the comparison with literature shows that the accuracy is generally sufficient to meet the usage requirements. It can also provide an initial solution for high precision solving, thereby saving a significant amount of time in searching for the initial solution, its effectiveness is preliminarily validated through applications demonstrated in the manuscript. In the future, we aim to advance the proposed RME solver for broad practical deployment in applications such as optimal control, edge computing, path tracking and so on.}, archive = {J_TASE}, author = {Pingdan Xiao and Junjie Fang and Zhengmiao Wei and Yu Dong and Sichun Du and Shiping Wen and Qinghui Hong}, doi = {10.1109/TASE.2025.3567981}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15163--15176}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A riccati matrix equation solver design based neurodynamics method and its application}, volume = {22}, year = {2025}, }
• (2025). Extended state observer based-optimal control of a microrobot navigation along a trajectory generated by a joystick device. TASE, 22, 15150--15162. (www)

  @article{ , abstract = {Trajectory tracking of microrobots presents a significant challenge in biomedical research applications such as drug delivery and artery unclogging. Microrobots are inherently subject to uncertainties and unmodeled dynamics, complicating their control. Additionally, the overheating of coils in the electromagnetic actuation system can limit controller performance. In this paper, we propose an observer-based optimal control strategy for microrobots, where the desired trajectory is generated online by means of a joystick. The control law addresses the problem of coils overheating by minimizing energy consumption while achieving trajectory tracking despite uncertainties and unmodeled dynamics. For practical implementation, we introduce a modified nonlinear extended state observer to estimate the entire state of the microrobot. Analytical tuning approaches for the control law and observer are developed using D-pole placement technique and Kharitonov’s theorem. This solution is an alternative to the use of cooling systems, which require more energy consummation and often reduce the performance of the controller. The tracking performance of the proposed control strategy is evaluated through simulation and experimental tests on a dedicated platform. Both simulation and practical results demonstrate the effectiveness of the developed control strategy. Note to Practitioners—This work is motivated by the fact that ensuring effective targeted drug delivery and safe navigation of magnetic microrobots in biomedical applications requires an electromagnetic actuation system (EMA) with minimal magnetic energy consumption and coil heating. This can be accomplished through the optimal design of the EMA and the incorporation of a cooling system. However, while this solution is interesting, it can make the EMA cumbersome and heavy, which reduces the performance of the controller in tracking the desired trajectory. Besides, the use of a cooling system implies increased energy consumption. Thus, in this work, we develop and implement a control strategy that combines optimal control with an observer to minimize magnetic energy and, consequently, coil heating. The performance of this control strategy in forcing the microrobot to track a reference trajectory, generated in real-time using a joystick, is demonstrated through experiments conducted in an endovascular environment at different scales. The results show that the proposed control strategy successfully achieves trajectory tracking despite uncertainties and environmental disturbances, with no coil heating.}, archive = {J_TASE}, author = {Meziane Larbi and Ahmed Maidi and El-Hadi Guechi and Karim Belharet}, doi = {10.1109/TASE.2025.3566475}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15150--15162}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Extended state observer based-optimal control of a microrobot navigation along a trajectory generated by a joystick device}, volume = {22}, year = {2025}, }
• (2025). Constrained tracking control of MIMO nonlinear systems with discontinuous references and unknown dynamics. TASE, 22, 15139--15149. (www)

  @article{ , abstract = {This paper is dealt with the tracking control problem for the multi-input multi-output (MIMO) block-triangular nonlinear systems with output constraints under discontinuous references. It is focused on the case where the system exhibits inherent nonlinearities, e.g., radically unbounded nonlinearities, and totally unknown dynamics. This renders the existing solutions infeasible. To surmount this challenge, a novel hybrid tracking control strategy composed of a robust decoupling constrained controller and a proportional controller is devised in this paper. It guarantees that the system outputs evolve within the prescribed constraint bands and track the discontinuous references with the tunable settling time and accuracy. Moreover, the controller shows a significant simplicity. No attempt is made for parameter identification, function approximation, disturbance estimation, derivative calculation or command filtering, despite the unknown system dynamics and the recursive control design. The theoretical findings are validated by a comparative experiment on a 2-DOF serial flexible link (2DSFL) robot. Note to Practitioners—MIMO nonlinear systems are widely present in practice, e.g., robotic manipulators, ground vehicles and aircrafts. For operational specifications and safety considerations, they are always subject to output constraints. On the other hand, building an exact mathematical model for a practical system is unfulfillable. The existing constrained control methods for uncertain MIMO nonlinear systems work for the case where the control distribution matrix is norm-bounded or the control distribution matrix, additive nonlinearity, Jacobian matrix or disturbance bound is known. Moreover, the control algorithms are complex due to the dependence on parameter identification, function approximation and disturbance estimation. Additionally, they fail to be applied to discontinuous references, which are common in vehicle guidance, industrial electronics and robotic applications. The related tracking control approaches are effective for the regular discontinuity of the reference, the bounded nonlinearities of the system, the known system dynamics, or the known bounding functions of the nonlinearities. To overcome the above challenges, a low-complexity hybrid tracking control strategy composed of a robust decoupling constrained controller and a proportional controller is developed in this paper. It steers the system outputs to track the discontinuous references without constraint violation. The result of a comparative experiment on a 2DSFL robot show the effectiveness and advantage of the proposed approach.}, archive = {J_TASE}, author = {Jin-Xi Zhang and Wei-Li Qi and Tianyou Chai}, doi = {10.1109/TASE.2025.3566443}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15139--15149}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Constrained tracking control of MIMO nonlinear systems with discontinuous references and unknown dynamics}, volume = {22}, year = {2025}, }
• (2025). Predefined-time event-triggered adaptive tracking control for uncertain time delay nonlinear systems. TASE, 22, 15129--15138. (www)

  @article{ , abstract = {This paper raises a predefined-time event-triggered adaptive tracking controller for nonlinear systems with time delays. To mitigate the effects of time-varying delays, a Lyapunov-Krasovskii function is initially introduced. A set of adaptive laws and virtual controllers with switching functions are designed to avoid singularity phenomena. In addition, the event-triggered mechanism with the switching threshold can effectively conserve communication resources while avoiding the occurrence of Zeno behavior. It is theoretically demonstrated that, under the devised control strategy, the system output successfully tracks the reference signal with negligible tracking error within a specified time. Lastly, the validity of the raised control strategy is demonstrated through examples. Note to Practitioners—This paper explores a predefined-time adaptive tracking controller for nonlinear systems with time delays. For instance, in engineering applications such as robotics, completing tasks within a predefined time frame is critical. Failure to account for time delays may cause the control strategy to fall short of achieving the desired performance or even lead to instability. In this context, the development and investigation of event-triggered control are essential in modern control theory, as they effectively address the challenges posed by resource constraints. Considering the aforementioned factors, this paper addresses time delays and proposes a predefined-time event-triggered adaptive tracking control strategy based on backstepping and the Lyapunov-Krasovskii function. The theoretical analysis and simulation results demonstrate that the proposed scheme achieves the expected control objectives.}, archive = {J_TASE}, author = {Lihong Gao and Zhen Wang and Xia Huang and Hao Shen}, doi = {10.1109/TASE.2025.3566407}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15129--15138}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined-time event-triggered adaptive tracking control for uncertain time delay nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Drone-based wall crack detection using model-agnostic meta-learning. TASE, 22, 15116--15128. (www)

  @article{ , abstract = {With the urbanization process and aging of buildings, wall crack detection plays a crucial role in the maintenance and safety of building structures. Due to the inherent characteristics of defects, however, cracks in the wall are relatively sparse, compared with the normal area. High-altitude regions are also difficult to access. As a result, large wall crack data sets are rare. This issue impairs the training of deep networks and may lead to the suboptimal detection result. To address this issue, we first capture a set of pure wall crack images using a drone, which are comprised of a new wall crack data set, namely, Ocean University of China Wall Crack Data Set (OUC-Crack). In contrast to existing crack data sets, OUC-Crack only contains wall crack images and hence it is particularly useful for wall crack detection. Then we propose a drone-based wall crack detection system, which consists of a drone platform and a crack detection network referred to as the Model-Agnostic Meta-Learning Based Segmentation Network (MAML-SegNet). Therefore, only a small number of training images are required. To fulfill the real-time detection task on drones, we further develop an efficient MAML-SegNet (EFF-MAML-SegNet). The proposed systems are tested on the OUC-Crack and two publicly available crack data sets, including Volker and Crack500. Experimental results show that our MAML-SegNet outperforms 14 baselines in terms of Dice Coefficient, with improvements of at least 1.26%, 13.19% and 7.12% on the three data sets, respectively. Given that the EFF-MAML-SegNet is deployed on a drone, real-time detection can be achieved with the Dice Coefficient values of 66.83%, 55.42% and 48.03% on the three data sets, respectively. These promising results should be due to the few-shot learning ability of meta-learning and the efficient network design. (Code, data and models are available at https://indtlab.github.io/projects/Wall-Crack-Detection) Note to Practitioners—The proposed system leverages the MAML that we deliberately design to perform the wall crack detection task using only a small number of training images. This system addresses the issue of insufficient training data in the scenario of wall crack detection. As a result, the challenge of data scarcity that practitioners in the community of wall crack detection have to encounter can be alleviated. In addition, the MAML-SegNet is built using a lightweight network design. This network enables the accurate crack detection with the less computational burden and a small number of training images, which increases the practical applications of the system. Furthermore, the deployment of the EFF-MAML-SegNet on the drone is able to perform the real-time wall crack detection task, due to the efficient design of the network. Compared with manual high-altitude inspection operations, the proposed system reduces both the cost and the risk while greatly improving the work efficiency. This system potentially facilitates the early detection and treatment of wall cracks, enabling the formulation of the effective maintenance and repair strategies. Moreover, the OUC-Crack data can be used by the researchers and engineers in the community for further research.}, archive = {J_TASE}, author = {Borong Xu and Wenxuan Shao and Xinghui Dong}, doi = {10.1109/TASE.2025.3565647}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15116--15128}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Drone-based wall crack detection using model-agnostic meta-learning}, volume = {22}, year = {2025}, }
• (2025). Exploring the synergistic effects of teleoperation scaling ratio and learning from demonstration. TASE, 22, 15106--15115. (www)

  @article{ , abstract = {Teleoperation and Learning from Demonstration (LfD) are complementary paradigms for robotic control, yet their synergistic potential remains underexplored. This work proposes a unified framework that dynamically adjusts teleoperation precision using task-specific priors from LfD while optimizing demonstration learning through teleoperation scaling. By analyzing human operator signals (e.g., muscle activity), our method autonomously scales robot movements to balance the precision requirements and execution efficiency during teleoperation. Conversely, teleoperation data informs the adaptive arrangement of motion primitives in LfD, improving trajectory accuracy in critical task phases. Experiments show that the proposed method can improve the efficiency of teleoperation tasks and the accuracy of task learning. This bidirectional synergy offers a practical pathway to enhance both human-robot interaction and skill learning, particularly in applications demanding variable precision, such as assembly and surgical robotics. Note to Practitioners—Robot teleoperation technology is widely used in medical, aviation and industrial fields, and LfD technology is a method for rapid programming and deployment of robot autonomous tasks. The work of this paper explores the common points of robot teleoperation technology and LfD technology, as well as the demand for task accuracy. Through the fusion and mutual promotion of the two methods, the autonomous zooming teleoperation method can be applied to teleoperation tasks such as surgery, and the time problem caused by the zooming adjustment can be reduced. At the same time, the LfD method, which considers scaling, is formed and can be used in learning and programming for different task precision requirements.}, archive = {J_TASE}, author = {Donghao Shi and Sihan Jin and Chenguang Yang and Zhenyu Lu and Qinchuan Li}, doi = {10.1109/TASE.2025.3565317}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15106--15115}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exploring the synergistic effects of teleoperation scaling ratio and learning from demonstration}, volume = {22}, year = {2025}, }
• (2025). High-order fully actuated system approach-based attitude stabilization for underactuated rigid and flexible spacecraft. TASE, 22, 15094--15105. (www)

  @article{ , abstract = {This paper addresses the challenge of attitude stabilization for a class of underactuated rigid and flexible spacecrafts by utilizing only two control inputs. The attitude stabilization problem of underactuated spacecraft poses substantial challenges to the High-Order Fully Actuated (HOFA) system approach, due to the fact that it heavily depends on the full actuation characteristics of the system. To circumvent the aforementioned limitation, we first derive a HOFA system for underactuated rigid spacecraft using homogeneity theory and averaged system approach. Subsequently, we design a controller based on the HOFA system approach, which yields a linear closed-loop system. This methodology is then extended to address the simplified attitude stabilization problem of underactuated flexible spacecraft. By utilizing the HOFA system approach in tandem with an observer for estimating modal variables, we propose a control strategy that not only stabilizes the attitude but also actively suppresses elastic vibration. The performance of the proposed control schemes are demonstrated through simulation studies. Note to Practitioners—The motivation of this paper is to address the application of the HOFA system approach to underactuated systems, specifically in underactuated rigid and flexible spacecraft. Due to insufficient control torques, underactuated spacecraft exhibits nonholonomic behavior, which increases the complexity of control. In practical scenarios, modern spacecraft often carries large flexible appendages, which induce sustained vibrations by disturbances and severely affect the attitude stability of spacecraft. The attitude control of underactuated flexible spacecraft is even more challenging. Traditionally, the attitude stabilization of underactuated spacecraft is solved via the first-order state-space method, which is suitable for solving state variables rather than control variables. Due to the limited computational resources on spacecraft, a concise control structure is required. Although the HOFA system approach can provide such a structure, it is designed for fully actuated systems and cannot apply to underactuated systems. To deal with these challenges, this paper proposes a method that transforms an underactuated spacecraft into a fully-actuated system. Building upon the attitude stabilization of underactuated rigid spacecraft, the paper also addresses the attitude stabilization of underactuated flexible spacecraft with vibration suppression. Notably, the proposed method provides a new method for applying the HOFA system approach to underactuated systems, expanding the engineering applications of this approach.}, archive = {J_TASE}, author = {Shixiang Jia and Jianbin Qiu and Tong Wang}, doi = {10.1109/TASE.2025.3567541}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15094--15105}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-order fully actuated system approach-based attitude stabilization for underactuated rigid and flexible spacecraft}, volume = {22}, year = {2025}, }
• (2025). Blockchain-integrated cyber-physical smart meter design and implementation for secured energy trading in virtual power plants. TASE, 22, 15083--15093. (www)

  @article{ , abstract = {A novel architecture for smart meters in a Virtual Power Plant (VPP) is introduced in this research. By integrating blockchain technology, the system not only measures and quantifies diverse consumer data but also facilitates immediate control, hence improving demand responsiveness in a VPP environment. Mathematical models were created to optimize profit, battery reserve, and power balance. A novel transaction and security algorithm that enables peer-to-peer (P2P) transactions in a secure setting is used in conjunction with a power flow algorithm for real time monitoring and control to implement the proposed model. The lightweight characteristics of the algorithms enable faster and more effective computer processing. The unique identifier issued to each smart meter facilitates seamless integration with a blockchain smart contract, therefore enabling improved and secure P2P transactions. An innovative experimental setup demonstrated the framework’s ability to effectively manage energy flows while maintaining seamless wireless connection with the grid and executing transactions. The smart meter demonstrated exceptional efficiency in load management, resulting in an average loss of 1.9524W. A dedicated dapp was created just for this purpose. Through the strategic integration of algorithms and blockchain technology, this framework enhances the efficiency and reliability of the metering infrastructure, while also enabling secure transactions. Note to Practitioners—This paper introduces a blockchain-integrated smart metering system tailored for VPPs, addressing challenges like energy inefficiency, cybersecurity risks, and real-time energy management. The design leverages blockchain technology to enable secure, P2P energy trading and real-time energy monitoring, presenting a scalable solution for decentralized energy markets. Practitioners in the energy sector, especially those managing distributed energy resources, can use this framework to enhance grid stability and improve energy distribution efficiency. The smart meter’s plug-and-play design allows for seamless integration into existing infrastructures, making it adaptable for various regulatory environments. The incorporation of IoT ensures real-time load management and communication, while blockchain safeguards transactional data, reducing risks of data breaches and fostering trust among prosumers. However, implementing this system requires careful consideration of hardware compatibility, scalability for larger networks, and user training to ensure adoption. The system’s reliance on real-time data processing and its hybrid approach to blockchain integration could challenge practitioners unfamiliar with these technologies or those having limited network capability. Potential extensions include integrating advanced predictive algorithms for improved energy forecasting and adapting the system for multi-energy management (e.g., heating and cooling). Further research on long-term system performance and cost-benefit analysis could enhance its practical value for stakeholders in energy markets and smart grid ecosystems.}, archive = {J_TASE}, author = {A. M. A. Daiyan Kaif and Khandoker Shahjahan Alam and Sajal K. Das and Guo Chen and Syed Islam and S. M. Muyeen}, doi = {10.1109/TASE.2025.3566938}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15083--15093}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Blockchain-integrated cyber-physical smart meter design and implementation for secured energy trading in virtual power plants}, volume = {22}, year = {2025}, }
• (2025). Rigid geometry formation subject to visibility constraints using heading angle correlation based on leader-follower system. TASE, 22, 15070--15082. (www)

  @article{ , abstract = {This article proposes a tracking model for nonholonomic constraint robots, enabling the realization of leader-steered rigid geometry formations. By employing cameras and local leader-based approaches, the challenge posed by traditional separation-bearing control methods, which are incapable of establishing rigid formation for both translational and rotational control, is resolved. In addition, to maintain the connectivity of the sensing topology, the field-of-view (FOV) constraints of the on-board cameras are integrated into the controller design. A conversion approach is used to translate the FOV constraints into a rigid geometry formation. Additionally, there is a trade-off between visibility constraints and the leader-steered rigid geometry formation, particularly when the trajectory of the global leader has significant curvature. To address this problem, a continuously smooth transition function is employed. Ultimately, a fixed-time distributed control protocol and distributed observers are developed to realize the formation framework. Experimental results demonstrate that the proposed control protocol effectively achieves rigid geometric formations and satisfies FOV constraints. Note to Practitioners—Robot formations with nonholonomic constraints are widely used in various applications. However, global communication conditions may not be available due to privacy concerns and network load. Additionally, in practical scenarios, it is preferable for the robot formation geometry to be rigid. Traditional formation error models, especially in scenarios involving local leaders, impede achieving rigid geometric formations. This paper proposes an error tracking model based on visual sensing to achieve rigid geometric formations. A PPC method is introduced to ensure that the FOV remains within specified limits. A switching function is employed to maintain the communication topology’s integrity consistently. Differentiators and fixed-time distributed control protocol are utilized to enhance formation performance. Experimental results validate that this approach facilitates rigid geometric formations while consistently maintaining visual constraints. Engineers can implement rigid geometric formations using this approach without needing to redesign controllers, simply by incorporating the additional angular offset proposed in this paper into the existing error model. With this change, we will further explore rigid formation transformations in the presence of obstacles.}, archive = {J_TASE}, author = {Zhongchao Liang and Zijian Wang and Zhongguo Li and Jun Yang}, doi = {10.1109/TASE.2025.3566367}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15070--15082}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rigid geometry formation subject to visibility constraints using heading angle correlation based on leader-follower system}, volume = {22}, year = {2025}, }
• (2025). Secure fault-tolerant control for nonlinear cyber-physical systems against multiple threats. TASE, 22, 15060--15069. (www)

  @article{ , abstract = {This paper mainly focuses on secure fault-tolerant control for nonlinear cyber-physical systems under the influence of multiple threats, i.e., sensor and actuator saturations, actuator fault, and intermittent denial-of-service attacks. A novel secure fault-tolerant control method is proposed within the Takagi-Sugeno fuzzy modeling framework, considering both saturation constraints and intermittent denial-of-service attacks, to ensure the secure performance of cyber-physical systems under multiple threats. By leveraging the advantages of the proposed method, the influence of outdated information caused by denial-of-service attacks is mitigated, and more precise state/fault estimations are obtained. Furthermore, the method guarantees the stability of both the estimation error system and the closed-loop system, even in the presence of intermittent denial-of-service attacks and saturation constraints. Finally, the effectiveness of this method is validated through experiment. Note to Practitioners—CPSs form the core of the next generation of manufacturing, being applicable to driverless autonomous systems, smart grids, smart healthcare, and various other fields. Nevertheless, this broad scope also exposes them to heightened risks. In practical engineering, actuator faults and saturations resulting from the complex environment or the inherent complexity of the system can impact the system’s performance. Moreover, as network technology continues to be applied in practical engineering, networks will not only enhance work efficiency and provide convenience but also introduce additional risks, that is, open networks are susceptible to network attacks, which will adversely affect actual production. Building upon the above analysis, this paper proposes a secure fault-tolerant control method to counter saturation constraints and intermittent DoS attacks. It can effectively depict the saturation phenomenon in the system, compensate for the impact of actuator failure, and reduce the impact of outdated information brought by denial of service attacks on the system control performance.}, archive = {J_TASE}, author = {Ximing Yang and Tieshan Li and Yue Long and Hanqing Yang}, doi = {10.1109/TASE.2025.3565567}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15060--15069}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure fault-tolerant control for nonlinear cyber-physical systems against multiple threats}, volume = {22}, year = {2025}, }
• (2025). Dynamic path planning for parafoil homing in surface wind disturbance environment. TASE, 22, 15045--15059. (www)

  @article{ , abstract = {The parafoil is a flexible aircraft with good load characteristics and endurance. Compared with common rotor Uncrewed Aerial Vehicles (UAVs), its excellent gliding characteristics make it widely used in fixed-point airdrop missions in the aviation field. However, while the flexible flight structure brings excellent flight characteristics, the homing of the parafoil will be seriously disturbed by wind. This paper studies the wind disturbance problem of parafoils, and proposes a dynamic homing planning strategy derived from the forward model based on the parafoil dynamic model. Focusing on the homing process of the parafoil, the wind repulsion potential field is established for the disturbance of the low-altitude wind field by Artificial Potential Field (APF). Recursive navigation planning is carried out based on Dynamic Path Rapid-exploration Random Tree (DP-RRT). We validate the dynamic programming algorithm through simulation and experimental environments. The results show that the dynamic programming algorithm can effectively improve the parafoil homing process’s flight control stability and homing accuracy. Note to Practitioners—The parafoil is a flexible aircraft widely used in both military and civilian applications, particularly in the aviation field. For instance, by combining the characteristics of a flexible flight structure with good controllability, parafoils have been extensively researched for Mars landing missions. This paper proposes a dynamic homing planning strategy based on DP-RRT to address issues of low stability and inadequate accuracy during the parafoil homing stage. Through a model-derived homing planning method, the parafoil can effectively complete homing tasks under specific dynamic guidance conditions. More importantly, this strategy examines the unstable flight states caused by variations in control inputs during the homing process of various conventional and aviation equipment. Ultimately, a homing guidance strategy that achieves more stable control effects is proposed.}, archive = {J_TASE}, author = {Zhenping Yu and Hao Sun and Qinglin Sun and Panlong Tan and Zengqiang Chen and Juntian Qu}, doi = {10.1109/TASE.2025.3563071}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15045--15059}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic path planning for parafoil homing in surface wind disturbance environment}, volume = {22}, year = {2025}, }
• (2025). Path integral policy improvement and dynamic movement primitives fusion-based impedance force control with error loop correction. TASE, 22, 15034--15044. (www)

  @article{ , abstract = {Path Integral Strategy Improvement(PI2)-based impedance control is a superior scheme for preventing damage to the physical structure of the fruit during the harvesting process. However, it is highly sensitive to disturbances and has limited generalization ability during the parameter learning process, making it difficult to apply the correct gripping force to fruits with uncertain stiffness. To solve this problem, this paper proposes a variable impedance force control method that integrates PI2 with Dynamic Movement Primitives (DMPs), supplemented by a force error correction loop. Firstly, an adaptive impedance parameter matching mechanism based on gain schedules is designed to facilitate dynamic estimation of impedance parameters and enable precise force control. To further accelerate impedance parameter matching in unknown environments, the PI2 algorithm is introduced to optimize gain schedules, and DMPs are integrated to suppress disturbances, thereby improving the generalization ability of the impedance model’s parameter learning. In addition, an additional force error control loop has been designed to minimize the deviation between the desired and actual gripping force. Finally, the effectiveness of the proposed method is verified through simulation and experiment in fruit teleoperation picking robot. Note to Practitioners—In the local-remote heterogeneous fruit teleoperation picking robot system, preventing fruit damage is critical to improving harvesting efficiency and quality. However, the existing PI2 algorithm suffers from limited generalization ability during the parameter learning process and is highly susceptible to environmental disturbances, presenting challenges in force control. To address this issue, this study proposes integrating DMPs into PI2 to optimize gain schedules and designing a force error control loop to refine trajectories, thereby achieving the force control objectives and effectively compensating for the shortcomings of PI2. Simulation and experimental results validate the effectiveness of the proposed method. Future research will focus on enhancing the algorithm’s adaptability to tackle the challenges posed by environmental variability in force control.}, archive = {J_TASE}, author = {Mujie Liu and Haifei Chen and Lijun Li and Zhiqiang Ma and Yong Xu and Hui Zhang}, doi = {10.1109/TASE.2025.3562677}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15034--15044}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Path integral policy improvement and dynamic movement primitives fusion-based impedance force control with error loop correction}, volume = {22}, year = {2025}, }
• (2025). Autonomous price-aware energy management system in smart homes via actor-critic learning with predictive capabilities. TASE, 22, 15018--15033. (www)

  @article{ , abstract = {The energy consumed by buildings is expected to significantly rise in the upcoming years, necessitating intelligent Home Energy Management Systems (HEMS) that create comfortable conditions for their inhabitants, while also offering sustainable and cost-effective solutions. The building environment, however, includes multiple time-varying parameters that cannot be controlled, such as the output of renewable energy sources, the market-dependent electricity prices, the outdoor temperature, as well as the occupants’ energy habits. To overcome these barriers, we propose a hybrid Machine Learning (ML) algorithm for smart HEMS control, leveraging the properties of a decision-making deep deterministic policy gradient model, enhanced by the predictive capabilities of long short-term memory networks. Hence, the proposed algorithm aims to achieve an optimal balance between energy cost and occupant comfort by continuously adjusting the energy provided to the heating, ventilation, and air conditioning system, as well as controlling the energy storage system of the smart home. The proposed hybrid method is validated with simulations using real-world data and compared against baseline approaches, showcasing its effectiveness to achieve an optimal trade-off between the indoor temperature deviation and the average energy cost. Note to Practitioners—This paper is motivated by the need of automated energy management in smart buildings to achieve minimal energy consumption. Continuous demands for heating, ventilation and air conditioning, as well as utilization of household loads lead to a significant increase in the daily energy consumption. This paper explores the trade-off between the price that a household pays in return for indoor temperature comfort and coverage of the energy requirements. For this purpose, we provide a comprehensive mathematical modeling of a smart home environment, followed by the description of a real-time decision-making energy management algorithm that is based on Machine Learning (ML). This algorithm takes into account the forecasting of energy consumption and energy production of the household in the future, optimally adjusting the power that is drawn from the grid. By implementing this algorithm using real-world data, we demonstrate that a smart home can maintain a comfortable indoor temperature, while minimizing the energy cost and boosting the use of local renewable energy. In this context, manufacturers, developers of smart home systems or energy providers that specialize in building automation systems can utilize this algorithm to offer consumers automated energy management that optimizes cost and comfort, leading to sustainable operations. We plan to extend this model to energy communities consisting of multiple smart homes that can also exchange energy surplus to achieve zero-sum energy footprint.}, archive = {J_TASE}, author = {Sotirios T. Spantideas and Anastasios E. Giannopoulos and Panagiotis Trakadas}, doi = {10.1109/TASE.2025.3566390}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15018--15033}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous price-aware energy management system in smart homes via actor-critic learning with predictive capabilities}, volume = {22}, year = {2025}, }
• (2025). Distribution locational marginal emission for carbon alleviation in distribution networks: Formulation, calculation, and implication. TASE, 22, 15006--15017. (www)

  @article{ , abstract = {Regulating the proper carbon-aware intervention policy is one of the keys to emission alleviation in the distribution network, whose basis lies in effectively attributing the emission responsibility using emission factors. This paper establishes the distribution locational marginal emission (DLME) to calculate the marginal change of emission from the marginal change of both active and reactive load demand for incentivizing carbon alleviation. It first formulates the day-head distribution network scheduling model based on the second-order cone program (SOCP). The emission propagation and responsibility are analyzed from demand to supply to system emission. Considering the complex and implicit mapping of the SOCP-based scheduling model, the implicit theorem is leveraged to exploit the optimal condition of SOCP. The corresponding SOCP-based implicit derivation approach is proposed to calculate the DLMEs effectively in a model-based way. Comprehensive numerical studies are conducted to verify the superiority of the proposed method by comparing its calculation efficacy to the conventional marginal estimation approach, assessing its effectiveness in carbon alleviation with comparison to the average emission factors, and evaluating its carbon alleviation ability of reactive DLME. Note to Practitioners—This paper proposes the novel distribution locational marginal emission (DLME) to calculate the marginal change of emission from the marginal change of both active and reactive load demand for incentivizing carbon alleviation in the distribution network. The proposed SOCP-based implicit derivation approach calculates the DLMEs effectively in a model-based way. The DLME can incentivize effective demand response for carbon alleviation in the distribution network. The proposed method can be applied to the distribution network management for carbon alleviation and emission reduction. We conduct case studies to verify the effectiveness of the proposed method in various distribution networks under various scenarios. Numerical results show that the proposed DLME can enhance 10%-200% of the carbon alleviation effectiveness compared to the average emission factors.}, archive = {J_TASE}, author = {Linwei Sang and Yinliang Xu and Hongbin Sun and Zaijun Wu and Qiuwei Wu and Wenchuan Wu}, doi = {10.1109/TASE.2025.3566098}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {15006--15017}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distribution locational marginal emission for carbon alleviation in distribution networks: Formulation, calculation, and implication}, volume = {22}, year = {2025}, }
• (2025). Design of predefined-time convergence parameter estimator with relaxed excitation conditions. TASE, 22, 14997--15005. (www)

  @article{ , abstract = {Parameter estimation plays a significant role in the performance of adaptive controllers, yet its convergence relies on persistent excitation (PE) condition. In this paper, a novel design method of parameter estimator is proposed for a class of nonlinear systems with linear-in-parameter (LIP) uncertain parameters. The method achieves non-increasing monotonicity of parameter estimation error and predefined-time convergence. By utilizing historical data and instant data to expand the parameter regression matrix and introducing a perturbation matrix in this process, a novel parameter estimator is constructed. Through stability analysis of the parameter estimation error, it was verified that when a sufficiently small perturbation matrix is selected, the estimator can achieve predefined-time convergence of the parameter estimation error under the sufficient excitation (SE) condition, which is strictly weaker than the PE condition. Meanwhile, the numerical simulation results considering identification and control problems demonstrate the effectiveness and superiority of the proposed method. Note to Practitioners—Uncertainty parameters are ubiquitous in numerous systems, such as spacecraft, robotic arms, and ships. When designing controllers for these systems with uncertainties, it is often necessary to incorporate parameter estimators to approximate the uncertain parameters, thereby enhancing the performance of the control system. Traditional parameter estimators typically rely on persistent excitation conditions (e.g., requiring the controlled object to track periodic trajectories) to ensure the convergence of estimation errors. However, we aim to avoid imposing additional motions solely to achieve accurate estimation. Furthermore, apart from estimation accuracy, we require the estimation errors to exhibit monotonically non-increasing behavior and for the convergence time to be predefined. Motivated by these practical considerations, this paper presents a parameter estimator that guarantees predefined-time convergence of estimation errors under relaxed excitation conditions, along with non-increasing monotonicity of the estimation errors. Notably, this approach imposes no additional requirements on the controller design, allowing it to be seamlessly integrated with existing controllers or used independently as a system identification method. Finally, the effectiveness and superiority of the proposed method are validated through numerical simulations.}, archive = {J_TASE}, author = {Jianhui Wang and Guangping He and Shixiong Geng}, doi = {10.1109/TASE.2025.3566013}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14997--15005}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design of predefined-time convergence parameter estimator with relaxed excitation conditions}, volume = {22}, year = {2025}, }
• (2025). Dual-mode model predictive control for constrained networked control system with DoS attacks and disturbances. TASE, 22, 14987--14996. (www)

  @article{ , abstract = {A dual-mode model predictive control (MPC) strategy is tailored for a networked control system (NCS) operating in a constrained environment that is vulnerable to additive disturbances as well as denial-of-service (DoS) attacks. DoS attacks disrupt communication, affecting data and control signals. This article optimizes and improves the traditional MPC strategy to solve the problems caused by DoS attacks on the system. A robustness constraint with scaling parameters is introduced to efficiently handle uncertainty in the MPC algorithm. The control strategy proposed in this article enables a dynamic shift in control modes. When the system state enters a predefined set, the control strategy transitions from optimization to a state feedback control law. In order to analyze the recursive feasibility of the optimization problem and system stability, the relationship between the system errors, disturbances, and DoS attacks is derived. We also derive conditions on disturbance upper bound, DoS attack maximum allowable duration, and scaling parameters should be satisfied to ensure the recursive feasibility of the optimization problem and system stability. Note to Practitioners—In network control systems, since the system status and control signals need to be transmitted between the controller and the controlled plant through the network, there are many interferences in the network and the signal transmission will be interrupted due to manual attacks, which makes the controller unable to effectively control the controlled plant. At the same time, because of the superiority of model predictive control over other control methods, we designed a model predictive control strategy that can cope with attacks and interference. In addition, since the more complex the algorithm in the actual control scenario, the longer the calculation time, the lower the real-time performance of the control signal, our control method combines feedback control with a simpler algorithm and designs a dual-mode control method.}, archive = {J_TASE}, author = {Li Qiu and Zhixing Shao and Lingtao Dong and Xiaoshan Bai}, doi = {10.1109/TASE.2025.3565799}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14987--14996}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-mode model predictive control for constrained networked control system with DoS attacks and disturbances}, volume = {22}, year = {2025}, }
• (2025). Active security control for networked jumping systems under asynchronous dual-channel DoS attacks: An HMM-based approach. TASE, 22, 14976--14986. (www)

  @article{ , abstract = {This paper aims to address the design problem of asynchronous controllers for networked jumping systems (NJSs), in which two communication channels are vulnerable to asynchronous denial-of-service (DoS) attacks. To this end, a hidden Markov model (HMM)-based active security control strategy is proposed to counter asynchronous dual-channel DoS attacks. Firstly, two maximum consecutive numbers are introduced to describe DoS attacks randomly initiated by adversaries. Subsequently, a mode-dependent predictor is designed to generate predictive states, which can be employed in the switched controller to stabilize the NJSs. Furthermore, sufficient conditions are derived by constructing Lyapunov functions to ensure the asymptotic mean-square stability of the NJSs under asynchronous dual-channel DoS attacks. Finally, a simulation using a space robot manipulator model is presented to validate the effectiveness and practicality of the proposed active security control strategy. Note to Practitioners—This paper tackles the challenge of designing robust controllers for networked systems that are susceptible to asynchronous DoS attacks over two communication channels. In practical terms, this research is particularly relevant for systems like space robots or other critical networked infrastructures where communication interruptions can severely impact performance. The proposed solution employs an HMM-based active security control strategy to counter these attacks. The method works by predicting system states even when communication channels are compromised, allowing the controller to maintain stability and performance. This strategy is designed to handle unpredictable and random DoS attacks, which can occur independently on either communication channel. The effectiveness of this method has been validated through simulations with a space robot manipulator, highlighting its potential applicability in real-world scenarios. Although the proposed strategy shows promise, further testing in diverse operational environments is recommended to fully evaluate its reliability and scalability in practice.}, archive = {J_TASE}, author = {Peng Cheng and Hu Ye and Di Wu and Weidong Zhang}, doi = {10.1109/TASE.2025.3565758}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14976--14986}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active security control for networked jumping systems under asynchronous dual-channel DoS attacks: An HMM-based approach}, volume = {22}, year = {2025}, }
• (2025). Automated nanomanipulation for repairing defects on nanoimprint lithography masters. TASE, 22, 14967--14975. (www)

  @article{ , abstract = {The fabrication of Nanoimprint Lithography (NIL) masters serves as the initial process in the manufacturing of devices such as silicon photonic chips using NIL. Repairing defects and modifying structures on an NIL master accurately is critical for NIL manufacturing. This study proposes an innovative scanning electron microscope (SEM)-based in-situ nanomanipulation technique to address the macro-micro-nano cross-scale nanopositioning issues necessary for repair functions. A macro/micro closed-loop control system was designed, which includes a frequency/voltage (f/u) proportional controller, a real-time direct inverse hysteresis compensation feedforward controller, a grating displacement sensor, and a cross-scale nanopositioning platform. The performance testing of the nano-manipulation approach resulted in a repair range of $22.18\times 20.92 \times 10.06$ mm3, meeting the repair requirements for most silicon photo chip NIL masters in terms of size. The system achieved a repair accuracy at 4.946 nm (X-axis), 4.663 nm (Y-axis), and 4.679 nm (Z-axis). As a demonstrate, the repair functionality testing confirmed the ability of the system to remove target structures such as cantilever beams and detach adhered particles from a master surface. Note to Practitioners—Although nanomanipulation systems theoretically show significant potential for repairing defects in micro-nanostructured NIL masters, there are still numerous challenges in terms of size, actuation methods, and other aspects for nanomanipulation systems suitable for master defect repair. To overcome these challenges, an in-situ nanomanipulation system suitable for the SEM environment is proposed, along with the design of a corresponding macro/micro closed-loop control system to achieve macro-micro-nano cross-scale motion positioning, meeting the requirements for master defect repair functions. The developed nanomanipulation system can provide a new technical means for the defect repair of micro-nanostructured NIL masters, thereby promoting the development of NIL in various important fields.}, archive = {J_TASE}, author = {Siyuan Meng and Jiankang Jiang and Fubo Wang and Junhui Zhu and Qianjun Zhang and Yong Wang and Wei Dong and Changhai Ru}, doi = {10.1109/TASE.2025.3564616}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14967--14975}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated nanomanipulation for repairing defects on nanoimprint lithography masters}, volume = {22}, year = {2025}, }
• (2025). Proactive robust hardening of resilient power distribution network: Decision-dependent uncertainty modeling and fast solution strategy. TASE, 22, 14953--14966. (www)

  @article{ , abstract = {To address the power system hardening problem, traditional approaches often adopt robust optimization (RO) that considers a fixed set of concerned contingencies, regardless of the fact that hardening some components actually renders relevant contingencies impractical. In this paper, we directly adopt a dynamic uncertainty set that explicitly incorporates the impact of hardening decisions on the worst-case contingencies, which leads to a decision-dependent uncertainty (DDU) set. Then, a DDU-based robust-stochastic optimization (DDU-RSO) model is proposed to support the hardening decisions on distribution lines and distributed generators (DGs). Also, the randomness of load variations and available storage levels is considered through stochastic programming (SP) in the innermost level problem. Various corrective measures (e.g., the joint scheduling of DGs and energy storage) are included, coupling with a finite support of stochastic scenarios, for resilience enhancement. To relieve the computation burden of this new hardening formulation, an enhanced customization of parametric column-and-constraint generation (P-C&CG) algorithm is developed. By leveraging the network structural information, the enhancement strategies based on resilience importance indices are designed to improve the convergence performance. Numerical results on 33-bus and 118-bus test distribution networks have demonstrated the effectiveness of DDU-RSO aided hardening scheme. Furthermore, in comparison to existing solution methods, the enhanced P-C&CG has achieved a superior performance by reducing the solution time by a few orders of magnitude. Note to Practitioners—This paper presents a robust-stochastic optimization method for enhancing power system resilience considering endogenous and exogenous uncertainties. A resilience-based network hardening model is formulated under the trilevel RSO framework incorporating the DDU set. Compared to the traditional RO built upon the static uncertainty sets, the DDU-RSO formulation has a dynamic and stronger modeling capability, which captures the endogenous nature of contingency uncertainties associated with proactive hardening decisions. Also, the exogenous uncertainty of power loads and available storage levels is considered by introducing a scenario-based SP formulation in the innermost level problem. Note that the real shape and size of DDU set are changeable depending on the proactive decisions, which makes our problem very challenging to compute. To exactly and efficiently solve the DDU-RSO model, we design and implement a customized P-C&CG algorithm. Also, the enhancement strategy based on resilience importance indices is applied, which helps address the multi-solution dilemma widely existing in network optimization problems. It is worthy noting that the resilience importance indices can be attained not only using numerical calculation, but also through the prior knowledge of system operators. The probability information of contingency events is also helpful for determining the resilience importance indices to reduce the conservatism of RO decisions. By exploiting the “deep knowledge” contained in power systems, the customized P-C&CG algorithm has demonstrated a superior scalable capability over existing methods (e.g., the basic C&CG and a type of Benders decomposition algorithm). With this advanced analytical tool, the proposed optimization framework can be readily implemented for resilient planning and operation of large-scale real-world power systems.}, archive = {J_TASE}, author = {Donglai Ma and Xiaoyu Cao and Bo Zeng and Qing-Shan Jia and Chen Chen and Qiaozhu Zhai and Xiaohong Guan}, doi = {10.1109/TASE.2025.3564483}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14953--14966}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Proactive robust hardening of resilient power distribution network: Decision-dependent uncertainty modeling and fast solution strategy}, volume = {22}, year = {2025}, }
• (2025). Prescribed-performance green dynamic positioning for fully actuated vessels under input magnitude and rate saturations. TASE, 22, 14940--14952. (www)

  @article{ , abstract = {For the green dynamic positioning (DP) of the fully actuated vessel under input magnitude and rate saturations, unknown disturbances, and safe operational region limitations, this paper develops a prescribed-performance Lyapunov-based nonlinear model predictive control (PL-NMPC) scheme with nonlinear thrust allocation (TA). The input magnitude and rate saturations are considered as constraints in the receding horizon optimization (RHO) model. A disturbance observer is designed to provide estimates of the unknown disturbances. A prescribed performance function and an associated error transformation are introduced. Leveraging the disturbance observer and the error transformation, a robust prescribed-performance auxiliary controller is designed to establish a contractive constraint for the RHO model. With the contractive constraint, the developed PL-NMPC scheme inherit the stability of the designed auxiliary controller. Recursive feasibility and the closed-loop stability under the PL-NMPC scheme are provided. Through the theoretical analyses, the fully actuated vessel maintains its position and heading within the safe operational region with minimized energy consumption. Simulation results and comparisons demonstrate the effectiveness and the energy-saving capability of the developed PL-NMPC scheme. Note to Practitioners—In engineering practice, thrusters of DP vessels are subject to physical limitations, and constraints on vessel positions, headings, and velocities arise due to specific DP missions, safety and energy-saving concerns. Unlike the existing NMPC-based DP controllers with TA, the developed PL-NMPC scheme is applicable to vessels with azimuth thrusters, able to avoid singular thruster configurations, and is supported by comprehensive analyses of recursive feasibility and closed-loop stability. Comparative results on a fully actuated model vessel show that the PL-NMPC scheme can achieve the energy-saving rate higher than 10%. Our scheme offers significant potential for improving the energy-saving capability and ensuring safe DP operations in offshore engineering and ocean transportation. In the future, our scheme can be applied to a wide range of multi-input and multi-output systems with factors of disturbance uncertainties, energy-saving and safety concerns.}, archive = {J_TASE}, author = {Zhihao Yu and Jialu Du and Jingyao Wang and Ze Lin}, doi = {10.1109/TASE.2025.3564323}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14940--14952}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-performance green dynamic positioning for fully actuated vessels under input magnitude and rate saturations}, volume = {22}, year = {2025}, }
• (2025). EEG-based motor imagery classification with tuned heuristic fusion graph convolutional network for rehabilitation training. TASE, 22, 14928--14939. (www)

  @article{ , abstract = {Motor imagery-based brain-computer interfaces (MI-BCIs) hold significant promise for rehabilitation training in individuals with neurological impairments such as stroke and spinal cord injury (SCI). Achieving precise and robust lower limb movement prediction for each patient is crucial. However, the variability in MI response frequencies and brain activation patterns among subjects presents a great challenge to the generalizability of MI-BCIs. This paper proposes a Tuned Heuristic Fusion Graph Convolutional Network (THFGCN) for limb movement prediction in rehabilitation scenarios. THFGCN innovatively designs a learnable EEG frequency band tuned module and a heuristic space topology module. These two modules allow for the intricate extraction of both frequency and spatial topological features, utilizing graph adjacency matrices that encapsulate channel correlations and spatial relationships, hence fostering individualized analysis and enhanced generalizability across subjects. Furthermore, a spatio-temporal convolution module paired with a feature map attention mechanism is proposed to extract the critical spatio-temporal features of electroencephalogram (EEG) data. Validation experiments on the PhysioNet and LLM-BCImotion datasets against six mainstream methods demonstrate that THFGCN outperforms state-of-the-art methods, achieving 88.41% and 82.82% accuracy in the within-subject case, and 65.93% and 60.56% accuracy in the cross-subject case, respectively. Detailed frequency band weight and T-distributed Stochastic Neighbor Embedding visualization validate the effectiveness of proposed modules. Furthermore, feature interpretability analysis proves the extracted features’ profound MI task relevance, underlining THFGCN’s exceptional interpretability. Note to Practitioners—Motor imagery-based brain-computer interfaces (MI-BCIs) have shown great potential in rehabilitation training for patients with neurological disorders such as stroke and spinal cord injury (SCI), due to their capacity to promote neural plasticity and functional recovery. An accurate and generalized MI-BCI can accelerate the neural remodeling process in patients and promote the clinical application of BCI technology. In this paper, we propose a new graph convolutional network (GCN)-based MI classification framework to predict limb movements during rehabilitation training. This framework incorporates a frequency tuned topology module and a heuristic space topology module to extract private frequency-domain topological features and shared spatial topological features during MI across subjects. Our method improves the generalization ability to new subjects while ensuring accurate limb movement prediction for single subjects. We show the superiority of our method on two MI datasets, showing that the extracted features have effective physiological interpretability. These findings suggest that the proposed method is not only effective but also transparent, and may support future research in understanding neural mechanisms and optimizing rehabilitation strategies for motor-impaired populations.}, archive = {J_TASE}, author = {Kecheng Shi and Rui Huang and Jianzhi Lyu and Zhe Li and Fengjun Mu and Zhinan Peng and Chaobin Zou and Hong Cheng and Jianwei Zhang and Bijoy Kumar Ghosh}, doi = {10.1109/TASE.2025.3564162}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14928--14939}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {EEG-based motor imagery classification with tuned heuristic fusion graph convolutional network for rehabilitation training}, volume = {22}, year = {2025}, }
• (2025). Development and application of coverage control algorithms: A concise review. TASE, 22, 14906--14927. (www)

  @article{ , abstract = {Coverage control is a foundational domain within multi-agent systems, which has recently undergone substantial advancements. Model-based and optimization-based control methods have achieved new breakthroughs and applications, while data-driven and learning-based algorithms in coverage control have also yielded significant results. This review examines the evolution and application of coverage control algorithms in multi-agent systems. Moreover, this review then focuses on how contemporary algorithms build on traditional control strategies and integrate cutting-edge data-driven and adaptive learning techniques. Finally, it explores potential future directions, emphasizing the importance of interdisciplinary approaches in overcoming existing challenges and seizing new opportunities in coverage control deployment. This review aims to be a valuable resource for researchers and practitioners, guiding continued exploration in this field. Note to Practitioners—This work was motivated by the need to provide an overview of the development and research of existing multi-agent coverage control algorithms. Different application scenarios and constraints require different methods of multi-agent coverage control, each with varying efficiencies. For researchers and practitioners, deciding which coverage control method to use in specific scenarios, while considering practical constraints such as energy limitations, can be challenging. Therefore, this work offers an overview of existing multi-agent coverage control algorithms and suggests potential application scenarios. This review serves as a comprehensive reference for researchers and practitioners in the field, helping them identify suitable solutions to ensure efficient coverage.}, archive = {J_TASE}, author = {Bin Cheng and Mingyuan He and Zhongpan Zhu and Bin He and Jie Chen}, doi = {10.1109/TASE.2025.3563489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14906--14927}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development and application of coverage control algorithms: A concise review}, volume = {22}, year = {2025}, }
• (2025). Digital twin-based smart manufacturing: Dynamic line reconfiguration for disturbance handling. TASE, 22, 14892--14905. (www)

  @article{ , abstract = {The increasing complexity of modern manufacturing, coupled with demand fluctuation, supply chain uncertainties, and product customization, underscores the need for manufacturing systems that can flexibly update their configurations and swiftly adapt to disturbances. However, current research falls short in providing a holistic reconfigurable manufacturing framework that seamlessly monitors system disturbances, optimizes alternative line configurations based on machine capabilities, and automates simulation evaluation for swift adaptations. This paper presents a dynamic manufacturing line reconfiguration framework to handle disturbances that result in operation time changes. The framework incorporates a system process digital twin for monitoring disturbances and triggering reconfigurations, a capability-based ontology model capturing available agent and resource options, a configuration optimizer generating optimal line configurations, and a simulation generation program initializing simulation setups and evaluating line configurations at approximately 400x real-time speed. A case study of a battery production line has been conducted to evaluate the proposed framework. In two implemented disturbance scenarios, the framework successfully recovers system throughput with limited resources, preventing the 26% and 63% throughput drops that would have occurred without a reconfiguration plan. The reconfiguration optimizer efficiently finds optimal solutions, taking an average of 0.03 seconds to find a reconfiguration plan for a manufacturing line with 51 operations and 40 available agents across 8 agent types. Note to Practitioners—This research is motivated by the need for an advanced framework that automates the reconfiguration of manufacturing lines in response to increasing disturbances within dynamic production environments. Existing approaches lack a holistic framework that automates disturbance monitoring, data management, line structure optimization, and pre-implementation simulation verification. This paper integrates all these components and develops an innovative configuration optimizer and simulator to improve the optimality and efficiency of disturbance handling. We demonstrate the applicability of our framework through a case study in a battery production line, focusing on agent delay issues. The approach is adaptable to a variety of disturbance scenarios in other types of manufacturing lines that share common elements such as workers, robots, machines, and auxiliary tools. Future investigations will expand on reconfiguration considerations, including spatial constraints, agent experience levels, and tool change requirements.}, archive = {J_TASE}, author = {Bo Fu and Mingjie Bi and Shota Umeda and Takahiro Nakano and Youichi Nonaka and Quan Zhou and Takaharu Matsui and Dawn M. Tilbury and Kira Barton}, doi = {10.1109/TASE.2025.3563320}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14892--14905}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Digital twin-based smart manufacturing: Dynamic line reconfiguration for disturbance handling}, volume = {22}, year = {2025}, }
• (2025). Nussbaum-based fixed-time tracking control for uncertain nonlinear systems driven by piezoelectric actuators. TASE, 22, 14880--14891. (www)

  @article{ , abstract = {The piezoelectric actuator will exhibit the problem of unknown control direction under high-speed operating conditions. Hence, the realization of high-speed control for nonlinear systems driven by piezoelectric actuators requires the fixed-time stability to enhance the convergence rate and the Nussbaum technique to handle the problem of unknown control directions. However, these two techniques are incompatible because the required conditional inequalities for each are different. To resolve this incompatibility, we have incorporated the projection operator into the design of the adaptive updating law. On the basis of ensuring that the adaptive estimation parameter is bounded, we have established the conditional inequality for Nussbaum technique. Then, some nonlinear damping terms are incorporated into the stability analysis to counteract a term containing Nussbaum function, thereby reducing the difficulty of the fixed-time stability analysis. Furthermore, the predefined tracking accuracy and fixed-time stability are theoretically proven through the stability analysis. Finally, the efficiency of the proposed control method is validated through semi-physical experiments on a cylindrical piezoelectric actuator. Note to Practitioners—This study is motivated by the control objectives of achieving high performance for uncertain nonlinear systems driven by piezoelectric actuators, with a focus on fast settling time and high tracking precision. Notably, some studies have reported that an increase in the frequency of the excitation input signal for a piezoelectric actuator can cause a reversal of control direction. However, existing control schemes, which assume the control direction is unchanged or known, may fail. This work proposes a novel control scheme that employs the Nussbaum technique to address the challenge of unknown control direction, and overcomes the technical incompatibilities between Nussbaum technique and fixed-time control by introducing a robust adaptive control law. The feasibility of the designed control approach is confirmed through semi-physical experiments.}, archive = {J_TASE}, author = {Guanyu Lai and Kai Huang and Zhi Liu and Hanzhen Xiao and C. L. Philip Chen}, doi = {10.1109/TASE.2025.3562598}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14880--14891}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nussbaum-based fixed-time tracking control for uncertain nonlinear systems driven by piezoelectric actuators}, volume = {22}, year = {2025}, }
• (2025). Robust cross-chamber one-class fault detection in semiconductor manufacturing. TASE, 22, 14867--14879. (www)

  @article{ , abstract = {Fault detection (FD) is essential for wafer quality control in semiconductor manufacturing (SM), as it can identify abnormal wafers in the early stages. However, frequent chamber discrepancies leads to distribution shifts in the data from different chambers, which causes performance degradation of the existing model. In this paper, we conceive a robust cross-chamber fault detection method in SM, which to the best of our knowledge is the first work that employs domain generalization to address the issue of cross-chamber fault detection in SM. Our basic idea is to map the samples from the source chambers and target chambers to the same hypersphere space, making normal samples cluster around a specific feature center while faulty samples stay away from it. Due to the scarcity of faulty samples, we propose a one-class classification-based fault detection method relying on normal samples to establish classification boundaries. To generalize the model to unseen target chambers, we design a meta-learning-based one-class domain generalization approach. We also devise a strategy to enhance distribution alignment within hypersphere, making the classification boundaries of various chambers be close to each other. The evaluations on real-world data collected from a wet process equipment in SM verify the high robustness of the model to chamber discrepancies with limited faulty samples. Note to Practitioners—The main challenges of fault detection in semiconductor manufacturing include the lack of faulty wafer samples and the frequent data distribution shifts across chambers. Traditional transfer learning-based fault detection methods, such as domain adaptation and domain generalization, often fail to effectively address the aforementioned challenges simultaneously. The key work in this paper is achieving robust cross-chamber fault detection with scarce faulty samples. We design a one-class classification model to learn classification boundaries from normal samples and propose a one-class domain generalization technique that achieves few-shot learning for fault samples through gradient alignment. We also introduce a new batch normalization technique to further align the classification boundaries. Engineers can apply the method to perform a intelligent fault detection on cross-chamber process tools for health management in semiconductor manufacturing.}, archive = {J_TASE}, author = {Qirui Bai and Shuangwu Chen and Huihuang Qin and Dong Jin and Xiaobin Tan and Huasen He and Guohao Wang and Jian Yang}, doi = {10.1109/TASE.2025.3562244}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14867--14879}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust cross-chamber one-class fault detection in semiconductor manufacturing}, volume = {22}, year = {2025}, }
• (2025). Bearing-based collision-free formation control for spacecrafts under dynamic event-triggered input. TASE, 22, 14855--14866. (www)

  @article{ , abstract = {This paper investigates a motion planning problem for multiple spacecrafts under dynamic event-triggered input and bearing constraint. As a safety-critical system, collision avoidance plays a key role in endowing the formation system with fully autonomous operating space under obstacle-cluttered environment. Given this fact, a control barrier function (CBF)-based quadratic programming (QP) is developed based on backstepping framework, such that each spacecraft is able to move to its predefined position without collisions among static/moving obstacles and its neighbors. Considering the limited communication and actuation capability, an input-based event-triggered mechanism (ETM) is then introduced with dynamic parameter to schedule the triggering instants. By doing so, the data transmission from the controller to the actuator can be well-regulated with system input only, which offers a more practical solution to limited communication resources. The stability of the system is proved via Lyapunov analysis and the simulations are conducted to show the effectiveness of the proposed strategy. Note to Practitioners—This paper was motivated by the problem of collision-free control for spacecraft formation. Practically, existing formation are commonly localized based on real-time position measurements under global framework. We introduced a bearing-based approach into the formation control of spacecrafts, which enables more natural self-localization using vision measurements. Under bearing-based framework, the collision-avoidance problem is then considered. In both initial stage and steady stage, the safety problem is nonnegligible. We developed the existing CBF-based algorithm to generate a collision-free path for each spacecraft. To minimize redundant actions of actuator, an ETM was incorporated to help improve the control efficiency. In simulation part, we will show how these methods work collaboratively in given tasks.}, archive = {J_TASE}, author = {Jun Li and Honglong Kang and Jianfeng Li and Shenmin Song}, doi = {10.1109/TASE.2025.3564742}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14855--14866}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bearing-based collision-free formation control for spacecrafts under dynamic event-triggered input}, volume = {22}, year = {2025}, }
• (2025). Modeling and control design for a musculoskeletal robot via adaptive dynamic programming. TASE, 22, 14843--14854. (www)

  @article{ , abstract = {In this study, for the musculoskeletal robot system, we aim at optimizing the angle tracking control based on the system model. First, we analyze the driving principle of muscles and establish a bionic muscle dynamics model. Further, based on the geometric relationship between muscles and skeletons, we establish the kinematics model and dynamics model of the musculoskeletal robot system. The adaptive dynamic programming (ADP) algorithm is used for solving the Hamilton-Jacobi-Bellman (HJB) equation. Based on the proposed updating law, the neural network is trained to approximate the solution. Using the Lyapunov’s direct method, we can prove that the system is stable with the proposed controller. Furthermore, simulation indicates that this presented control law is feasible. Note to Practitioners—The motivation of this study is to investigate the dynamic modeling and optimal control design of musculoskeletal robots. Existing research on musculoskeletal robots has mostly focused on the mechanical design, with less consideration given to the analysis of system dynamics. Typically, model-free control algorithms are employed, which affect the precise control performance of the system. This paper aims to develop a mathematical representation of musculoskeletal robots based on the biological and mechanical characteristics of muscles and bones. The proposed method can optimize the tracking performance of the system, avoiding repetitive trial and error. Aseries of simulation experiments have been conducted to validate the effectiveness and applicability of this method.}, archive = {J_TASE}, author = {Yuhua Song and Weiying Wan and Lifeng Zhu and Aiguo Song}, doi = {10.1109/TASE.2025.3564943}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14843--14854}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling and control design for a musculoskeletal robot via adaptive dynamic programming}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant adaptive synchronization for high-order discrete-time multiagent systems with stochastic noises. TASE, 22, 14833--14842. (www)

  @article{ , abstract = {This paper addresses fault-tolerant adaptive synchronization for a category of high-order discrete-time multi-agent systems (MASs) influenced by stochastic noise. Built upon a spanning directed acyclic graph (SDAG), a backstepping-based framework is introduced for implementing controllers. To be specific, under the proposed framework, all unknown information are accumulated at last step and only one neural network (NN) is utilized to implement the controller design for each agent. As a consequence, the corresponding intermediate NN approximations for virtual controllers are released. Furthermore, circular dependencies are avoided with the assistance of SDAG, ensuring the feasibility of the controllers in multiagent setting. By utilizing the Lyapunov difference approach, all signals in the MASs are guaranteed to be exponential mean-square (EMS) bounded. Ultimately, simulation studies involving five pendulum systems are performed to demonstrate the effectiveness of the proposed strategy. Note to Practitioners—In contemporary applications, MASs are increasingly prone to experiencing various faults, arising from both complex operational environments and intricate interactions among agents. Such faults, if not promptly identified and addressed, can severely compromise the performance and reliability of MAS-based applications. Additionally, MASs are typically characterized by inherent nonlinearity and operate under digital control, which imposes the significance to develop nonlinear discrete-time control strategies. Consequently, this article proposes novel adaptive NN control strategies for a type of high-order nonlinear MASs to achieve output synchronization. Furthermore, actuator faults and stochastic noises are considered. The proposed fault-tolerant control strategies integrate adaptive NN approximations and a compensation term for fault mitigation, ensuring that the synchronization errors converge to a neighborhood of the origin through proper parameter adaptation.}, archive = {J_TASE}, author = {Wenqi Xu and Xiaoping Liu and Tong Wang and Xiaokun Liu}, doi = {10.1109/TASE.2025.3562514}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14833--14842}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant adaptive synchronization for high-order discrete-time multiagent systems with stochastic noises}, volume = {22}, year = {2025}, }
• (2025). Synergistic constrained control of 6-DOF fixed-wing multi-UAVs with dynamic self-triggered communication. TASE, 22, 14818--14832. (www)

  @article{ , abstract = {A coordinated control challenge is addressed in 6-degree-of-freedom (6-DOF) fixed-wing multiple autonomous aerial vehicle (multi-AAV) systems under communication and state constraints. The primary obstacle in achieving this goal arises from managing frequent information interactions and the assurance that UAV states converge within prescribed bounds. On the one hand, a novel dynamic self-triggering mechanism is effectively proposed. Unlike current state-of-the-art approaches, the proposed dynamic self-triggering communication mechanism features a larger triggering threshold and eliminates the need for continuous monitoring of system state information. This reduces the demand on system communication and sensor resources. On the other hand, a new time-varying constraint bounded function is introduced to effectively relax restrictions on the initial system state. Then, the coordinated translational/rotational controllers are designed to ensure minimal consensus tracking error. Semi-physical simulations highlight the effectiveness of the proposed control algorithm. Note to Practitioners—In actual environment, the multi-UAVs flight always requires inter-communication to ensure the stable performance of the entire formation. However, period-based communication leads to a waste of communication resources. The event-triggering communication mechanism lowers the communication frequency of UAVs, thereby reducing energy consumption. Nevertheless, most existing control results on event-triggered communication overlook the fact that continuous monitoring of state information still causes unnecessary energy consumption. To further investigate the problem, a dynamic self-triggering mechanism is proposed in this study, which can determine the subsequent triggered moment based on the state information of the current triggered moment. In addition, the state of UAVs due to safety and physical constraints ought to be constrained. Therefore, a prescribed-time constrained control strategy is proposed, which not only improves the transient performance (e.g. small overshoot and fast adjustment time), but also ensures that the UAV state converges within a given constraint bound.}, archive = {J_TASE}, author = {Yuyuan Shi and Jing Li and Maolong Lv and Ning Wang and Yuan Yuan and Jing Chang}, doi = {10.1109/TASE.2025.3545933}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14818--14832}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synergistic constrained control of 6-DOF fixed-wing multi-UAVs with dynamic self-triggered communication}, volume = {22}, year = {2025}, }
• (2025). Trajectory-attracted adaptive tracking control for robotic systems based on a hybrid guiding vector field in flexible environments. TASE, 22, 14808--14817. (www)

  @article{ , abstract = {This paper proposes a trajectory-attracted adaptive tracking control (TAATC) scheme based on a hybrid guiding vector field (HGVF) for robotic systems, addressing the high operational difficulty and safety concerns inherent in flexible environments. The HGVF is constructed using the characteristics of different task spaces in flexible environments to enable smooth transitions between free space and contact space with adjustable operating velocity, thereby enhancing robustness against environmental interactions and uncertainties. The HGVF can attract the state trajectories of the robotic systems by path convergence of an auxiliary dynamic system, simplifying the trajectory planning process with a time-independent representation of the desired path. In addition, a neural network is employed to compensate for uncertainties in the robotic systems, while a nonlinear duffing function represents the dynamic contact force model of flexible environments. In this way, the TAATC scheme is constructed by using the HGVF and the adaptive neural network, which unifies trajectory planning and tracking control. By using the proposed TAATC scheme, the robotic systems can achieve smooth interaction with flexible environments and obtain the desired tracking force without complex trajectory planning. The stability of the HGVF and the whole control scheme are analyzed by using the Lyapunov theorem. Finally, the effectiveness of the proposed method is validated through both simulation and experimental tests. Note to Practitioners—The motivation of this article is that operational tasks in flexible environments are the most common for robotic systems, where the deformation and fragile nature exist widely in actual applications. The risks of overshooting, oscillations, and other harmful effects during the task execution, especially when transitioning from non-contact to contact tasks, usually degrade the control performances or even damage the flexible objects. In this article, a HGVF-based trajectory-attracted adaptive tracking control scheme is proposed for robotic systems with uncertainty operating in flexible environments. The theoretical derivation demonstrates that the proposed control scheme can attract the robotic system to track a desired trajectory and interaction with flexible environments safely, where the smooth transitioning process is achieved by using the HVGF-based time-independent control method to unify the trajectory planning and tracking control. The satisfactory control performances of the simulation and experimental tests on a robotic manipulator, which interacts with a flexible object, show the practical application on a real robotic system.}, archive = {J_TASE}, author = {Penghui Fan and Jinzhu Peng and Shuai Ding and Yaqiang Liu and Yaoyu Yang and Nan Zhao and Yaonan Wang}, doi = {10.1109/TASE.2025.3564644}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14808--14817}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory-attracted adaptive tracking control for robotic systems based on a hybrid guiding vector field in flexible environments}, volume = {22}, year = {2025}, }
• (2025). Adaptive NN observer-based synthesize strategy for connected nonlinear system. TASE, 22, 14798--14807. (www)

  @article{ , abstract = {In this paper, the issues of tracking and synchronization for connected nonlinear servo system is considered. An adaptive neural network (NN) observer-based synthesize strategy is proposed. Firstly, a mathematical model of connected nonlinear isomorphism multi-motor servo system with disturbance is established, where the motors are connected through wired or wireless networks. Then, an adaptive disturbance observer based on sliding-mode is proposed, and the finite time convergence of observation errors has been proven. Thirdly, an adaptive NN tracking strategy is designed, and we have proved the tracking error is bounded and the full-state asymmetric constraints are satisfied. Furthermore, a synchronous control technique on sliding-mode is presented, and both the boundedness of synchronous error and reachability of sliding surface are verified. Finally, a numerical simulation and a semi-physical simulation are carried out to illustrate the validity of proposed methods. Note to Practitioners—This paper was motivated by the tracking and synchronization of connected nonlinear servo system, which encounters with the adverse effect of disturbance. We present a novel adaptive NN observer-based synthesize strategy to achieve the desired control performance. First of all, we established the dynamic of connected nonlinear isomorphism multi-motor servo system under disturbance. In order to obtain prior knowledge of disturbances, we design an adaptive sliding-mode disturbance observer and the observation errors can converge to a small domain within finite time. To ensure that all motors can track the desired trajectory without violating the full state constraints, we propose an adaptive NN tracking scheme based on coordinate transformation and the boundedness of the tracking error is demonstrated. Additionally, to achieve synchronization of all motors, we present a sliding-mode synchronous control approach, and the reachability of sliding surface and the boundedness of synchronous error are proved, simultaneously. Both simulation and experiment indicate that designed algorithms are effective and feasible.}, archive = {J_TASE}, author = {Meng Li and Yong Chen and Meng Zhang and Haiyu Song}, doi = {10.1109/TASE.2025.3564331}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14798--14807}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive NN observer-based synthesize strategy for connected nonlinear system}, volume = {22}, year = {2025}, }
• (2025). Proportional-difference observer-based fault estimation design for discrete-time nonlinear singular systems with disturbances. TASE, 22, 14788--14797. (www)

  @article{ , abstract = {This article presents a robust fault reconstruction scheme for a class of discrete-time nonlinear singular systems with disturbances. The Lipschitz condition is utilized to describe the nonlinearity under investigation. To effectively estimate multiple types of faults such as time-varying faults and constant ones, a brand-new proportional-difference (PD) observer design method is established. What is worth mentioning is that all the gains of PD observer can be calculated at the same time by our method instead of manually selecting some gains and then solving other gains as has been done in the pertinent literature, which serves as a pioneering attempt for the similar design. Besides, by using the Lyapunov theory, less conservative linear matrix inequalities (LMIs) criteria with slack scalars and matrices are obtained to ensure that error systems are robust stability with the prescribed $H_{\infty }$ performance. Finally, two examples on RC circuits and tanks level control models are presented to prove the superiority and practicability of the proposed fault estimation procedure. Note to Practitioners—This article is motivated to realize the PD observer-based fault estimation design of nonlinear singular systems in the discrete-time domain even if there are external disturbances and faults in the system model. The practical significance of this article is mainly twofold. First, since aerospace engineering, power systems, electrical circuit networks and industrial process systems, etc. can be modeled by singular systems, to get closer to reality, it is of great importance to research such systems especially in the presence of nonlinearities and disturbances. Second, the increasing complexity of modern control systems has led to the prevalence of faults in all engineering environments. What is more, the faults often occur in many forms and not just in the form of constant faults in practice. In this case, the achievement of fault estimation for multiple types of faults makes our method applicable in actual engineering, especially in the application of active fault-tolerant control.}, archive = {J_TASE}, author = {Yunfei Mu and Qiancheng Wang and Dongsheng Yang and Huaguang Zhang}, doi = {10.1109/TASE.2025.3564067}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14788--14797}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Proportional-difference observer-based fault estimation design for discrete-time nonlinear singular systems with disturbances}, volume = {22}, year = {2025}, }
• (2025). Attention with system entropy for optimizing credit assignment in cooperative multi-agent reinforcement learning. TASE, 22, 14775--14787. (www)

  @article{ , abstract = {In cooperative multi-agent reinforcement learning (MARL), value function factorization methods have been proposed to address the dimensionality explosion problem encountered in centralized training and decentralized execution (CTDE). The existing value function factorization methods lack a perspective from the global system when addressing credit assignment, failing to measure each agent’s contribution to the current system. A typical limitation is that these factorization approaches primarily rely on the proximity of local individual groups to assign credit. There are challenges such as the sensitivity of credit assignment weights to local information and instability in the learning process. In this study, we focus on redistributing agents’ features during the system evolution and suggest employing an enhanced attention mechanism with system entropy measure to factorize value function. Specifically, this method emphasizes each agent’s representation between their local and global contributions and then redesigns multi-head attention to optimize the credit assignment in value function factorization. To evaluate the effectiveness of our method, we conduct a series of cooperative multi-agent tasks on the StarCraft II platform and compare the results with several state-of-the-art techniques, including Qatten, QMIX, QTRAN, COMA, and VDN. The experimental results demonstrate that our method achieves faster overall convergence speed, higher stability, and robust performance across various scenarios. Note to Practitioners—The fully cooperative multi-agent reinforcement learning (MARL) is extensively applied in practical multi-agent systems, such as coordinating multiple vehicle traffic conditions. Motivated by these practical application scenarios, our study is to enhance MARL’s learning efficiency by introducing an entropy-centered improved multi-head attention mechanism to address credit assignment issues. This method is aimed at measuring the importance weights of each agent within the agent system, thereby further boosting the learning efficiency of MARL. This work holds promise for application in large-scale intelligent agent systems, including multi-robot systems and collaborative multi-vehicle systems, among others.}, archive = {J_TASE}, author = {Wei Wei and Haibin Li and Shiyuan Zhou and Baifeng Li and Xue Liu}, doi = {10.1109/TASE.2025.3563725}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14775--14787}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attention with system entropy for optimizing credit assignment in cooperative multi-agent reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Online credibility assessment of equipment digital twin for discrete manufacturing. TASE, 22, 14763--14774. (www)

  @article{ , abstract = {The equipment digital twins (EDTs) for discrete manufacturing should be calibrated quickly to avoid irreversible physical damage to the equipment caused by biased control commands. Therefore, an online credibility assessment method for EDTs is urgently needed. However, existing assessment approaches consume too much time, and thus could not reveal dynamic faults in time. In this paper, the dynamic relationship between online evolution and actual applications of EDTs is investigated. Then, two steps are proposed to accelerate the assessment process significantly. One involves pre-constructing a performance-deviation-agent (PDA), and the other involves dynamically fitting the application-time-window (ATW) probability distribution. The methodology is applicable to discrete manufacturing processes. The dynamic credibility of EDT evolution process can be updated after every iteration of the model evolution. Sorting manufacturing equipment was used as a case study to demonstrate the effectiveness of this method. The time consumption was reduced by 90% compared with traditional assessment methods in the case. Note to Practitioners—The application of digital twins (DTs) for discrete manufacturing equipment typically aims to enhance the control effectiveness of the physical equipment. To ensure the effectiveness of control signals derived from DT-based simulations, the credibility of the equipment DTs must be known and guaranteed. This paper addresses the quick updating of the credibility of equipment DTs as they evolve according to real-time data collected from the equipment. The proposed methods are currently applicable to individual pieces of equipment with straightforward workflows. By adapting the training method for the error prediction model to scenarios involving multiple pieces of equipment and more complex workflows, the overall control efficacy within a workshop can be significantly improved.}, archive = {J_TASE}, author = {Han Lu and Lin Zhang and M. Jamal Deen and Kunyu Wang and Hongbo Cheng and Laurence T. Yang}, doi = {10.1109/TASE.2025.3563246}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14763--14774}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online credibility assessment of equipment digital twin for discrete manufacturing}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered adaptive asymptotic tracking control of uncertain systems with unified prescribed performance. TASE, 22, 14752--14762. (www)

  @article{ , abstract = {This paper presents a dynamic event-triggered unified performance adaptive control method for nonparametric strict-feedback nonlinear systems. In contrast to most existing static/dynamic event-driven controllers that only guarantee uniformly ultimately bounded tracking results, here by introducing an external auxiliary variable into the dynamic threshold strategy and using the robust technique with the integral function, not only the communication overhead is reduced and the Zeno phenomenon is precluded, but also the asymptotic zero-error tracking is achieved. The control design and stability analysis become quite complicated and challenging when the performance constraint is taken into account. By constructing a series of functional transformations in conjunction with the core information technique to handle the nonparametric uncertainty, the proposed controller is able to guarantee multiple prescribed performance characteristics by appropriately adjusting the key design parameter, eliminating the need to redesign the controller and reanalyze the stability. Finally, simulation results are conducted to demonstrate the effectiveness of the theoretical discussion. Note to Practitioners—The study of dynamic event-triggered control is an important topic for practical networked systems such as autonomous aerial vehicles, boiler turbine systems, chemical processes and robotic systems. If the impact of the event-triggered mechanism on the controller design is not considered, the system performance may deteriorate significantly. Although numerous works on this topic can be found in the literature, there are still some limitations in the design and implementation. In this paper, we propose a unified prescribed performance control method for some classical practical systems via the dynamic event trigger framework, which can effectively clarify the relationship between system performance and transmission resource consumption. By introducing an external auxiliary variable into the dynamic threshold strategy and using the integral function, not only the communication overhead is saved in the practical networked systems, but also the asymptotic zero error tracking is achieved. Moreover, by constructing a series of function transformations and using the core information technique, the proposed controller is able to guarantee multiple prescribed performance behaviors under a fixed control framework, eliminating the need to redesign the controller and reanalyze the stability, which is more in line with actual requirements.}, archive = {J_TASE}, author = {Kai Zhao and Yuhang Huang and Yongcheng Zhou}, doi = {10.1109/TASE.2025.3563063}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14752--14762}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered adaptive asymptotic tracking control of uncertain systems with unified prescribed performance}, volume = {22}, year = {2025}, }
• (2025). GNN topology representation learning for deformable multi-linear objects dual-arm robotic manipulation. TASE, 22, 14738--14751. (www)

  @article{ , abstract = {Deformable Multi-Linear Objects (DMLOs), or Branched Deformable Linear Objects (BDLOs), are flexible objects that possess a linear structure similar to DLOs but also feature branching or bifurcation points where the object’s path diverges into multiple sections. The representation of complex DMLOs, such as wiring harnesses, poses significant challenges in various applications, including robotic systems’ perception and manipulation planning. This paper proposes an approach to address the robust and efficient estimation of a topological representation for DMLOs leveraging a graph-based description of the scene obtained via graph neural networks. Starting from a binary mask of the scene, graph nodes are sampled along the objects’ estimated centerlines. Then, a data-driven pipeline is employed to learn the assignment of graph edges between nodes and to characterize the node’s type based on their local topology and orientation. Finally, by utilizing the learned information, a solver combines the predictions and generates a coherent representation of the objects in the scene. The approach is experimentally evaluated using a test set of complex real-world DMLOs. Within an offline evaluation, the proposed approach achieves a Dice score exceeding 90% in predicting graph edges. Similarly, the identification accuracy of branch and intersection points in the graph topology is above 90%. Additionally, the method demonstrates efficient performance, achieving a runtime of over 20 FPS. In an online assessment employing a dual-arm robotic setup, the approach is successfully applied to disentangle three automotive wiring harnesses, demonstrating the effectiveness of the proposed approach in a real-world scenario. Note to Practitioners—This research is motivated by the lack of effective robotic perception and manipulation solutions for deformable multi-linear objects such as wiring harnesses. These objects play a crucial role in several manufacturing industries. However, their intricate branching structures, inherent deformable nature, and limited distinctive visual characteristics present formidable challenges when automating tasks that involve them. The proposed approach addresses several of these challenges: 1) It maintains robustness in the presence of noise and disruptions in the input image mask due to its graph-based structure; 2) It remains unaffected by the existence of large connectors, clips, fixtures, and other objects along the branch sections of the object; 3) It avoids making any assumptions about the structure of the object; 4) It is capable of simultaneously identifying branch-points and resolving intersection-points among different branch sections of these objects. The experiments have verified the effectiveness of this solution when applied to real-world deformable multi-linear objects found in automotive settings. This underscores the potential to integrate this solution into industrial scenarios that involve the assembly and routing of such objects.}, archive = {J_TASE}, author = {Alessio Caporali and Kevin Galassi and Riccardo Zanella and Gianluca Palli}, doi = {10.1109/TASE.2025.3562231}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14738--14751}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GNN topology representation learning for deformable multi-linear objects dual-arm robotic manipulation}, volume = {22}, year = {2025}, }
• (2025). Only pick once: Algorithms for efficiently picking an exact number of multiple identical objects. TASE, 22, 14723--14737. (www)

  @article{ , abstract = {Picking up multiple objects at once is a grasping skill that makes a human worker efficient in many domains. This work tackles the problem of getting a requested number of identical objects in a shallow bin by only pick once (OPO) using a simple parallel gripper. The proposed system contains several graph-based algorithms that convert the layout of objects into a graph, cluster vertices in the graph, rank and select candidate clusters based on their topology. Our algorithm also has a multi-object picking predictor based on a convolutional neural network for estimating how many objects would be picked up with a given gripper location and orientation. This paper presents four evaluation metrics and three protocols to evaluate the proposed system. The results show that our proposed system has very high success rates for two and three objects when only picking once. Utilizing our approach can significantly outperform single object picking two to three times in terms of efficiency. The results also show our algorithm can be applied to one unseen shape (hexagon) and unseen sizes cube and cylinder during training without fine-tuning to achieve decent accuracy. Note to Practitioners—This paper is motivated by the current bottleneck in robotics picking which restricts the mass deployment of robots in tasks that are similar to batch-picking in logistics. The state of the art (SOTA) picking algorithm is fast but only focuses on picking one object at a time, which inspires us to look into picking multiple objects at once to increase the efficiency. We design a custom-sized parallel-jaw gripper based on the object sizes’ need, and we train a deep-learning model for each designed gripper to predict the number of objects that a grasp pose can retrieve. Our simulation and real-life evaluations have very good results in picking 2 to 3 objects at once in a random setting. Our future work will consider more complex scenarios and rearrangements that can break some hard settings.}, archive = {J_TASE}, author = {Zihe Ye and Ricardo Frumento and Yu Sun}, doi = {10.1109/TASE.2025.3560956}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14723--14737}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Only pick once: Algorithms for efficiently picking an exact number of multiple identical objects}, volume = {22}, year = {2025}, }
• (2025). A hierarchical multi-task and multi-agent assignment approach: Learning DQN strategy from execution. TASE, 22, 14712--14722. (www)

  @article{ , abstract = {This article investigates the problem of real-time task assignment with heterogeneous agents while considering resource constraints. A hierarchical reinforcement learning-based(HRL) method to address the problem has been proposed. Unlike most existing studies, the method is to assign the tasks to agents such that the resource consumption is minimized while respecting the resource constraints and realizing task distribution balance among agents. At the high level, the real-time task assignment problem within a heterogeneous multi-agent system is formulated as a sequence optimization model which considers multiple constraints. At the low level, a novel reinforcement learning-based hierarchical framework that decomposes the large-scale task assignment problem into the target assignment layer and resource assignment layer is developed. In the first layer, a task assignment strategy closer to the optimal one in real-time by reducing the dimension of the DQN state-action space and learning from execution is produced. In the second layer, a novel resource assignment method is designed to minimize resource consumption and reserve as many agents as possible to handle new tasks by optimizing the resource distribution among agents. Simulation experiments in multi-UAV collaborative emergency material delivery demonstrate that the proposed method can generate high-quality solutions for various problem scales and greatly improve resource balance and real-time performance. Note to Practitioners—This work is motivated by the task assignment problem involving multiple heterogeneous agents in a highly real-time scenario, while taking into account various practical constraints. Each agent can carry various types and quantities of resources within its capacity, and the objective is to minimize resource consumption and travel distance for agents while reserving as many agents as possible to handle new tasks. The proposed approach can be utilized for securing the transportation of valuable assets, pesticide spraying using uncrewed aerial vehicles, and environmental monitoring. The emergency material delivery is used as an illustrative example to demonstrate how UAVs can ensure the rapid transportation of materials and safeguard the lives of affected individuals. To resolve the computationally challenging problem, this paper proposes a HRL method to efficiently solve the task assignment problem among multiple heterogeneous agents, which effectively enhances computational efficiency and improves task completion by optimizing the distribution of resources among agents.}, archive = {J_TASE}, author = {Yu Wang and Huiping Li and Qingliang Shen}, doi = {10.1109/TASE.2025.3560121}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14712--14722}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical multi-task and multi-agent assignment approach: Learning DQN strategy from execution}, volume = {22}, year = {2025}, }
• (2025). Attack-defense game of heterogeneous multi-agent systems under actuator faults. TASE, 22, 14699--14711. (www)

  @article{ , abstract = {This paper investigates N versus N attack-defense game problem of heterogeneous nonlinear multi-agent systems under the condition of actuator faults on the defensive side. Specifically, N attackers need to approach the target area as closely as possible in order to attack it, while N defenders need to approach the attackers as closely as possible and intercept them. We decompose this problem into two N-player nonzero-sum game problems. Based on differential game method, the optimal attacking game strategies for the attackers are first presented. Then, considering that the defenders contains actuator faults, this paper designs a fault observer for each defending agent based on adaptive estimation theory to estimate unknown actuator fault and proposes a novel optimal defense game strategy applicable for actuator fault. To address the problem that it is difficult to directly solve the optimal control strategies in nonlinear systems, we propose an online learning algorithm based on adaptive dynamic programming to obtain approximate solutions for the optimal cost functions and the optimal control strategies. The effectiveness of the proposed method is verified through theoretical analysis and simulation examples. Note to Practitioners—The problem of attack-defense confrontation for multi-agent systems has a broad practical background, such as resource competition, missile interception, and base defense, etc. By introducing game theory, complex attack-defense confrontation problems for multi-agent systems can be clearly formulated. However, most studies on attack-defense game confrontation are still limited to a small number of intelligent agents. Additionally, actuator faults with frequent and great harm pose significant challenges to the safe and stable operation of multi-agent systems in complex and changeable practical applications. Especially in the attack-defense confrontation scenario, actuator faults may be caused not only by the system’s own physical faults but also by the malicious attacks from attackers. Based on the above considerations, this paper studies the N versus N attack-defense game problem of heterogeneous nonlinear multi-agent systems when the defenders have actuator faults. The proposed control scheme not only ensures the smooth implementation of interception tasks for the defenders with actuator faults, but also significantly saves time and control input energy compared with the traditional formation tracking control method without actuator faults. It is hoped that this paper can provide some inspiration for the attack-defense confrontation problem of multi-agent systems under actuator faults.}, archive = {J_TASE}, author = {Yadong Li and Bin Hu and Tao Li and Zhi-Hong Guan}, doi = {10.1109/TASE.2025.3559704}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14699--14711}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attack-defense game of heterogeneous multi-agent systems under actuator faults}, volume = {22}, year = {2025}, }
• (2025). Learning-based fault-tolerant control with high-order control barrier functions. TASE, 22, 14689--14698. (www)

  @article{ , abstract = {Fault-tolerant control can maintain acceptable performance in faulty systems but may violate safety constraints, posing potential risks. Additionally, model uncertainties present further challenges. In this paper, we propose a novel fault-tolerant probabilistic safe control framework that integrates Gaussian process (GP) regression, fault-tolerant control, and the high-order control barrier function (HOCBF) method. To handle inherent uncertainties, we adopt a probabilistic method that combines control Lyapunov function (CLF) with HOCBF using GP regression. Building on this foundation, we design a fault-tolerant GP-based CLF-HOCBF method to address unknown actuator faults. Furthermore, we establish two theoretical criteria to ensure the probabilistic safety and stability of the proposed control framework. To validate our method, we implement it in the autonomous driving simulator CARLA, demonstrating its effectiveness and competitiveness compared to existing approaches. Note to Practitioners—Control systems should maintain stability and performance in the presence of model uncertainties, especially by ensuring during faults until appropriate engineering measures are implemented. However, existing methods on addressing uncertainties will influenced by observation noise and imperfect measurements, and current fault-tolerant control strategies struggle to handle safety-critical control problems. Given this, this paper suggests a new approach using GP regression and HOCBF, considering the cases of unknown system dynamics and actuator bias faults. Accordingly, sufficient conditions are established for guaranteeing the probabilistic safety and stability. Our proposed method faces challenges when addressing unknown gain dynamics or unknown actuator gain faults. In future research, we will design an affine dot product compound kernel method to overcome these limitations.}, archive = {J_TASE}, author = {Linhao Zhao and Lu Wang and Yuting Cao and Yin Yang and Shiping Wen}, doi = {10.1109/TASE.2025.3563189}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14689--14698}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based fault-tolerant control with high-order control barrier functions}, volume = {22}, year = {2025}, }
• (2025). High-performance robust synchronous control of dual-drive gantry system via composite adaptive B-spline wavelet neural network. TASE, 22, 14679--14688. (www)

  @article{ , abstract = {Complex uncertain dynamics and external disturbances have always been critical factors restricting the performance improvement of dual-drive gantry system (DDGS), especially the synchronization and robustness. To enhance the performance of DDGS in uncertain industrial environments, in this paper, a high-performance robust control method is proposed based on a novel composite adaptive B-spline wavelet neural network (CABWNN) estimator. Specifically, a composite adaptive robust synchronization control method with desired compensation is proposed based on a rotational dynamic coupling model of the system. This method combines the benefits of both direct and indirect adaptive robust control, achieving superior control performance while simultaneously mitigating the impact of measurement noise in the system. Additionally, a novel CABWNN estimator with a specially designed composite adaptive weight updating law is introduced to approximate the system uncertainties, which enhances the approximation accuracy. Under the proposed control method, the boundedness of all signals in the closed-loop has been proved in theory. Furthermore, comparative experiments on a dual-drive gantry system validate the superiority of the proposed control strategy. Note to Practitioners—Achieving synchronization control of dual-drive gantry systems under complex conditions is a key challenge in industrial production. This paper proposes a high-performance robust synchronization control method based on a rigid-flexible coupling model. First, a composite adaptive robust control with desired compensation enhances parameter uncertainty compensation while reducing measurement noise effects. To handle complex operating conditions, a novel CABWNN estimator improves estimation of unmodeled dynamics and disturbances. The proposed controller ensures high-precision trajectory tracking and synchronization, overcoming traditional methods’ trade-off between tracking and beam rotation, thereby enhancing system performance for industrial applications.}, archive = {J_TASE}, author = {Yuntong Wang and Yanbin Liu and Weichao Sun}, doi = {10.1109/TASE.2025.3562139}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14679--14688}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-performance robust synchronous control of dual-drive gantry system via composite adaptive B-spline wavelet neural network}, volume = {22}, year = {2025}, }
• (2025). Congruent-transformation-based stability criterion of T-S fuzzy systems with time-varying delay via the generalized line-integral lyapunov function. TASE, 22, 14669--14678. (www)

  @article{ , abstract = {This paper is concerned with the stability analysis of Takagi-Sugeno (T-S) fuzzy systems with time-varying delays. First, by constructing a generalized line-integral Lyapunov function (GLILF), applying the free-matrix-based inequality and the delay-derivative-dependent free-weighting-matrices method, a basic stability criterion is derived. Besides, the structure of the matrices in the stability criterion is reconstructed through congruent transformation, and an improved stability criterion is derived accordingly. To clarify the advancement of the proposed method, a stability condition is given without the GLILF. Finally, two numerical examples and a practical example are applied to show the advancement and merits of the presented method. Note to Practitioners—This study proposes a new stability-related method for time-delay fuzzy systems, which can broaden the application scope of existing theories. Firstly, the new method provides a more flexible choice for modeling existing nonlinear time-delay systems such as truck trailer systems, power systems, multi-agent systems, etc. as fuzzy systems, which can more effectively represent the original system using fuzzy models. Secondly, the proposed method reduces the conservatism of stability analysis and design and improves the time delay margin that can be tolerated under the premise of system stability.}, archive = {J_TASE}, author = {Zhou-Zhou Liu and Li Jin and Yong He}, doi = {10.1109/TASE.2025.3558917}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14669--14678}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Congruent-transformation-based stability criterion of T-S fuzzy systems with time-varying delay via the generalized line-integral lyapunov function}, volume = {22}, year = {2025}, }
• (2025). Practical fixed-time consensus of discontinuous multi-agent systems: Adaptive intermittent control with time-varying gains approaches. TASE, 22, 14657--14668. (www)

  @article{ , abstract = {In this paper, the fixed-time (FxT) and practical FxT consensus of discontinuous multi-agent systems (MASs) via adaptive intermittent control strategy are studied. New intermittent FxT stability lemmas incorporating the indefinite function and the unified exponent condition are established, which can include the existing periodically intermittent stability lemmas and aperiodic intermittent lemmas with negative definite derivatives. Considering that the MASs cannot realize the exact consensus under the inevitable external interferences, intermittent-type practical FxT stability lemmas are further obtained. The intermittent controllers and adaptive intermittent controllers are designed, which demonstrates the improvements over the existing control strategies since the chattering phenomenon and singularity do not appear. Notably, the control gains are time-varying, which are also different from the previous time-invariant ones. Based on the established stability lemmas, the FxT consensus and practical FxT consensus are achieved. Clearly, the practical FxT consensus of discontinuous MASs under the discontinuous-time controllers is achieved for the first time. The simulation confirms the applicability of the proposed methods and shows strong support for our theoretical findings. Note to Practitioners—Many studies in multi-agent systems (MASs) rely heavily on continuous information exchange among agents, which can lead to significant communication overhead. In practical applications, however, such continuous communication may be inefficient or even infeasible due to network constraints, high energy consumption, or intermittent connectivity. To overcome these challenges, this paper presents a novel adaptive intermittent control strategy aimed at achieving both fixed-time (FxT) and practical FxT consensus. The key innovation of this strategy lies in the development of adaptive intermittent controllers that dynamically adjust the feedback gain based on system states and external disturbances. This method reduces communication load while ensuring fast and reliable consensus convergence. Moreover, new intermittent stability lemmas are introduced, extending existing stability results by incorporating indefinite functions and unified exponential conditions. These lemmas facilitate the achievement of multi-agent consensus within a fixed-time or practical-time framework. By utilizing these new stability results, the adaptive intermittent control strategy effectively addresses challenges posed by discontinuous dynamics in MASs, such as chattering and singularities, thereby ensuring desired consensus behavior. The proposed strategy and lemmas represent significant advancements over traditional consensus algorithms, with promising applications in systems where communication efficiency and robustness are crucial. Future research will focus on expanding the applicability of this approach to larger and more complex systems.}, archive = {J_TASE}, author = {Honglin Ni and Chaoxu Mu and Fanchao Kong and Song Zhu}, doi = {10.1109/TASE.2025.3563055}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14657--14668}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical fixed-time consensus of discontinuous multi-agent systems: Adaptive intermittent control with time-varying gains approaches}, volume = {22}, year = {2025}, }
• (2025). Quasi-multisynchronization and quasi-monosynchronization of delayed neural networks with parameter mismatch via impulsive control. TASE, 22, 14646--14656. (www)

  @article{ , abstract = {This article focus on the quasi-multisynchronization and quasi-monosynchronization of delayed neural networks (DNNs) with parameter mismatch via impulsive control. At first, a kind of actication functions (AFs) is proposed and multistability/monostability criteria of DNNs are given. By judging $\boldsymbol {2n}$ algebraic inequalities, the $\boldsymbol {n}$ -neuron DNNs with this kind of AFs can produce $\boldsymbol {2^{n}}$ locally stable equilibrium points (EPs) or one globally stable EP. In addition, a kind of impulsive controller is designed. Impulsive control strategy proposed in this paper can save the bandwidth and reduce the communication cost compared with continuous-time control strategy. Moreover, the concept of quasi-multisynchronization of DNNs is proposed for the first time and sufficient conditions are given to ensure quasi-multisynchronization and quasi-monosynchronization of DNNs by building the comparison system and using the Lagrange method of variation of parameters. Under this control method, a monostable system can become a multistable system, which means that the storage capacity (SC) of DNNs is improved. Lastly, two examples are illustrated to testify the validity of the proposed theory and method. Note to Practitioners—Impulsive control strategy can save the bandwidth and reduce the communication cost compared with conventional continuous-time control. In addition, in order to seek out a new way to increase the SC of NNs, combining quasi-synchronization and multisynchronization, this paper studies the quasi-multisynchronization and quasi-monosynchronization of DNNs with parameter mismatch via impulsive control. A kind of ETI controller is designed and the concept of quasi-multisynchronization of DNNs is proposed for the first time. Except for destroying the monotonicity or continuity of the AFs, and designing the state-dependent switching NNs, another method to improve the SC of NNs is found. That is, by achieving quasi-multi-synchronization, the monostable NNs can become the multistable NNs, which can produce multiple locally stable ESs.}, archive = {J_TASE}, author = {Zhen Wang and Yang Liu and Xia Huang and Hao Shen}, doi = {10.1109/TASE.2025.3562995}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14646--14656}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quasi-multisynchronization and quasi-monosynchronization of delayed neural networks with parameter mismatch via impulsive control}, volume = {22}, year = {2025}, }
• (2025). Generalized critical observability of labeled petri nets under nondeterministic observations. TASE, 22, 14633--14645. (www)

  @article{ , abstract = {This paper addresses the property of critical observability in labeled Petri nets. A system is deemed critically observable if its state estimate (formed by observing the system output) can be determined to belong to a set of either critical markings or non-critical markings. More generally, for a critically observable system, there may be multiple types of critical markings and the operator may desire to know the type to which the current state estimate belongs. We propose a generalized version of critical observability, namely generalized critical observability, and provide two necessary and sufficient conditions to check it based on basis markings and integer linear programs, effectively circumventing the need for an exhaustive enumeration of all reachable markings of the net. In addition, we investigate (generalized) critical observability of Petri nets under nondeterministic observations, which may arise due to the loss of transmission signals or due to interference with sensor measurement outputs. To handle this, an augmented system is constructed from the original Petri net by adding new transitions that capture the effect of transitions associated with nondeterministic labels. Thus, we transform the verification of (generalized) critical observability of a Petri net under nondeterministic observations to that of the augmented model. Two real-world examples illustrate the applicability of the presented notion and methodology. Note to Practitioners—In most real safety-critical applications, such as train control systems, air traffic management systems, and electric power networks, it is crucial to determine whether the system state estimate falls within a set of dangerous or faulty states (or, more broadly, states of interest). Critical observability guarantees that the current state estimate, i.e., the set of states consistent with observations, is either a subset of critical states or a subset of non-critical states. In practical scenarios, there may be multiple types of faulty states, and operators need to accurately identify the fault type when a system operates abnormally, i.e., they need to determine the type to which the current state estimate belongs, in order to make correct decisions. Therefore, a generalized version of critical observability (namely, generalized critical observability) is proposed, which allows for the determination of the specific type of critical states. Furthermore, to handle unreliable sensor readings (e.g., due to influences from external factors, such as extreme environments), we investigate (generalized) critical observability for labeled Petri nets (LPN) under nondeterministic observations. The two proposed basis reachability graph-based approaches are provably computationally efficient when verifying generalized critical observability of LPNs, which facilitates engineers to apply the methods developed in this research to real-world systems. Two examples, one involving a manufacturing system and the other involving an airport system, demonstrate the proposed approach.}, archive = {J_TASE}, author = {Yuling Zhang and Christoforos N. Hadjicostis and Zhiwu Li}, doi = {10.1109/TASE.2025.3560435}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14633--14645}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Generalized critical observability of labeled petri nets under nondeterministic observations}, volume = {22}, year = {2025}, }
• (2025). Finite-time consensus of second-order multiagent systems with input saturation via hybrid sliding-mode control. TASE, 22, 14623--14632. (www)

  @article{ , abstract = {This paper addresses the finite-time consensus (FTC) issue for second-order multi-agent systems (MASs) with nonlinear disturbances. To tackle the challenges posed by increasingly complex communication environments, an innovative integral sliding-mode surface is designed using event-triggered control. Furthermore, the study explores communication constraints between the leader and followers, employing impulsive control to facilitate communication only at specific intervals. A novel hybrid integral sliding-mode control protocol is advanced for the second-order MASs, which effectively eliminates the “Zeno phenomenon” and notably reduces both communication frequency and energy consumption, thereby enhancing overall communication efficiency. Ultimately, the efficacy of the put forward protocol is verified through simulation examples including comparison experiment. Note to Practitioners—The motivation of this paper lies in the fact that the problem of limited control resources and communication bandwidth is prevalent in the practical applications of MASs. Failing to address these limitations is likely to result in performance deterioration. To further reduce communication costs and achieve greater energy savings, a hybrid sliding mode control protocol is put forward for nonlinear second-order MASs under disturbances in this paper. Theoretical deduction shows that the proposed control protocol can achieve leader-following FTC. Additionally, simulation experiments validate its superiority, both of which indicate that it can greatly reduce the number of communications in applications with certain practical significance.}, archive = {J_TASE}, author = {Xiaowei Jiang and Ranran Jiao and Bo Li and Xianhe Zhang and Huaicheng Yan}, doi = {10.1109/TASE.2025.3559947}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14623--14632}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time consensus of second-order multiagent systems with input saturation via hybrid sliding-mode control}, volume = {22}, year = {2025}, }
• (2025). IIIM-SAM: Zero-shot texture anomaly detection without external prompts. TASE, 22, 14610--14622. (www)

  @article{ , abstract = {Anomaly detection is a crucial aspect in ensuring the reliability of industrial products in smart manufacturing. Despite employing unsupervised anomaly detection methods to mitigate the challenges associated with obtaining defect data, there remains the challenge of adapting anomaly detectors to drift within normal data distributions, especially when there are significant intra-class variations in normal samples. Consequently, zero-shot anomaly detection techniques, which do not require training data and thus avoid interference from training data, have emerged as a new research direction. Meanwhile, the zero-shot segmentation capability demonstrated by the Segment Anything Mode (SAM) visual foundation model has further fueled our research enthusiasm. In this paper, we focus on texture images and propose a no-external prompt SAM-based method for zero-shot texture anomaly detection, called Image Internal Information Mining SAM (IIIM-SAM). We utilize the Image Internal Information Mining Prompter to mine relationships between internal regions of the image, identifying specific background and potential defect points. The two types of prompt points are fed into the SAM model in different stages, imparting the SAM with incorporating category information segmentation capability. Unlike current methods, our zero-shot texture anomaly detection method requires no external prompts or expert knowledge, and also avoids the uncertainty introduced by text prompts necessary for CLIP-based approaches. Our experimental results demonstrate the superiority of our zero-shot texture anomaly detection method compared to other approaches. Specifically, in the texture subset of MVTecAD/KolektorSDD2, our IIIM-SAM achieves 99.2%/93.6% image-level AUROC and 98.6%/92.3% pixel-level AUROC. Note to Practitioners—The motivation of this paper arises from a practical challenge in anomaly detection in smart flexible manufacturing—customized customer demands lead to significant intra-class variations and even entirely different types of products. Currently, most anomaly detection methods struggle to adapt quickly to this challenge, severely impeding production efficiency. This paper proposes a zero-shot texture anomaly detection method based on SAM, which automatically generates background and potential defect points through an image internal information mining prompter. Compared to existing zero-shot anomaly detection algorithms based on foundation models, our method’s most significant uniqueness is without any expert knowledge and the absence of any external text prompts. Through comparisons with four advanced methods on six instances using two evaluation metrics, the results demonstrate the outstanding anomaly detection capability of our approach.}, archive = {J_TASE}, author = {Zhe Zhang and Yuhang Zhou and Jiahe Yue and Runchu Zhang and Jie Ma}, doi = {10.1109/TASE.2025.3561776}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14610--14622}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {IIIM-SAM: Zero-shot texture anomaly detection without external prompts}, volume = {22}, year = {2025}, }
• (2025). Context-aware sim-to-real unsupervised domain adaptation for lane detection via disentangled feature alignment. TASE, 22, 14593--14609. (www)

  @article{ , abstract = {While existing supervised deep learning-based lane detection methods achieve exceptional detection performance, constructing a large real-world dataset with labels is a cost-intensive task. Therefore, we propose a novel sim-to-real unsupervised domain adaptation method specialized in lane detection. In this paper, we present a disentangled feature alignment approach that performs selective adaptation for lane and background features. By performing the disentangled prototype-based local and global feature alignments, we solve the negative transfer problem of an existing adversarial feature alignment-based domain adaptation for lane detection. In addition, the lane detection task requires accurate localization of the lanes even in occluded parts. Therefore, we adopt a consistency regularization strategy using masked images to improve the detection accuracy in the occluded area. Additionally, we introduce an adaptive resolution adjustment of the masked-out patch based on training maturity for context learning. By applying a simple framework, we achieve superior lane detection accuracy with lower false positives and false negatives in the target domain compared to conventional methods. Extensive experiments demonstrate the superiority of the proposed method. Note to Practitioners—Lane detection plays a major role in autonomous vehicles by predicting the location of lanes on a road. In particular, deep learning-based lane detection models aim to output robust detection results in various environments, including different weather and road types. Because deep learning models are highly dependent on the distribution of training datasets, models trained with datasets acquired in a given environment can cause significant performance degradation when tested in different environments. Therefore, improving the generalizability of the lane detection model in the real-world application requires a vast training dataset. However, it is impractical to collect and label the real-world datasets including multiple road environments. We propose a novel training strategy for the lane detection models trained with simulation datasets to achieve high detection performance even on the real-world dataset. In other words, we introduce knowledge transfer methods for enhancing stability of the models trained with simulation datasets on real roads. In addition, we expect that the proposed method will transform the manual process of collecting and labeling real-world datasets into automatic creation of simulation-based learning datasets.}, archive = {J_TASE}, author = {Yeon Jeong Chae and Ji Sun Byun and Yun Hak Lee and Jae Yun Lee and Sang Hoon Han and Joo Hyeon Jeon and Sung In Cho}, doi = {10.1109/TASE.2025.3560458}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14593--14609}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Context-aware sim-to-real unsupervised domain adaptation for lane detection via disentangled feature alignment}, volume = {22}, year = {2025}, }
• (2025). AI-driven landing zone detection module for vertical take-off and landing vehicles using projection-based LiDAR-navigation pipelines. TASE, 22, 14577--14592. (www)

  @article{ , abstract = {This paper introduces an artificial intelligence-based landing zone detection module (LZDM) for vertical take-off and landing (VTOL) navigation. It employs a projection-based point cloud semantic segmentation (PCSS) convolutional neural network model combined with point cloud accumulation and a range image generation module. The proposed method addresses the limitations of existing projection-based PCSS methods, which often struggle with low-resolution and non-repetitive scan raw light detection and ranging (LiDAR) data commonly found in aerial datasets. The proposed LZDM was developed using three sets of aerial datasets collected from a DJI M600 hexacopter drone, a DJI M300 RTK quadrotor, and a Bell412 helicopter. The results were evaluated using both qualitative and quantitative metrics, demonstrating its robustness and effectiveness. In terms of quantitative results, the proposed method achieved mean intersection over union and accuracy values greater than 0.93 and 98 percent, respectively, across all three datasets, highlighting its accuracy in identifying safe landing zones (LZs). To assess the real-time feasibility of the proposed LZDM, it was deployed on a reconfigurable hardware-accelerated module. This setup achieved processing rates higher than 10 Hz for all three datasets and a throughput of over 5 million pts/s on the Jetson AGX Xavier dedicated hardware combined with the PyTorch TensorRT optimization module. The supplementary materials, including the inference code, sample testing data, and instructions are available in our GitHub repository at https://github.com/nira16herath/CENet-LZ-detection/tree/main Note to Practitioners—VTOLs, such as small drones and helicopters, are increasingly becoming popular in time-critical and safety-critical operations, including parcel delivery to remote areas, search and rescue, and other autonomous missions. While advanced trajectory planning systems are available to guide these vehicles to designated (or targeted) positions, the system also demands reliable and accurate landing of the vehicle at nominal and emergency situations. Therefore, ensuring reliable and rapid LZ detection to improve safe autonomous landing and expedite decision-making remains a significant challenge, particularly in unknown and unprepared terrains. Autonomous systems must identify safe LZs with precision and should avoid landing on unsafe areas, such as water bodies, inclined surfaces, or other unstable surfaces. Furthermore, VTOL-based aerial datasets typically generate low-resolution and non-repetitive scan LiDAR point clouds, making accurate LZ detection more challenging. The work proposed in the paper provides a fast, accurate, and semantic-aware LiDAR-based LZ detection that provides a low risk of misidentification, which ensures safe landings in complex environments. The proposed method has been trained and validated using real-world custom-labeled aerial datasets and deployed on reconfigurable hardware, ensuring real-time suitability for LZ detection in diverse environments. Since the method relies solely on LiDAR sensors, practitioners can implement the proposed method without additional sensor modalities or high-resolution LiDAR point clouds of environments.}, archive = {J_TASE}, author = {Nirasha Herath and Oscar De Silva and George K. I. Mann and Awantha Jayasiri}, doi = {10.1109/TASE.2025.3563822}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14577--14592}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AI-driven landing zone detection module for vertical take-off and landing vehicles using projection-based LiDAR-navigation pipelines}, volume = {22}, year = {2025}, }
• (2025). Efficient trotting of soft robotic quadrupeds. TASE, 22, 14562--14576. (www)

  @article{ , abstract = {Soft robots hold significant potential in legged locomotion due to their inherent deformability, enabling enhanced adaptability to various environmental conditions and the generation of diverse locomotion gaits. While various soft robots have been proposed for terrestrial locomotion, research on dynamically-stable locomotion, such as trotting, with actuated soft bending limbs remains limited. We introduce a pneumatically-actuated soft quadruped featuring a soft body capable of a variety of dynamically-stable trotting locomotion. We utilize soft limb kinematics and parameterize fundamental limb locomotion to obtain quadrupedal locomotion trajectories for both linear and curvilinear motions. We also employ a physics-enabled dynamic model to optimize and evaluate trotting locomotion trajectories for dynamic stability. We further validate the stable locomotion trajectories through empirical experiments conducted on a soft quadruped prototype. The results demonstrate that the quadruped trots at a peak speed of 1.24 body lengths per second when traversing flat and uneven terrains, including slopes, cluttered areas, and naturalistic irregular surfaces. Furthermore, we compare the energy efficiency between trotting and crawling locomotion. The findings reveal that trotting is significantly more energy-efficient than crawling, with an average energy saving of up to 42%.Note to Practitioners—This paper was motivated by the challenge of achieving dynamically stable and efficient locomotion in soft quadrupeds. Many soft-legged robots are typically designed for statically stable, albeit inefficient and slow, locomotion gaits such as crawling. Our research aims to address this practical challenge of improving mobility in soft-legged robots. We develop a novel soft quadruped with pneumatically-actuated soft limbs that achieves efficient trotting that is 42% more energy-efficient than crawling. This work is particularly relevant for industries requiring adaptable and efficient navigation in environments, such as search and rescue, agricultural monitoring, and exploration. The development and optimization of trotting gaits through a physics-enabled dynamic model for dynamic stability provide a foundational framework for enhancing the adaptability and operational utility of soft robots. While our findings mark a significant step forward, challenges remain in deploying these locomotion strategies on autonomous untethered robots with onboard sensor feedback. Future research will focus on these areas, aiming to improve the practical deployment and robustness of soft robotic locomotive systems.}, archive = {J_TASE}, author = {Dimuthu D. K. Arachchige and Tim Sheehan and Dulanjana M. Perera and Sanjaya Mallikarachchi and Umer Huzaifa and Iyad Kanj and Isuru S. Godage}, doi = {10.1109/TASE.2025.3553082}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14562--14576}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient trotting of soft robotic quadrupeds}, volume = {22}, year = {2025}, }
• (2025). Dynamic scheduling with task migration in cloud manufacturing using hybrid federated deep reinforcement learning and graph neural network model. TASE, 22, 14547--14561. (www)

  @article{ , abstract = {Cloud Manufacturing (CMfg) serves as a pivotal platform, seamlessly integrating enterprise resources and consumer demands, thus playing a central role in task scheduling and service allocation. However, the dynamic nature of cloud environments, especially during service interruptions, necessitates systems capable of promptly responding to real-time changes and demands, presenting formidable challenges for task migration. Traditional methods often struggle to adapt swiftly, leading to suboptimal real-time decision-making. To address these challenges, we introduce the DRL-GNN-FL SynergyNet model, a hybrid solution that prioritizes both time efficiency and cost-effectiveness. This model leverages advanced technologies, including Deep Reinforcement Learning (DRL), Graph Neural Networks (GNN), and Federated Learning (FL). DRL enables rapid adaptability and real-time decision-making, GNN effectively processes complex network structures, and FL facilitates distributed learning and knowledge sharing. This comprehensive approach effectively tackles the intricacies of task migration in CMfg environments. Experimental results demonstrate the model’s exceptional performance, adaptability, and scalability in real-time scenarios, showcasing its potential as a practical and effective solution to the dynamic task migration challenges encountered in CMfg. The DRL-GNN-FL SynergyNet model offers a promising avenue for enhancing the resilience and responsiveness of cloud manufacturing systems. Note to Practitioners—To optimize production planning in the domain of smart manufacturing systems, it is imperative to streamline the utilization of manufacturing resources effectively. This paper centers its focus on addressing the challenges associated with task migration within the context of dynamic scheduling in cloud manufacturing, presenting a robust solution by introducing Deep Reinforcement Learning (DRL), Graph Neural Networks (GNN), and Federated Learning (FL). Conventional dynamic scheduling methods often struggle to swiftly adapt to real-time fluctuations and demands while frequently falling short in formulating effective task migration strategies. To overcome these limitations, this paper introduces an innovative DRL-GNN-FL SynergyNet model. This model amalgamates the rapid adaptability and real-time decision-making capabilities offered by DRL, the adept handling of complex network structures by GNN, and the benefits of distributed learning and knowledge sharing inherent to FL. Through comprehensive simulation experiments conducted in our study, we conclusively demonstrate that the proposed approach surpasses existing scheduling techniques in terms of operational efficiency and the proficiency to make well-informed decisions regarding task migration. Moreover, our approach readily lends itself to practical implementation within real production environments. Looking forward, future research endeavors could further refine this methodology by service broken associated with point-to-point and hub-and-spoke networks across diverse transportation routes.}, archive = {J_TASE}, author = {Wenxin Li and Lingyan Li and Sicheng Liu and Xiao Song and Lin Zhang}, doi = {10.1109/TASE.2025.3562381}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14547--14561}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic scheduling with task migration in cloud manufacturing using hybrid federated deep reinforcement learning and graph neural network model}, volume = {22}, year = {2025}, }
• (2025). Distributed cauchy-kernel-based maximum correntropy filter in interconnected multi-area power system. TASE, 22, 14538--14546. (www)

  @article{ , abstract = {This paper introduces a novel distributed Cauchy-kernel-based maximum correntropy filter designed to address the state estimation problem in multi-area power systems under non-Gaussian noise conditions. In the framework of load frequency control (LFC), the proposed filter uses local and neighboring sensor data, leveraging the coupling characteristics of multi-area power systems. A method for determining the coupling estimation gain is provided. Unlike traditional methods that use minimum mean square error, this filter employs the Cauchy-kernel-based maximum correntropy condition to evaluate estimation performance. A fixed-point iterative algorithm is also presented to enhance the estimation performance in non-Gaussian environments, and the convergence of the algorithm is proved by Banach’s fixed point theory. Furthermore, the filter’s performance is compared with that of the Kalman filter, with an analysis of estimation unbiasedness. Finally, we validate the proposed distributed filter by simulation using a ten-area power system. Note to Practitioners—In the multi-area interconnected power system, the LFC needs to monitor state variables such as generator frequency and tie-line power in real time. However, the amplitude of impulse noise caused by the on-off behavior of renewable energy sources in practical systems is significantly higher than that of conventional Gaussian noise. Power systems using the traditional minimum mean square error estimation methods would produce substantial errors. Therefore, this paper presents a Cauchy-kernel-based maximum correntropy distributed estimation method, which can effectively address the state estimation problem in multi-area power systems under non-Gaussian noise environments.}, archive = {J_TASE}, author = {Shaojie Xu and Dan Ye and Dongsheng Yang}, doi = {10.1109/TASE.2025.3563488}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14538--14546}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed cauchy-kernel-based maximum correntropy filter in interconnected multi-area power system}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time practical consensus of nonlinear multi-agent systems subject to DoS attacks via event-triggered mechanism. TASE, 22, 14525--14537. (www)

  @article{ , abstract = {This article studies the prescribed-time practical consensus problem for nonlinear multi-agent systems (MASs) subject to denial-of-service (DoS) attacks via an intermittent-based event-triggered mechanism. To this end, we first propose a time-varying scaling function based on the intermittent period, and then use this function to derive a new practical prescribed-time stability theory. Second, this paper further designs an intermittent-based event-triggered control protocol, along with a corresponding triggered function. The main innovation of this protocol is its ability to avoid singular phenomena as time approaches the prescribed instant and to reduce communication costs among all agents. Additionally, some sufficient conditions for achieving leader-following practical consensus are deduced by the proposed prescribed-time stability theory while also excluding Zeno behavior. Finally, two examples are given to verify the effectiveness and feasibility of the theoretical results. Note to Practitioners—This paper is motivated by security consensus challenges in cyber-physical systems, with a particular focus on cost efficiency and convergence performance. Existing prescribed-time security consensus solutions often overlooked the singularity of prescribed instant. Our work addresses a critical issue in MASs vulnerable to DoS attacks. The proposed control strategy is especially relevant for systems susceptible to network attacks, such as microgrids, UAVs, and submarine systems. Preliminary experiments have demonstrated the feasibility of this approach. Future research will tackle the complete consensus problem for MASs within the prescribed-time.}, archive = {J_TASE}, author = {Xuening Xu and Zhiyong Yu and Haijun Jiang and Xuehui Mei}, doi = {10.1109/TASE.2025.3561548}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14525--14537}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time practical consensus of nonlinear multi-agent systems subject to DoS attacks via event-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Energy consumption modeling and process parameter optimization of internal gear power honing under multi-axis coupling. TASE, 22, 14510--14524. (www)

  @article{ , abstract = {The internal gear power honing process is widely used in gear machining for electric vehicles because of the advantages of high-precision and high-efficiency machining. The gear honing process involves six axes and high spindle speeds, this process contributes a substantial amount of energy consumption but has less attention on energy saving. To improve the energy efficiency of gear honing process, this paper proposes an energy consumption modeling and process parameter optimization method under multi-axis coupling. The multi-axis coupling motion and energy consumption characteristics of gear honing process are analyzed. The energy consumption model of gear honing process under multi-axis coupling is then established, and the influence law of process parameters on honing energy consumption is investigated. Furthermore, a multi-objective process parameter optimization model for minimizing energy consumption and machining time is constructed. An improved multi-objective atomic orbital search (IMOAOS) algorithm is developed to solve the multi-objective optimization problem. The gear honing experiment results demonstrate that the optimized scheme reduces energy consumption by 39.75% and machining time by 8.48% compared to the empirical scheme. The proposed multi-objective optimization scheme also significantly balances the energy consumption and machining time of gear honing process compared with single-objective optimization. Note to Practitioners—The internal gear power honing process is increasingly used in the electric vehicle and aerospace gear fields due to the high-precision machining. This process consumes a lot of energy but has fewer energy-saving concerns. This article analyzes the energy consumption characteristics of gear honing process, and establishes its energy consumption model for the first time. A multi-objective honing process parameter optimization model considering energy consumption and machining time is then constructed. The experimental results show that the energy consumption and time of the optimization scheme are significantly reduced compared with the empirical scheme. The proposed modeling and optimization method provides a honing parameter decision-making scheme for practical operators.}, archive = {J_TASE}, author = {You Zhang and Congbo Li and Ying Tang and Huajun Cao and Guibao Tao}, doi = {10.1109/TASE.2025.3560253}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14510--14524}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy consumption modeling and process parameter optimization of internal gear power honing under multi-axis coupling}, volume = {22}, year = {2025}, }
• (2025). Human–Robot intrinsic skill transfer and programming by demonstration system. TASE, 22, 14498--14509. (www)

  @article{ , abstract = {Robots can often learn skills from human demonstrations. Robot operations via visual perception are commonly influenced by the external environment and are relatively demanding in terms of external conditions, manual segmentation of tasks is time-consuming and labor-intensive, and robots do not perform complex tasks with sufficient accuracy and naturalness in their movements. In this work, we propose a programming by demonstration framework to facilitate autonomous task segmentation and flexible execution. We acquire surface electromyography (sEMG) signals from the forearm and train the gesture datasets by transfer learning from the sign language dataset to achieve action classification. Meanwhile, the inertial information of the forearm is collected and combined with the sEMG signals to autonomously segment operational skills into discrete task units, and after comparing with ground truth, it is demonstrated that multi-modal information can lead to higher segmentation accuracy. To make the robot movements more natural, we add arm stiffness information to this system and estimate the arm stiffness of different individuals by creating a muscle force map of the demonstrator. Finally, human manipulation skills are mapped onto the UR5e robot to validate the results of human-robot skill transfer. Note to Practitioners—The motivation of this work is to record the characteristics of human operations and transfer them to robots for more flexible and safer robot programming, with eventual application in structured industrial scenarios. Differences between our study and others include 1) The use of a portable multi-source signal acquisition device that operates in an uncoupled fashion. 2) Consider the demonstration programming process from the external demonstration information and the inner force information. 3) The results obtained by different people are the same, indicating that the method has good generalization.}, archive = {J_TASE}, author = {Fei Wang and Jinxiu Wu and Siyi Lian and Yi Guo and Kaiyin Hu}, doi = {10.1109/TASE.2025.3559661}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14498--14509}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human–Robot intrinsic skill transfer and programming by demonstration system}, volume = {22}, year = {2025}, }
• (2025). A hierarchical progressive perception system for autonomous luggage trolley collection. TASE, 22, 14484--14497. (www)

  @article{ , abstract = {Advancements in intelligent vehicle technologies and autonomous driving are now used for tasks like collecting luggage trolleys at airports. The robotic autonomous luggage trolley collection system employs robots to gather and transport scattered luggage trolleys. However, existing methods for detecting and locating luggage trolleys often fail when they are not fully visible. To address this, we introduce the hierarchical progressive perception system, which enhances the detection and localization of luggage trolleys under partial occlusion. The proposed system integrates the hierarchical process structure and progressive perception strategy. This innovative structure processes the luggage trolley’s position and orientation separately. It can accurately determine the luggage trolley’s position with just one well-detected keypoint and estimate its orientation when partially occluded. Once the luggage trolley’s initial pose is detected, the progressive perception strategy continuously refines this information until the robot begins grasping. The proposed system only needs RGB images for labeling and training, eliminating the need for complex data collection and annotation. The experiments on detection and localization demonstrate that the proposed system is more reliable under partial occlusion compared to existing methods. Its effectiveness and robustness have also been confirmed through practical tests in actual luggage trolley collection tasks. A website about this work is available at https://sites.google.com/view/robot-perception. Note to Practitioners—The motivation of this article is to detect and locate the luggage trolleys even when they are partially occluded. The effective perception of luggage trolleys is the basis for autonomous luggage trolley collection tasks, but the existing methods cannot work well under occlusion. Then, we propose a new perception system that integrates the hierarchical process structure and progressive perception strategy. The innovative hierarchical structure separately identifies the position and orientation of the luggage trolley. The progressive perception continuously updates the luggage trolley’s pose during the navigation. The proposed system works well even when the luggage trolleys are partially occluded, ensuring they are detected and located reliably. Additionally, this system only needs RGB images for labeling and training, enhancing the scalability and practical implementation. The effectiveness and practicality of the proposed system have promoted the implementation of the robotic autonomous luggage trolley collection system at airports.}, archive = {J_TASE}, author = {Zhirui Sun and Zhe Zhang and Jieting Zhao and Hanjing Ye and Jiankun Wang}, doi = {10.1109/TASE.2025.3562621}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14484--14497}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical progressive perception system for autonomous luggage trolley collection}, volume = {22}, year = {2025}, }
• (2025). Dissipative analysis for multi-area power systems in multi-delay environments and PI-type load frequency controller design via LMIs method. TASE, 22, 14474--14483. (www)

  @article{ , abstract = {This article focuses on the dissipative stability issue with load frequency control (LFC) in multiple time-varying delay (MTD) power systems. Firstly, the model is constructed considering the effects of MTD. Next, Lyapunov-Krasovskii functionals (LKF) are constructed which include delay-product-type terms (DPT) with MTD. Then, the low conservative stability criteria with respect to linear matrix inequalities (LMIs) are derived through the application of auxiliary function-based integral inequalities (AFBII) and improved reciprocal convex inequalities (IRCI) to handle the DPT. Next, an effective dissipative control strategy is developed based on the above criteria. Finally, simulation experiments are carried out, and the effectiveness of the proposed design method is highlighted by comparing with the previous results. Note to Practitioners—This paper focuses on the dissipative stability of LFC in power systems with MTD. It provides a new perspective for the control and optimization of power systems and other dynamic systems involving MTD. Specifically, when constructing LKF with DPT, this paper designs for different time-varying delay conditions, which can better deal with the uncertainty caused by time delay changes and help to improve the accuracy and reliability of system stability analysis. Then, the new stability criterion proposed in this paper provides more relaxed conditions, which reduces the conservatism of system stability analysis and makes the stability criterion more practical application value. Based on the new stability criterion, a dissipative controller design method is presented in this paper. This method can provide better control effect when dealing with complex multi-delay systems and help to improve the overall operating efficiency and stability of the power system. Finally, the effectiveness of the scheme is verified by simulating the multi-region power system.}, archive = {J_TASE}, author = {Wenqing Zhu and Jun Wang and Kaibo Shi and Xiao Cai and Ming Zhu}, doi = {10.1109/TASE.2025.3555722}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14474--14483}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dissipative analysis for multi-area power systems in multi-delay environments and PI-type load frequency controller design via LMIs method}, volume = {22}, year = {2025}, }
• (2025). Aperiodic data-driven model predictive control with feasibility and stability guarantees. TASE, 22, 14461--14473. (www)

  @article{ , abstract = {This paper proposes an event-triggered data-driven model predictive control method with terminal ingredients for unknown linear time-invariant (LTI) systems under the input constraints. Specifically, the proposed method utilizes an implicit model description based on behavioral systems theory as well as input-output measurements, without requiring prior system identification steps. We explicitly design terminal ingredients: the terminal data-driven controller, terminal weight matrix and terminal region. To reduce computational costs, we develop an event-triggered scheme to activate the optimization problem only when necessary, rather than periodically. Moreover, a terminal data-driven controller is employed in the terminal region to avoid solving the optimization problem, further reducing computational requirements. We prove that the proposed method guarantees recursive feasibility and closed-loop stability under sufficiently small channel noise. Finally, the effectiveness of our method is verified by simulations. Note to Practitioners—Model predictive control (MPC) is a widely-used advanced control methodology that relies on explicit system models. This requirement limits its applicability to unknown systems, such as unknown LTI systems. Data-driven MPC offers a promising alternative by bypassing the need for explicit knowledge of system models, and it relies solely on initial measurement data and online input-output measurements to make control decisions. The existing data-driven MPC methods solve the optimization problem in a periodic manner, resulting in significant computational burden. Event-triggered scheme can alleviate the computational burden by reducing the frequency of solving the optimization problem. However, designing such a scheme typically requires knowledge of the system model, which poses a challenge for unknown systems. To address this challenge, this paper develops an event-triggered data-driven MPC approach for unknown LTI systems under input constraints and channel noise. Compared with the existing data-driven MPC approaches, the proposed event-triggered data-driven MPC approach significantly reduces the computational burden while ensuring recursive feasibility and closed-loop stability.}, archive = {J_TASE}, author = {Pengbiao Wang and Xuemei Ren and Dongdong Zheng}, doi = {10.1109/TASE.2025.3559986}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14461--14473}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Aperiodic data-driven model predictive control with feasibility and stability guarantees}, volume = {22}, year = {2025}, }
• (2025). SSFold: Learning to fold arbitrary crumpled cloth using graph dynamics from human demonstration. TASE, 22, 14448--14460. (www)

  @article{ , abstract = {Robotic cloth manipulation poses significant challenges due to the fabric’s complex dynamics and the high dimensionality of configuration spaces. Previous approaches have focused on isolated smoothing or folding tasks and relied heavily on simulations, often struggling to bridge the sim-to-real gap. This gap arises as simulated cloth dynamics fail to capture real-world properties such as elasticity, friction, and occlusions, causing accuracy loss and limited generalization. To tackle these challenges, we propose a two-stream architecture with sequential and spatial pathways, unifying smoothing and folding tasks into a single adaptable policy model. The sequential stream determines pick-and-place positions, while the spatial stream, using a connectivity dynamics model, constructs a visibility graph from partial point cloud data, enabling the model to infer the cloth’s full configuration despite occlusions. To address the sim-to-real gap, we integrate real-world human demonstration data via a hand-tracking detection algorithm, enhancing real-world performance across diverse cloth configurations. Our method, validated on a UR5 robot across six distinct cloth folding tasks, consistently achieves desired folded states from arbitrary crumpled initial configurations, with success rates of 100.0%, 100.0%, 83.3%, 66.7%, 83.3%, and 66.7%. It outperforms state-of-the-art cloth manipulation techniques and generalizes to unseen fabrics with diverse colors, shapes, and stiffness. Project page: https://zcswdt.github.io/SSFold/ Note to Practitioners—In this paper, we introduce SSFold, a novel framework for robotic cloth manipulation that integrates human demonstrations with advanced learning techniques, providing a practical solution for real-world applications. Practitioners in industries such as textile manufacturing, automated laundry services, and even medical fabric handling can leverage this method to improve operational efficiency and reduce reliance on manual labor significantly. By using a two-stream architecture to handle complex cloth dynamics and self-occlusions, SSFold unifies smoothing and folding tasks into a single policy model, which adapts effectively to diverse fabric types and conditions. A key advantage of this method lies in its ability to utilize low-cost, easy-to-set-up hand tracking systems for human demonstration data collection, reducing the need for expensive, complex setups. The framework has been successfully validated in real-world scenarios with a UR5 robot, achieving high success rates across a variety of folding tasks. Furthermore, the method demonstrates strong generalization capabilities, making it suitable for a wide range of applications beyond the tasks it was initially trained on. This scalability and flexibility offer a highly practical and cost-effective solution for integrating robotic cloth manipulation into various industries.}, archive = {J_TASE}, author = {Changshi Zhou and Haichuan Xu and Jiarui Hu and Feng Luan and Zhipeng Wang and Yanchao Dong and Yanmin Zhou and Bin He}, doi = {10.1109/TASE.2025.3560680}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14448--14460}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SSFold: Learning to fold arbitrary crumpled cloth using graph dynamics from human demonstration}, volume = {22}, year = {2025}, }
• (2025). Practical reinforcement learning using time-efficient model-based policy optimization. TASE, 22, 14436--14447. (www)

  @article{ , abstract = {In this paper, we propose practical model-based policy optimization (PMBPO) to address the time efficiency issue caused by overly frequent model updates in recent probabilistic model-based reinforcement learning (MBRL) methods that accelerate learning by generating samples from the model. PMBPO enhances the reliability of the generated samples by introducing an expressive probabilistic model that focuses on the system’s dynamic features over continuous time steps. A time-efficient learning framework is proposed by offline updating and interacting with the model at the end of each epoch. One policy fallback mechanism is further designed to mitigate the negative impact of model bias on the learned policy. Evaluated on five Mujoco control benchmarks and one quadruped robot control scenario, PMBPO reduces the one-step computation time by 90% while achieving 70% more cumulative rewards compared to the state-of-the-art MBRL approaches. It extends the feasibility of MBRL in practical control scenarios. The code of PMBPO is available at https://github.com/mrjun123/PMBPO Note to Practitioners—Recent model-based reinforcement learning methods like MBPO, which efficiently learn tasks by generating samples from an approximated system model, are difficult to apply in practical control scenarios due to extensive system pauses caused by overly frequent model updates. This paper proposes a practical MBPO specific for control problems. It reduces the computational burden of real-time learning while surpassing the recent MBPO approaches in learning efficiency and control capability.}, archive = {J_TASE}, author = {Wenjun Huang and Yunduan Cui and Lei Peng and Xinyu Wu}, doi = {10.1109/TASE.2025.3561771}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14436--14447}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical reinforcement learning using time-efficient model-based policy optimization}, volume = {22}, year = {2025}, }
• (2025). Safe data-driven predictive control. TASE, 22, 14422--14435. (www)

  @article{ , abstract = {In the realm of control systems, model predictive control (MPC) has exhibited remarkable potential; however, its reliance on accurate models and substantial computational resources has hindered its broader application, especially within real-time nonlinear systems. This study presents an innovative control framework to enhance the practical viability of the MPC. The developed safe data-driven predictive control aims to eliminate the requirement for precise models and alleviate computational burdens in the nonlinear MPC (NMPC). This is achieved by learning both the system dynamics and the control policy, enabling efficient data-driven predictive control while ensuring system safety. The methodology involves a spatial temporal filter (STF)-based concurrent learning for system identification, a robust control barrier function (RCBF) to ensure the system safety amid model uncertainties, and a RCBF-based NMPC policy approximation. An online policy correction mechanism is also introduced to counteract performance degradation caused by the existing model uncertainties. Demonstrated through simulations on two applications, the proposed approach offers comparable performance to existing benchmarks with significantly reduced computational costs. Note to Practitioners—Model predictive control (MPC) implementation causes high computational cost due to solving an online optimization problem at each time step. To prevail over this challenge, one prevalent approach involves the utilization of model-reduction techniques to simplify the system dynamics. However, even after the model reduction, the computational burden often remains substantial. Another sound approach involves the implementation of function approximators, such as the NNs and the GPR, to learn the MPC policy. However, the trained controllers suffer from several shortcomings: 1) Unlike the MPC that ensures system safety, trained controllers offer no assurance on safety, and 2) The trained controllers inevitably cause performance degradation due to the model uncertainties. These issues have significantly hindered the adoption of such controllers in real-world engineering applications.}, archive = {J_TASE}, author = {Amin Vahidi-Moghaddam and Kaian Chen and Kaixiang Zhang and Zhaojian Li and Yan Wang and Kai Wu}, doi = {10.1109/TASE.2025.3560600}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14422--14435}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe data-driven predictive control}, volume = {22}, year = {2025}, }
• (2025). Analysis of conventional robot task sequences during lot switching periods in cluster tools. TASE, 22, 14407--14421. (www)

  @article{ , abstract = {The semiconductor manufacturing system operates in lot units comprising wafers with identical recipes. Production predominantly occurs within a cluster tool, encompassing multiple single-wafer processing chambers and a wafer-handling robot. Recently, the increasing popularity of small lot sizes has been driven by the diversification of customer demand and larger wafer diameters. Thus, cluster tools frequently encounter lot switching periods, where lots with different recipes are processed consecutively. Existing research on cluster tool scheduling primarily relies on conventional backward and swap sequences optimized for cyclic robot operations. This approach persists even during lot switching period scheduling, which represents non-cyclic scenarios. However, it remains unclear whether these conventional robot task sequences remain optimal or ensure consistent performance during lot switching periods. Therefore, we introduce a theoretical analysis of the performance of the conventional robot sequences, such as backward and swap sequences, in lot switching periods for the first time. We examine various dominant relations, establishing the conditions under which the conventional sequences provide an optimal schedule. We also show worst-case performance bounds. Finally, given the limited performance of the conventional swap sequence during lot switching periods for dual-armed robots, we propose a new form of the robot task sequence and demonstrate its effectiveness. Note to Practitioners—The conventional robot task sequences, including backward and swap, are widely known to perform optimally in cyclic situations where the same lot is processed consecutively. Therefore, many cluster tool scheduling studies assumed these sequences were the basic movements of robots and proceeded with their research based on them. Such assumptions are applied even during the lot switching period when two different lots are considered. The performance of the robot task sequence is crucial because the tool performance depends on it. However, when the recipes of two consecutive lots differ, it is uncertain whether the conventional backward and swap sequences still offer optimal performance. Hence, this study theoretically verifies whether the conventional sequence still performs well during the lot-switching period. We believe that our findings will provide sufficient theoretical support for utilizing the conventional sequence in non-cyclic scheduling in the future. Furthermore, by proposing a new form of sequence suitable for the non-cyclic situations of dual-armed cluster tools, we suggest the potential for expanding to a more diverse range of sequences beyond the conventional ones.}, archive = {J_TASE}, author = {Jeongsun Ahn and Hyun-Jung Kim}, doi = {10.1109/TASE.2025.3562191}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14407--14421}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Analysis of conventional robot task sequences during lot switching periods in cluster tools}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant soft sensor modeling based on a two-dimensional group distributionally robust optimization framework. TASE, 22, 14396--14406. (www)

  @article{ , abstract = {In industrial automation and intelligence, fault tolerance mechanisms have always been an attractive topic. To develop soft sensors with fault tolerance for different types of faults and unforeseen new faults, this article proposes a two-dimensional group distributionally robust optimization (2D-GDRO) framework for fault-tolerant soft sensor modeling. We propose to describe the potential distributions of new fault conditions with an uncertainty set and optimize the soft sensor model by minimizing the worst-case risk over the uncertainty set. Considering the restricted representation range of the uncertainty set constructed directly from a mixture distribution of a limited number of existing fault conditions in the training set, a two-dimensional uncertainty set is designed at the group dimension and the sample dimension. To efficiently train a fault-tolerant soft sensor within the 2D-GDRO framework, we introduce a triple-interleaved optimization algorithm. This algorithm integrates mini-batch stochastic gradient descent, exponentiated gradient ascent, and group-wise SoftMax techniques. Finally, the fault tolerance of the 2D-GDRO framework based soft sensor is verified using the Tennessee-Eastman process and the real three-phase flow facility. The experimental results show that 2D-GDRO outperforms other training frameworks in average soft sensing accuracy under new fault conditions. Note to Practitioners—This article addresses the need for fault-tolerant soft sensors in industrial automation, where systems must continue to function reliably despite various types of faults and unforeseen faults. The solution proposed is a 2D-GDRO framework, which aims to make soft sensors more resilient to such faults by preparing them for worst-case scenarios during the training phase. This framework enhances the fault tolerance of soft sensors by designing an uncertainty set at two dimensions: the group dimension and the sample dimension, which allows soft sensors to better generalize beyond known fault conditions. The 2D-GDRO framework uses an efficient triple-interleaved optimization algorithm. While the 2D-GDRO framework significantly improves fault tolerance, its performance depends on the diversity of fault conditions represented in the training data. In cases where fault scenarios are highly unpredictable, additional refinement may be required. The fault-tolerant soft sensor based on 2D-GDRO can be applied across various industries, such as chemical manufacturing, oil and gas, and energy production, ensuring that the predictive monitoring remain reliable even in the presence of unknown new faults.}, archive = {J_TASE}, author = {Xiangrui Zhang and Chunyue Song and Jun Zhao and Biao Huang}, doi = {10.1109/TASE.2025.3561754}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14396--14406}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant soft sensor modeling based on a two-dimensional group distributionally robust optimization framework}, volume = {22}, year = {2025}, }
• (2025). CBGTE: Neural network aided extended kalman filter for dual-band infrared attitude estimation. TASE, 22, 14383--14395. (www)

  @article{ , abstract = {Due to the numerous advantages of dual-band infrared radiation (DBIR) attitude measurement (AM) technology, it has garnered significant attention from industry and academia. However, geometric errors caused by sensor measurement noise, assembly positions, motor interference and other sensor-related commonalities, along with random errors induced by the sensitivity of DBIR characteristics to infrared radiation interference, collectively undermine the reliability of the estimated attitude information. To address this issue, the bidirectional gated recurrent unit (Bi-GRU) and Transformer were combined to assist extended kalman filter (EKF) for DBIR attitude estimation (CBGTE). The core concept involves two aspects: 1) utilizing EKF to mitigate geometric errors during the process of DBIR AM, and 2) combining Bi-GRU and Transformer to aid EKF in compensating for random errors in the measurement process. A semi-physical experimental platform is established to validate the performance of CBGTE. Through experimental validation with real-world data, the proposed CBGTE algorithm has demonstrated significant improvements in accuracy when compared with several state-of-the-art algorithms, achieving roll angle error of ±0.4° and pitch angle error of ±0.2°. Note to Practitioners—The neural network aided extended kalman filter for dual-band infrared radiation attitude estimation can be applied to various fields, including autonomous aerial vehicles, satellite systems, flight projectiles, and etc. This study focuses on the flight projectile system as the research object. Flying projectiles serve as primary tools for striking adversaries, where precision attitude is paramount. Compared to other attitude estimation technologies, dual-band infrared radiation attitude estimation technology has several advantages, including without cumulative error, resistance to electromagnetic interference, passive operation, and low cost. However, the flying projectiles are a nonlinear complex system characterized by multiple variables and strong coupling. The attitude information is susceptible to noise interference, which consequently undermines its reliability. To address this issue, we propose the CBGTE algorithm. The proposed CBGTE algorithm enhances the robustness of dual-band infrared attitude measurement technology and enables precise attitude angle calculation through experimental validation. Furthermore, this method effectively reduces hardware costs and facilitates practical applications.}, archive = {J_TASE}, author = {Xiaoyu Han and Miaomiao Xu and Yihan Cao and Xiongzhu Bu and Qiang Sun}, doi = {10.1109/TASE.2025.3561781}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14383--14395}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {CBGTE: Neural network aided extended kalman filter for dual-band infrared attitude estimation}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered networked adaptive tracking control of wind turbine systems. TASE, 22, 14371--14382. (www)

  @article{ , abstract = {This paper addresses the co-design problem of adaptive control for networked wind turbine systems that track the desired rotor speed while efficiently scheduling network communication. Unlike existing approaches, the proposed method integrates the control design and the communication considerations, ensuring asymptotic tracking of rotor speed under the generator torque saturation with significantly reduced communication load. Firstly, the communication scheme is developed using the dynamic event-triggering mechanism, which introduces the feedback signal in the sampling loop. Secondly, an auxiliary signal is designed to mitigate the negative effect of inevitable generator torque saturation, ensuring that the bounded controller asymptotically exits from saturation. Then, the adaptive torque controller is constructed with the compensation signals to guarantee asymptotic stability, and the measurement function for dynamic event-triggering is co-designed alongside the controller to regulate the sampling-induced error. Furthermore, the requirement for exact knowledge of the wind turbine systems is eliminated by utilizing an online approximator to learn the uncertain aerodynamics and parameters. Finally, it is theoretically proven that all the signals in the closed-loop system are bounded, and the rotor speed asymptotically tracks the reference rotor speed. The feasibility and advantages of the proposed method are demonstrated on the NREL 5-MW wind turbine using the high-fidelity OpenFAST simulation platform. Note to Practitioners—This work is motivated by the prevalent networked control system paradigm of wind energy systems, where network communication has become ubiquitous due to the increasing scale of wind energy and the integration of internet of things devices. Therefore, the co-design of communication and adaptive control of wind turbine systems is imperative for developing modern wind energy systems. This paper presents a co-design approach that integrates the adaptive asymptotic tracking control method with an aperiodic communication scheme designed by the dynamic event-triggering mechanism. This approach significantly reduces communication costs while achieving competitive control performance compared to the traditional methods that require continuous communication. To enhance the practical applicability, the proposed method concurrently addresses several challenges in the networked control of wind turbine systems, including redundant network communication, uncertain nonlinearities, and input saturation encountered in practical applications, providing a networked control solution with efficient communication and guaranteed stability. The validation results demonstrate the effectiveness and advantages of the proposed method.}, archive = {J_TASE}, author = {Jun Chen and Wenchao Meng and Yingjie Gong and Qinmin Yang}, doi = {10.1109/TASE.2025.3557185}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14371--14382}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered networked adaptive tracking control of wind turbine systems}, volume = {22}, year = {2025}, }
• (2025). ICODE: Modeling dynamical systems with extrinsic input information. TASE, 22, 14358--14370. (www)

  @article{ , abstract = {Learning models of dynamical systems with external inputs, which may be, for example, nonsmooth or piecewise, is crucial for studying complex phenomena and predicting future state evolution, which is essential for applications such as safety guarantees and decision-making. In this work, we introduce Input Concomitant Neural ODEs (ICODEs), which incorporate precise real-time input information into the learning process of the models, rather than treating the inputs as hidden parameters to be learned. The sufficient conditions to ensure the model’s contraction property are provided to guarantee that system trajectories of the trained model converge to a fixed point, regardless of initial conditions across different training processes. We validate our method through experiments on several representative real dynamics: Single-link robot, DC-to-DC converter, motion dynamics of a rigid body, Rabinovich-Fabrikant equation, Glycolytic-glycogenolytic pathway model, and heat conduction equation. The experimental results demonstrate that our proposed ICODEs efficiently learn the ground truth systems, achieving superior prediction performance under both typical and atypical inputs. This work offers a valuable class of neural ODE models for understanding physical systems with explicit external input information, with potentially promising applications in fields such as physics and robotics. Our code is available online at https://github.com/EEE-ai59/ICODE.git Note to Practitioners—Understanding and accurately predicting the behavior of complex dynamical systems is essential for a wide range of practical applications, from robotics and control systems to energy management and biology. Many real-world systems involve external inputs that can be irregular, nonsmooth, or change abruptly, which complicates the modeling process. The method introduced in this paper, Input Concomitant Neural ODEs (ICODEs), offers a new way to integrate these inputs directly into the system model, rather than treating them as unknown or hidden factors. This approach ensures that the model can better reflect real-world conditions, where inputs play a critical role in system behavior. One of the key advantages of ICODEs is their ability to ensure that the system stabilizes to a predictable state, even when trained on different data or starting conditions. This property is particularly useful in safety-critical applications where consistent and reliable system behavior is essential, such as in robotics or in power electronics like DC-to-DC converters. By learning dynamic systems more accurately, ICODEs provide practitioners with a powerful tool for developing systems that can better handle varying inputs while capturing system characteristics. Whether for designing safer robots, optimizing industrial systems, or improving energy efficiency, ICODEs offer promising potential for real-world applications requiring precise modeling of input-driven dynamics.}, archive = {J_TASE}, author = {Zhaoyi Li and Wenjie Mei and Ke Yu and Yang Bai and Shihua Li}, doi = {10.1109/TASE.2025.3560450}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14358--14370}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ICODE: Modeling dynamical systems with extrinsic input information}, volume = {22}, year = {2025}, }
• (2025). Deep reinforcement learning-assisted robust cubature kalman filter for power system dynamic state estimation with multi-rate measurements. TASE, 22, 14346--14357. (www)

  @article{ , abstract = {The coexistence of high-frequency phasor measurement units (PMUs) and conventional SCADA systems raises the challenge of heterogenous-source and multi-rate measurements which significantly degrades the performance of power system dynamic state estimation. In this work, a deep reinforcement learning (DRL) assisted robust cubature Kalman filtering (CKF) scheme is proposed to handle measurements from hybrid sources and with different time scales. In specific, a multi-rate measurement function reconstruction approach is designed with an independent discretization mechanism to lift the present limitation of requiring an integer multiple relationship of the sampling rates from multiple sources in most of existing works. Embedded with this discretization mechanism, a deep reinforcement learning assisted two-parameter linear exponential smoothing method is proposed to reconstruct the slow measurement model with online adjustable estimation parameters. A generalized correntropy loss criterion is also included in the robust CKF to counter the non-Gaussian noise and the noise distribution variation caused by the reconstruction. Comparisons results demonstrate that the proposed DRL-based robust CKF method can achieve better accuracy and robustness under various operating scenarios. Note to Practitioners—This work aims to improve the performance of the power system state estimation function in power system control center. Most of the power system state estimation approaches handle measurements with same sampling rates and/or with a multiple relationships of sampling rates. However, in power system practice, PMU and SCADA units exist simultaneously and their sampling rates may be not with a multiple relationship. This work proposes a deep reinforcement learning assisted robust cubature Kalman filter embedded with a multi-rate sampling integration mechanism. It will be able to achieve better estimation accuracy under time-varying operation conditions in power systems such as sudden shifts of load and renewable generations. The method can be integrated into the existing GUI in power system control center where the state estimation also exists. The deep reinforcement learning may increase the need of computing capability of control center which has been upgrading by most of power utilities.}, archive = {J_TASE}, author = {Haoli Gu and Shichao Liu and Bo Chen and Rusheng Wang and Li Yu and Okyay Kaynak}, doi = {10.1109/TASE.2025.3560081}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14346--14357}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep reinforcement learning-assisted robust cubature kalman filter for power system dynamic state estimation with multi-rate measurements}, volume = {22}, year = {2025}, }
• (2025). Observer-based multi-agent reinforcement learning for pursuit-evasion game with multiple unknown uncertainties. TASE, 22, 14332--14345. (www)

  @article{ , abstract = {This paper aims to investigate the challenging problem of a multi-agent game with multiple pursuers and a single evader in an environment with multiple unknown uncertainties. A coupled approach combining decentralized observers and reinforcement learning (RL) controllers is proposed to deal with this scenario. Firstly, decentralized observers driven by auxiliary control laws are introduced to estimate the states of uncertain systems, with their best responses obtained through the adaptive dynamic programming (ADP) method. The estimated states, which reflect the actual states of the pursuers’ systems, are concurrently transmitted to the RL controllers. Subsequently, the controllers are trained with observer-based heterogeneous-agent proximal policy optimization (OHAPPO) algorithm, in which a novel global multi-function cost is designed. The algorithm utilizes the advantage decomposition for policy updates in the way of credit assignment, resulting in more stable and efficient updates compared to traditional value decomposition. Moreover, to further enhance the performance of both observers and controllers, a sequential game is established between them, where observers’ policies are influenced by controllers’ optimal control and vice versa. Finally, the simulation results verify the effectiveness of the designed OHAPPO algorithm in the pursuit-evasion game. Note to Practitioners—Numerous military applications, such as missile interception and pursuit or encirclement on sea and land, can be effectively delegated to a group of experienced and intelligent agents. These agents offer safety and efficiency advantages in complex and dynamic adversarial environments. RL methods are good at dealing with such scenarios, but their application in complex real-world scenarios is often hindered, as multiple unknown uncertainties in the environment can affect the effectiveness and stability of strategies, as well as the generalization of simulation to reality transfer and the high cost of training directly in real environments. This study takes the pursuit-evasion game as an example task, forming a universal coupled control method. This method allows active observers with game properties to overcome multiple unknown uncertainties. It enables RL to be trained in relatively real environments, overcoming application limitations caused by uncertainty and facilitating RL research in real scenarios. Due to hardware limitations, the next step has not yet been taken, and it will address testing and deployment issues in real-world environments.}, archive = {J_TASE}, author = {Yangyang Liu and Chun Liu and Yizhen Meng and Bin Jiang and Xiaofan Wang}, doi = {10.1109/TASE.2025.3560757}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14332--14345}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based multi-agent reinforcement learning for pursuit-evasion game with multiple unknown uncertainties}, volume = {22}, year = {2025}, }
• (2025). Graph-based variation propagation network for modeling and prediction of hybrid multi-stage manufacturing systems. TASE, 22, 14317--14331. (www)

  @article{ , abstract = {Multistage Manufacturing Systems (MMS) are common in industries involving complex processes with multiple stages, each impacting the final product quality. Traditional product quality modeling approaches struggle with the intricate interdependencies and variable structures within these systems, further complicated by the extensive sensor-generated data. In this paper, we introduce Graph-based Variation Propagation Network (GVPNet), an innovative approach for end-to-end learning of product quality representations and their propagation through MMS stages. GVPNet utilizes a heterogeneous Graph Attention Network (hGAT) architecture that uses a graph representation specifically tailored for MMS, facilitating the effective application of graph neural networks (GNN) in this context. The network’s ability to aggregate information and learn node embeddings enables it to predict quality variables several stages ahead, thus offering a proactive tool for quality control. GVPNet was applied in two real world case studies and demonstrated its superiority in predictive accuracy and robustness over benchmarks. Note to Practitioners—This paper introduces the Graph-based Variation Propagation Network (GVPNet), designed to enhance quality prediction in Multistage Manufacturing Systems (MMS) where traditional models fall short. Integrating a tailored graph representation with the Graph Attention Network (hGAT), GVPNet excels in capturing the complex dynamics of MMS, including inter-stage dependencies. This approach enables more accurate and proactive quality control in various manufacturing contexts, providing a significant tool for practitioners to optimize processes and improve product quality, even in data-rich environments.}, archive = {J_TASE}, author = {Dhari F. Alenezi and Jianjun Shi and Jing Li}, doi = {10.1109/TASE.2025.3560174}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14317--14331}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Graph-based variation propagation network for modeling and prediction of hybrid multi-stage manufacturing systems}, volume = {22}, year = {2025}, }
• (2025). A systematic literature review of the characteristics of automation systems that impact debugging of PLC software. TASE, 22, 14306--14316. (www)

  @article{ , abstract = {The debugging capabilities of software development tools in the automation domain lag behind other domains. Basic capabilities from regular debuggers, such as breakpoints and step-by-step execution, are hard to find or do not support all requirements imposed by the domain. This work presents the results of a Systematic Literature Review carried out to understand the characteristics of the automation domain that influence the debugging capabilities of Programmable Logic Controller (PLC) software and to find the current state and future directions of debugging tools. The process of the review was prepared and reviewed before its execution to avoid any type of bias. The review successfully finds the major characteristics affecting the debugging of PLC software which include the presence of interconnected components and heavy machinery from different domains, the participation of different stakeholders at different phases of the development of the automation systems, and the restricted software capabilities specific to the domain. The existing debugging capabilities are mainly limited to monitoring and forcing the current state of the PLC software. However, research is being done in different areas that help debug PLC software, such as testing and simulation, and static and dynamic code analysis. The latter helps develop more powerful debugging capabilities such as the reproduction of real-time scenarios and replay debugging. The usability of debugging tools is also a subject of study in the automation domain. The work ends with an overview of future research directions to close the gap to other domains. Note to Practitioners—This paper was motivated by the limited capabilities found when developing software for Programmable Logic Controllers (PLC) used in the automation domain. A Systematic Literature Review (SLR) was conducted to understand the reasons for the limited debugging capabilities and to know the current state of debugging tools in the automation domain. For the SLR, search terms were defined and used to look for research studies in different online libraries. The result of the search was filtered based on inclusion and exclusion criteria, and the remaining resources were read in detail. From the information found, the debugging capabilities in PLCs are limited due to the machinery connected to them, the different people involved while creating automation systems, and restrictions of programming standards used to develop software for PLCs. The current debugging capabilities for PLCs are mainly limited to monitoring the current state and forcing new ones. However, research has been done to extend the debugging capabilities in the areas of testing, simulation, static and dynamic code analysis, and reproduction of real-time scenarios. This paper provides at the end research directions to further improve debugging capabilities for PLC software.}, archive = {J_TASE}, author = {Jose Cabral and Alois Zoitl}, doi = {10.1109/TASE.2025.3559790}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14306--14316}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A systematic literature review of the characteristics of automation systems that impact debugging of PLC software}, volume = {22}, year = {2025}, }
• (2025). Orchestrated robust controller for precision control of heavy-duty hydraulic manipulators. TASE, 22, 14284--14305. (www)

  @article{ , abstract = {Vast industrial investment along with increased academic research on heavy-duty hydraulic manipulators has unavoidably paved the way for their automatization, necessitating the design of robust and high-precision controllers. In this study, an orchestrated robust controller is designed to address the mentioned issue for generic manipulators with an anthropomorphic arm and spherical wrist. Thanks to virtual decomposition control (VDC), the entire robotic system is decomposed into subsystems, and a robust controller is designed at each local subsystem by considering unknown model uncertainties, unknown disturbances, and compound input nonlinearities. As such, radial basis function neural networks (RBFNNs) are incorporated into VDC to tackle unknown disturbances and uncertainties, resulting in novel decentralized RBFNNs. All robust local controllers designed at each local subsystem, then, are orchestrated to accomplish high-precision control. In the end, for the first time in the context of VDC, a semi-globally uniformly ultimate boundedness is achieved under the designed controller. The validity of the theoretical results is verified by performing extensive simulations and experiments on a 6-degrees-of-freedom industrial manipulator with a nominal lifting capacity of $600\, kg$ at 5 meters reach. Comparing the simulation results with the state-of-the-art controllers along with provided experimental results, demonstrates that proposed method fulfilled all promises and performed excellently. Note to Practitioners—Heavy-duty hydraulic manipulators (HHMs) play a pivotal role in various industrial applications, where their automation is critical for enhancing productivity, precision, and operational efficiency. However, automating HHMs presents significant challenges due to their complex dynamics, highly nonlinear behavior, and presence of unknown real-world uncertainties, such as unknown disturbances, non-ideal devices, unmodeled dynamics, and limited computational resources. These challenges underscore the need for control algorithms that are robust to such uncertainties, computationally efficient for real-time deployment, and capable of achieving high accuracy for reliable execution of free-motion tasks. This study designs a control scheme for generic 6-degree-of-freedom (DoF) HHMs, which ensures robustness and accuracy. Theoretical analysis along with experimental validations on a 6-DoF industrial HHM perfectly display stability of the method, robustness to real-world unknown uncertainties, efficiency in real-time implementation, and excellent accuracy in task execution. These results highlight the scheme’s reliability and universality for industrial HHM applications, making it a significant contribution to advancing automation in this field. Additionally, detailed performance index provided with results enhances replicability and measurability, paving the way for meaningful comparisons in the future.}, archive = {J_TASE}, author = {Mahdi Hejrati and Jouni Mattila}, doi = {10.1109/TASE.2025.3559595}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14284--14305}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Orchestrated robust controller for precision control of heavy-duty hydraulic manipulators}, volume = {22}, year = {2025}, }
• (2025). Event-driven robust geometric control for a nano unmanned helicopter. TASE, 22, 14272--14283. (www)

  @article{ , abstract = {This work presents the design of an event-driven position tracking control law for a nano unmanned helicopter with a body length less than 25 centimeters and a mass less than 70 grams. A robust geometric controller based on robust integral sign error (RISE) is devised to compensate for the unmanned helicopter’s modeling uncertainties as well as unknown external disturbances. An event-driven robust nonlinear control technique is formulated for the nano helicopter’s position loop control taking into account the very limited onboard computing resources of the nano unmanned helicopter. The stability of the helicopter’s closed-loop system is analyzed. And practical flying experiments are performed to verify the efficiency of the control method. The experimental results demonstrate that not only the tracking accuracy is guaranteed, but also the proposed control method can also successfully reduce the consumption of the onboard flight control unit’s computing resources. Note to Practitioners—The motivation for this paper stems from the need to develop a control algorithm with low computational resource consumption that is suitable for nano unmanned helicopters. Due to the unique structure of nano unmanned helicopters, existing control algorithms designed for quadrotor drones cannot be directly applied to nano unmanned helicopters. Additionally, the small size and limited payload capacity of nano unmanned helicopters further restrict their available onboard computational resources. To address this issue, we propose an event-driven robust controller that can ensure tracking accuracy and reduce the use of computational resources.}, archive = {J_TASE}, author = {Aochen Ma and Xun Gu and Bin Xian}, doi = {10.1109/TASE.2025.3559506}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14272--14283}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-driven robust geometric control for a nano unmanned helicopter}, volume = {22}, year = {2025}, }
• (2025). Wireless capsule endoscopy diagnosis using prototype self-attention and dynamic curriculum learning. TASE, 22, 14260--14271. (www)

  @article{ , abstract = {Wireless capsule endoscopy is a non-invasive and painless approach for diagnosing gastrointestinal diseases. An automated medical decision support system can significantly enhance clinician efficiency and reduce the incidence of misdiagnosis when analyzing lesions within the 50,000 to 100,000 image frames generated for each individual. However, only a small number of previous studies have focused on fine-grained recognition characteristics, and very few have simultaneously addressed both fine-grained recognition and class imbalance issues in wireless capsule endoscopy. To address these issues, we propose a novel medical decision support network that includes a prototype attention enhancement module and a dynamic curriculum learning approach. The prototype attention enhancement module improves lesion-sensitive compact representation learning by leveraging cosine similarity between the representation and a prototype memory with multi-class centers. The dynamic curriculum learning adopts triplet loss and weighted cross-entropy, facilitated by a progressive factor that controls between fine-grained representation learning and class-sensitive learning. Through extensive comparative experiments on three public datasets, the proposed network demonstrated competitive performance, outperforming previous methods with an F1-score of 96.6% on the 10-class Kvasir-Capsule dataset, an accuracy of 98.3% on the CAD-CAP dataset, and an accuracy of 93.7% on the mixed KID dataset. Two additional gastrointestinal histopathology datasets confirm the generalization of the proposed network. The code will be available at https://github.com/Xingcun-Li/MDSN-WCE Note to Practitioners—The primary aim of this study is to address diagnostic challenges in automated wireless capsule endoscopy (WCE). WCE provides a non-invasive diagnostic method for gastrointestinal diseases, but manual analysis of the resulting large volume of image data is time-consuming and error-prone. To this end, we propose a novel medical decision support network, which is also applicable to similar medical image analysis issues characterized by class imbalance and fine-grained recognition. Our approach integrates a prototype attention enhancement (PAE) module and dynamic curriculum learning (DCL), enabling the model to accurately identify lesions within WCE images and exhibit favorable performance in gastrointestinal histopathology image recognition beyond the WCE domain. Additionally, our approach can be applied to various common deep learning models without incurring significant additional computational costs, thereby aiding endoscopists in significantly reducing the time required for diagnosing WCE video frames. The code will be uploaded to GitHub for direct use in the automated diagnosis of WCE video frames. While our method shows promise, its main focus is on rapidly identifying a large number of frames within WCE videos. However, further work is needed to extract key frames and improve the interpretability of outputs from WCE videos, which is also a future research direction.}, archive = {J_TASE}, author = {Xingcun Li and Qinghua Wu and Yuning Chen and Kun Wu and Lin Meng}, doi = {10.1109/TASE.2025.3558934}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14260--14271}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Wireless capsule endoscopy diagnosis using prototype self-attention and dynamic curriculum learning}, volume = {22}, year = {2025}, }
• (2025). Admittance-based output feedback fuzzy switching control for PAM-driven parallel robots via nonsingular terminal sliding mode. TASE, 22, 14247--14259. (www)

  @article{ , abstract = {As a kind of soft actuator with inherent compliance, pneumatic artificial muscles (PAMs) have great application potential in robots. However, some challenging issues, such as high nonlinearities, sensor noises, and external disturbances, inevitably bring enormous difficulties to the accurate control of PAM-driven robots. To this end, this paper proposes an adaptive output feedback fuzzy switching control method for switched-form PAM-driven parallel robot systems, utilizing admittance models to rebuild compliant trajectories. Specifically, based on the nonrecursive high-order sliding mode (HOSM) differentiators with fixed-time convergence, unmeasurable velocity signals can be reconstructed to eliminate the adverse effects of measurement noises, decreasing the time delay of feedback signals. Moreover, a soft switching strategy is designed to flexibly adjust the switching weights and intervals of fuzzy structures, maintaining smooth control commands. Further, by introducing a nonsingular terminal sliding manifold, tracking errors can rapidly converge to a small neighborhood around the origins within a finite time, and all closed-loop variables are proved to be bounded through the Lyapunov stability theory. Finally, several groups of experiments are carried out on a self-built PAM-driven parallel robot to verify the effectiveness of the suggested method. Note to Practitioners—More compliant requirements need to be satisfied when facing an increasing number of human-robot interaction tasks. Naturally inheriting flexibility, pneumatic artificial muscles (PAMs) can achieve positioning/tracking control with various merits, including large actuated forces, low friction, clean power, etc. Once injecting a large amount of compressed gas, the overall stiffness of the system will significantly increase; hence, it is still an open challenge to combine compliant methods with high-efficiency tracking control for PAM-driven robots. To address this problem, based on Backlash-like hysteresis models, the whole switching dynamics of the PAM-driven parallel robots can be fully described, considering more comprehensive modeling elements. Then, by using admittance models, the closed-loop control is endowed with adjustable flexibility. Together with rapid velocity estimation, a nonsingular terminal sliding mode with finite-time convergence and a switching fuzzy update law are both introduced to further enhance the tracking performance and response speed. This combination gives the first solution to simultaneously improve the interaction compliance, response speed, and control accuracy for this kind of soft actuator-driven robots, which provides potential applications for practitioners.}, archive = {J_TASE}, author = {Xinlin Zhang and Gendi Liu and Shuzhen Diao and Tong Yang and Yongchun Fang and Ning Sun}, doi = {10.1109/TASE.2025.3558891}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14247--14259}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Admittance-based output feedback fuzzy switching control for PAM-driven parallel robots via nonsingular terminal sliding mode}, volume = {22}, year = {2025}, }
• (2025). Robust output regulation with prescribed performance for second-order non-polynomial nonlinear systems. TASE, 22, 14234--14246. (www)

  @article{ , abstract = {This article focuses on the robust output regulation problem with prescribed performance for uncertain second-order nonlinear systems with non-polynomial nonlinearity. According to the existing framework for the robust output regulation problem of such nonlinear systems, we introduce some standard assumptions, and then construct two internal models of nonlinear forms to reproduce the steady-state information. Then, we convert the considered problem into the robust stabilization problem with prescribed performance of an augmented system containing the studied nonlinear system and the internal models. We incorporate a novel error transformation into the existing technique to design a stabilization controller, based on which we synthesize our control law to solve our original problem. Finally, we present two simulation examples to test our results. Note to Practitioners—This paper is motivated by a goal to ensure satisfaction of prescribed performance when achieving output regulation, rather than only considering steady-state performance traditionally. In addition to steady-state performance, transient performance also matters in the practical application. For example, regarding the microelectromechanical systems, guarantee of transient performance can reduce or prevent collisions between the movable plate and the fixed one, thus prolonging the device lifetime. We introduce the concept of prescribed performance to include transient and steady-state performance. Satisfaction of prescribed performance can protect both devices and humans in practice, effectively expanding the practical applicability of the output regulation theory. The theory of robust output regulation can avoid applying the solution of regulator equations related to the plant, which is quite practical since many parameters in the models of devices can be hardly obtained accurately. In addition, we consider second-order nonlinear systems with non-polynomial nonlinearity, including more physical systems compared with polynomial ones. For example, the models of single-link robotic manipulator systems contain the sine function, i.e., non-polynomial nonlinearity, for which our results are also suitable. All in all, the results in this paper have not only profound theoretical significance, but also important practical application value.}, archive = {J_TASE}, author = {Wei Liu and Jialei Luo and Weijie Sun}, doi = {10.1109/TASE.2025.3560099}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14234--14246}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust output regulation with prescribed performance for second-order non-polynomial nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). A cross-modal mutual knowledge distillation framework for alzheimer’s disease diagnosis: Addressing incomplete modalities. TASE, 22, 14218--14233. (www)

  @article{ , abstract = {Early detection of Alzheimer’s Disease (AD) is crucial for timely interventions and optimizing treatment outcomes. Integrating multimodal neuroimaging datasets can enhance the early detection of AD. However, models must address the challenge of incomplete modalities, a common issue in real-world scenarios, as not all patients have access to all modalities due to practical constraints such as cost and availability. We propose a deep learning framework employing Incomplete Cross-modal Mutual Knowledge Distillation (IC-MKD) to model different sub-cohorts of patients based on their available modalities. In IC-MKD, the multimodal model (e.g., MRI and PET) serves as a teacher, while the single-modality model (e.g., MRI only) is the student. Our IC-MKD framework features three components: a Modality-Disentangling Teacher (MDT) model designed through information disentanglement, a student model that learns from classification errors and MDT’s knowledge, and the teacher model enhanced via distilling the student’s single-modal feature extraction capabilities. Moreover, we show the effectiveness of the proposed method through theoretical analysis and validate its performance with simulation studies. In addition, our method is demonstrated through a case study with Alzheimer’s Disease Neuroimaging Initiative (ADNI) datasets, underscoring the potential of artificial intelligence in addressing incomplete multimodal neuroimaging datasets and advancing early AD detection. Note to Practitioners—This paper was motivated by the challenge of early AD diagnosis, particularly in scenarios when clinicians encounter varied availability of patient imaging data, such as MRI and PET scans, often constrained by cost or accessibility issues. We propose an incomplete multimodal learning framework that produces tailored models for patients with only MRI and patients with both MRI and PET. This approach improves the accuracy and effectiveness of early AD diagnosis, especially when imaging resources are limited, via bi-directional knowledge transfer. We introduced a teacher model that prioritizes extracting common information between different modalities, significantly enhancing the student model’s learning process. This paper includes theoretical analysis, simulation study, and real-world case study to illustrate the method’s promising potential in early AD detection. However, practitioners should be mindful of the complexities involved in model tuning. Future work will focus on improving model interpretability and expanding its application. This includes developing methods to discover the key brain regions for predictions, enhancing clinical trust, and extending the framework to incorporate a broader range of imaging modalities, demographic information, and clinical data. These advancements aim to provide a more comprehensive view of patient health and improve diagnostic accuracy across various neurodegenerative diseases.}, archive = {J_TASE}, author = {Min Gu Kwak and Lingchao Mao and Zhiyang Zheng and Yi Su and Fleming Lure and Jing Li}, doi = {10.1109/TASE.2025.3556290}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14218--14233}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A cross-modal mutual knowledge distillation framework for alzheimer’s disease diagnosis: Addressing incomplete modalities}, volume = {22}, year = {2025}, }
• (2025). CDPMM-DMP: Conditional dirichlet process mixture model-based dynamic movement primitives. TASE, 22, 14201--14217. (www)

  @article{ , abstract = {Movement Primitives(MPs) are compact generators for representation and generalization of modular movements, which are usually used to implement learning from demonstration tasks in robotics. Existing works on MPs mostly utilize combinations of basis functions to represent diverse movements, whether employing probabilistic or dynamic approaches. However, applying these approaches requires manual specification of hyperparameters related to basis functions, resulting in inconvenience and a reliance on specific expertise. In this paper, we develop a Conditional Dirichlet Process Mixture Model-based Dynamic Movement Primitive (CDPMM-DMP) to achieve a nonparametric improvement for the Dynamic Movement Primitive (DMP). First, inspired by Bayesian nonparametric theory, we explore the use of the Dirichlet Process Mixture Model (DPMM) to replace the original radial basis functions in the DMP, and construct the required training set from demonstrations. Then, we study the output generation mechanism driven by the DPMM, particularly by employing conditional sampling to avoid the anomalous outputs caused by direct sampling from the DPMM. Finally, we provide analyses of the various properties brought by our nonparametric transformation of DMP. The analyses and validation results show that the proposed CDPMM-DMP can significantly reduce the parameter tuning burden in usage with its nonparametric learning property. Besides, our method still retains the inherent properties of DMP, while also incorporating some properties of probabilistic MPs, such as multi-sample learning and co-activation. Note to Practitioners—This paper is motivated by alleviating the parameter tuning burden of the Dynamic Movement Primitive (DMP) in its learning process. In our research, we address this issue by introducing a nonparametric probabilistic model (namely, Dirichlet process mixture model) into the learning process of DMP to substitute its original deterministic basis functions. This substitution effectively enables our method to automatically identify the required parameters from the demonstration data. To address the subsequent influences of the probabilistic components on the generation process of DMP and its inherent properties, on one hand, we focus on conditional sampling to establish a mechanism for the generation process. This mechanism can prevent the discontinuity and non-convergence of outputs caused by probabilistic sampling. On the other hand, we explore our method to achieve some probabilistic properties of generalization, which the DMP lacks. Simulation and experimental results show our method can achieve nonparametric learning from demonstrations. Additionally, it not only produces smooth and continuous movement from any initial to final states (similar to DMPs), but also holds probabilistic properties such as multi-sample learning and co-activation (similar to probabilistic primitive methods). In practical applications, due to the CDPMM-DMP’s ability of nonparametric learning, it is more user-friendly for general users lacking specialized experience, aligning with the development goals of robot leaning from demonstrations. Moreover, its properties combining deterministic and probabilistic primitive methods make it more potent when facing complex scenarios and various requirements.}, archive = {J_TASE}, author = {Hao Jiang and Jianping He and Xiaoming Duan}, doi = {10.1109/TASE.2025.3559696}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14201--14217}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {CDPMM-DMP: Conditional dirichlet process mixture model-based dynamic movement primitives}, volume = {22}, year = {2025}, }
• (2025). An efficient real-time railway container yard management method based on partial decoupling. TASE, 22, 14183--14200. (www)

  @article{ , abstract = {Sea-rail intermodal transportation is an essential infrastructure in global supply chains nowadays. Since the yard is the interface between sea and land, optimizing the transportation process in yards is of significant interest for increasing transportation efficiency. However, the yard management problem usually suffers from the large state space and cannot be solved effectively. We consider this important problem from a decomposition perspective, focusing on real-time scheduling in railway container yards, and make the following contributions. We show the partial decomposition property of the yard management problem. This property establishes the equivalence between joint optimization of yard components and independent optimization of each component under mild assumptions, achieving optimality and effectiveness simultaneously. Utilizing this property, we further simplify each subproblem while preserving the optima, and develop algorithms for each simplified subproblem. Numerical experiments demonstrate the performance of proposed algorithms and the contribution of optimizing each subproblem from various perspectives. Note to Practitioners—Yard is the interface between sea and land in sea-rail intermodal transport, thus being the bottleneck of the transport efficiency. Optimizing yard management is important but challenging, since the matching between trains and yards, the allocation of yard storage, and the movement of cranes should be jointly optimized, which compose a complex problem with an exponentially large decision space. Besides, as modifications to already-deployed systems always come with significant costs, improving transport efficiency with minimal adjustments is significant, which emphasizes the importance of identifying the most crucial component. We address these difficulties in this work and show that train matching, yard storage allocation, and crane scheduling can be decoupled under mild assumptions. To utilize the partial decomposition property, algorithms for each subproblem are further proposed respectively, the effectiveness of which is shown in numerical simulation. And the contribution of optimizing each sub-problem is compared, which provides valuable rule-of-thumb for practice.}, archive = {J_TASE}, author = {Yao Luan and Qing-Shan Jia and Yi Xing and Zhiyu Li and Tengfei Wang}, doi = {10.1109/TASE.2025.3559241}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14183--14200}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An efficient real-time railway container yard management method based on partial decoupling}, volume = {22}, year = {2025}, }
• (2025). Secure cloud-edge collaborative method for dynamic industrial process monitoring using self-updating dictionary learning. TASE, 22, 14170--14182. (www)

  @article{ , abstract = {Modern industrial systems possess the capacity to accumulate substantial data, thereby enabling data-driven process monitoring. However, the dynamic industrial processes give rise to new working conditions continuously. Traditional methods require substantial new data for updates, resulting in considerable delays. Furthermore, edge devices face resource limitations, which complicates the ability to meet the increasing demands for storage and computing power. On the other hand, the frequent transmission of data between the cloud and edge introduces potential security risks. To address these challenges, this paper proposes a Secure Monitoring Method based on Self-Updating Dictionary Learning (SUDL-SM) within a cloud-edge collaboration framework. Specifically, to tackle the issue of poor model adaptability caused by limited data of new modes, this paper first proposes a dictionary learning method based on multi-task hardness evaluation. By evaluating the multi-dimensional contributions of samples, those with strong generalization are extracted, and the dictionary is updated online accordingly, ensuring adaptability to both new and historical conditions. Subsequently, due to the resource constraints inherent in edge devices, a dictionary distillation compression method has been proposed. This method aims to maximize dictionary compression while preserving the original monitoring performance, thereby ensuring efficient and accurate inference on edge devices. Finally, a hybrid encryption-based cloud and edge data transmission protocol is designed to effectively address malicious activities such as data theft and tampering by ensuring reliable interaction between the cloud and the edge. Extensive experiments verified the effectiveness and superiority of the proposed method. Note to Practitioners—The continual emergence of new working conditions in actual industrial processes leads to serious model mismatch issues. However, resource limitations affect real-time updating and monitoring. This paper aims to address critical issues in practical industrial applications, particularly focusing on effective dynamic industrial processes in resource-constrained environments and ensuring data privacy protection within the industrial Internet. By combining cloud edge collaboration, dictionary self-updating and dictionary distillation compression technology, it can adapt to the complex and constantly changing operating environment while ensuring the monitoring accuracy. In addition, the cloud-edge data transmission protocol based on hybrid encryption effectively guarantees a secure data transmission process, thereby providing a more efficient and reliable monitoring solution for industrial production. Compared with the conventional method, the proposed method overcomes the inherent resource limitations on the edge devices and the information security problems between the cloud and edge. Overall, it is suitable for actual dynamic industrial systems.}, archive = {J_TASE}, author = {Keke Huang and Qinzhe Wang and Zixuan Chen and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2025.3557285}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14170--14182}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure cloud-edge collaborative method for dynamic industrial process monitoring using self-updating dictionary learning}, volume = {22}, year = {2025}, }
• (2025). Adaptive human movement compensation control of supernumerary robotic limb for overhead support task with non-zero-sum differential game theory. TASE, 22, 14158--14169. (www)

  @article{ , abstract = {The supernumerary robotic limb (SRL) mounted on the shoulder has been demonstrated to be able to serve as a third arm to assist human in overhead support task. However, the mechanical connection between the wearer and the SRL means that human operator’s movement will continually disturb the SRL and may lead to instability. Moreover, there may be physical conflicts between SRL and human operator due to their different intentions, which may potentially increase the load on human operator. Therefore, it’s necessary to control SRL to ensure stable supporting and transparent interaction. Here, we propose an adaptive controller for human movement compensation. Firstly, we model the human-SRL coordinative behavior based on non-zero-sum differential game theory aiming to enhance the support stability and reduce the operator’s load, which is a framework capable of dynamically regulating the control strategies between two interacting agents. We then implement an adaptive control strategy that adjusts the SRL’s input optimally responsive to the human operator’s input in the sense of Nash equilibrium to meet predefined control objectives. From experimental results during overhead support task, the proposed controller reduces the peak reaction force from 15.59 N to 4.50 N compared to the state-of-the-art controller that disregards human input while ensuring the stability of support, thereby providing advantage for human-SRL coordination in overhead support task. Note to Practitioners—To ensure optimal overhead assistance, the stability of the support and the safety of the human operator are of paramount importance. Therefore, this study investigates how to provide stable support and transparent interaction through movement compensation. Previous compensation methods lacked the understanding for the wearer and the optimal adjustment for the control of SRL. Consequently, this paper proposes a non-zero-sum differential game to model the desired behaviors between human operator and SRL. Furthermore, it solves for the optimal input to adaptively adjust the control gains of the SRL at the Nash equilibrium. The method proposed in this paper, compared to existing compensation methods, not only ensures the stability of the support but also enhances safety. Moreover, this approach avoids the need for expensive force sensing and precise human modeling, making it easy to implement and integrate into actual production environments.}, archive = {J_TASE}, author = {Jianxi Zhang and Hong Zeng and Jia Liu and Dapeng Chen and Aiguo Song}, doi = {10.1109/TASE.2025.3559522}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14158--14169}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive human movement compensation control of supernumerary robotic limb for overhead support task with non-zero-sum differential game theory}, volume = {22}, year = {2025}, }
• (2025). Bipartite output synchronization of fuzzy fractional output-coupled networks via membership function-dependent adaptive control. TASE, 22, 14147--14157. (www)

  @article{ , abstract = {This article focuses on the bipartite output synchronization for a type of fuzzy fractional output-coupled networks based on fractional-order fuzzy adaptive strategies. Firstly, in view of the unavailability of state information caused by irresistible factors and the coexistence of cooperative and competitive relations in reality, a class of T-S fuzzy fractional networks with output couplings is established under the signed graph framework. Next, a type of membership function-dependent fractional adaptive schemes is presented to automatically regulate the control gains, some criteria of bipartite output synchronization are derived based on the characteristic of the signed topology instead of traditional gauge transformation method. Particularly, a pinning adaptive control scheme is employed to investigate bipartite output synchronization for fuzzy fractional output-coupled networks with the positive-definite output matrix, which determines the pinning nodes just dependent of the cooperative links among nodes in signed graph. The developed criteria are finally confirmed by some examples. Note to Practitioners—Considering the unavailability of state information caused by irresistible factors and the coexistence of cooperative and competitive relations in real networked systems, the bipartite output synchronization issue of fuzzy fractional cooperative-antagonistic networks is concerned in this study, which has potential applications in opinion dynamics of social networks, bidirectional formation of uncrewed aerial vehicles. A type of membership function-dependent fractional adaptive scheme and its pinning form are developed to automatically regulate the control gains, which reflects the influence of fuzzy rules on control and improves the previous fuzzy-independent controller design. In theoretical analysis, based on the decomposition of Laplacian matrix of the signed graph, a direct analysis method is developed instead of traditional gauge transformation, which not only provides a new and more effective tool for the investigation of bipartite synchronization or consensus, but also weakens the requirement on topology connectivity. As a result, the proposed fuzzy-dependent adaptive controller and direct analysis technique have great theoretical values and wide practical engineering applications.}, archive = {J_TASE}, author = {Yu Sun and Cheng Hu and Juan Yu and Shiping Wen and Hong-Li Li}, doi = {10.1109/TASE.2025.3558624}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14147--14157}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bipartite output synchronization of fuzzy fractional output-coupled networks via membership function-dependent adaptive control}, volume = {22}, year = {2025}, }
• (2025). Non-cyclic scheduling of single-armed cluster tools for processing two wafer types. TASE, 22, 14133--14146. (www)

  @article{ , abstract = {Cluster tools are extensively used for processing wafers in semiconductor manufacturing. To enhance the throughput of cluster tools, concurrent processing of multiple types of wafers has become a popular approach. In practice, due to mass customization and the shrinking of chips’ feature sizes, the lot size of wafers tends to become smaller, leading to more frequent transient and non-cyclic periods. Therefore, the scheduling problem of single-armed cluster tools that concurrently process two types of wafers from start-up to close-down period should be urgently addressed. To do so, a mixed integer programming (MIP) model is established to optimize the release sequence of wafer types and robot task sequence so that the total completion time of the two wafer types can be minimized. Since solving the MIP model for large-sized instances is hard, a two-combined backward sequence (2-CBS) is further proposed. Experimental results show that the 2-CBS can save a significant amount of computing time and increase the total completion time by only 1.09% on average. However, by quickly determining the completion time of each wafer lot, the throughput of cluster tools could be enhanced to some extent, as the idle time caused by the arrival delay of overhead hoist transports can be reduced. Note to Practitioners—When considering the concurrent processing of multiple wafer types, almost all existing studies focus on cyclic scheduling and thus fail to address the growing importance of non-cyclic scenarios. To fill this gap, the present work develops an MIP model to optimize the release sequence of wafer types and robot task sequence such that the total completion time of two wafer types can be minimized. Such a model is efficient for small-sized instances. Meanwhile, a 2-CBS strategy is developed for large-sized instances, which can be solved quickly and results in an increase in the total completion time by only 1.09%, demonstrating its effectiveness and applicability.}, archive = {J_TASE}, author = {Fajun Yang and Xiang Xu and Zhiyuan Yang and Chao Li and Lu Zhen and Naiqi Wu}, doi = {10.1109/TASE.2025.3554273}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14133--14146}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Non-cyclic scheduling of single-armed cluster tools for processing two wafer types}, volume = {22}, year = {2025}, }
• (2025). A three-step optimization framework with hybrid models for a humanoid robot’s jump motion. TASE, 22, 14120--14132. (www)

  @article{ , abstract = {High dynamic jump motions are challenging tasks for humanoid robots to achieve environment adaptation and obstacle crossing. The trajectory optimization is a practical method to achieve high-dynamic and explosive jumping. This paper proposes a 3-step trajectory optimization framework for generating a jump motion for a humanoid robot. To improve iteration speed and achieve ideal performance, the framework comprises three sub-optimizations. The first optimization incorporates momentum, inertia, and center of pressure (CoP), treating the robot as a static reaction momentum pendulum (SRMP) model to generate corresponding trajectories. The second optimization maps these trajectories to joint space using effective Quadratic Programming (QP) solvers. Finally, the third optimization generates whole-body joint trajectories utilizing trajectories generated by previous parts. With the combined consideration of momentum and inertia, the robot achieves agile forward jump motions. A simulation and experiments of forward jump with a distance of 1.0 m and 0.5 m height are presented in this paper, validating the applicability of the proposed framework. Note to Practitioners—The motivation of this paper stems from the need to improve jumping performance of humanoid robots. By comprehensively considering factors such as robot posture, centroidal angular momentum, and landing foot placement, the algorithm enhances the robot’s ability to navigate complex environments. This capability is crucial for applications that require overcoming obstacles, such as in search and rescue or inspection tasks. Improved jumping ability can significantly boost environmental adaptability, allowing robots to perform effectively in diverse conditions, and it also represents an exploration of the high-dynamic motion capabilities of humanoid robots. Future research will focus on integrating visual and perceptual information to enhance decision-making.}, archive = {J_TASE}, author = {Haoxiang Qi and Zhangguo Yu and Xuechao Chen and Qingqing Li and Yaliang Liu and Chuanku Yi and Chencheng Dong and Fei Meng and Qiang Huang}, doi = {10.1109/TASE.2025.3557162}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14120--14132}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A three-step optimization framework with hybrid models for a humanoid robot’s jump motion}, volume = {22}, year = {2025}, }
• (2025). Rolling dispatch for AAVs inspection based on task adaptive clustering. TASE, 22, 14108--14119. (www)

  @article{ , abstract = {The assignment and scheduling of AAV power inspection tasks constitute a representative Mixed-Integer Nonlinear Programming (MINLP) problem. However, due to the complexity of the workflow, manual methods and existing heuristics struggle to balance time costs and solution quality for large-scale tasks. To address this challenge, this paper introduces a rolling window approach for batch processing and proposes a clustering strategy before scheduling. Building on this strategy, we develop a rolling dispatch algorithm utilizing task-adaptive clustering (RDA-TAC), which alleviates problem complexity and enhances search efficiency. Specifically, we present a novel MINLP model designed to minimize total equivalent endurance losses, incorporating multi-level and sequential inspection constraints derived from actual inspection processes. We then batch and segment large-scale tasks, treating task fragments as the smallest units for assignment, and adaptively adjust them to convert the problem into a single-drone path planning task. Through nine scenarios based on real power grid data, RDA-TAC attains optimal solutions for small-scale tasks and outperforms manual and heuristic methods in more complex cases, demonstrating a tenfold increase in solving speed. Note to Practitioners—Autonomous aerial vehicles (AAVs) are the leading method for power line inspection, offering rapid deployment, cost-effectiveness, and improved safety. However, traditional manual assignment and path planning algorithms often struggle with large-scale and complex tasks. To address this, we propose a strategy that clusters tasks before scheduling, reducing complexity and improving efficiency. Empirical validation demonstrates that our method balances solution time and quality effectively. Comparative results are included in the attached digital format.}, archive = {J_TASE}, author = {Jize Ren and Nanfeng Song and Xian Li and Lei Wang}, doi = {10.1109/TASE.2025.3553745}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14108--14119}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rolling dispatch for AAVs inspection based on task adaptive clustering}, volume = {22}, year = {2025}, }
• (2025). Observer-based h∞ fault-tolerant tracking control of multi-agent systems with nonideal communication links and external disturbances. TASE, 22, 14096--14107. (www)

  @article{ , abstract = {This paper studies the problem of robust $H_{\infty } $ active fault-tolerant tracking control for multi-agent systems (MASs) under nonideal communication links and external disturbances. Unlike most existing studies that assume idealized or simplified communication links, such as uniform constant or time-varying delays, this paper considers a more realistic communication scenario involving non-uniform time-varying communication delays. The communication cases discussed encompass existing common communication scenarios as special cases. To achieve active fault-tolerant control, an intermediate variable-assisted proportional-integral observer (PIO) is introduced to estimate the system’s state and faults. Based on fault estimation values and considering non-ideal communication constraints, a fault-tolerant control scheme is designed to achieve the leader-follower tracking control objectives. With the help of Lyapunov theory and matrix inequality method, sufficient criteria are proposed for co-designing the observer and controller gains. Finally, two simulation examples verify the effectiveness of the theoretical results. Note to Practitioners—Multi-agent systems have been widely applied in fields such as autonomous vehicle platooning, distributed robotics, and smart grids. Their exceptional coordination capabilities and flexibility make them an ideal choice for performing complex tasks. However, these systems rely on network communications and often face communication delays and external disturbances, posing significant challenges to the control system, especially when a fault occurs that may cause the system to lose stability. Therefore, it is particularly important to carry out active fault-tolerant control in the case of nonideal communication conditions, which can maintain the normal operation of the system and provide compensation in time when faults occur to ensure the robustness and stability of the system. The observer-based active fault-tolerant control method designed in this article offers a practical solution for multi-agent systems that require high-reliability control in complex communication environments.}, archive = {J_TASE}, author = {Huichao Lin and Jiuxiang Dong and Ju H. Park}, doi = {10.1109/TASE.2025.3557951}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14096--14107}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based h∞ fault-tolerant tracking control of multi-agent systems with nonideal communication links and external disturbances}, volume = {22}, year = {2025}, }
• (2025). Closed-loop placement planning for regrasping and reconstruction with single-view RGB-D images. TASE, 22, 14084--14095. (www)

  @article{ , abstract = {In robotic reorientation, achieving certain positions or orientations within a workspace can be difficult due to collisions and interference between objects and the environment. Regrasping is a technology that overcomes limitations on object reorientation using multiple pick-and-place actions. However, constructing a regrasping graph, which consists of predicted intermediate placements, is challenging with only partial observations of unknown objects. To address this challenge, we design a closed-loop placement planning system using single-view RGB-D images. This system seamlessly integrates perception and manipulation processes, capturing RGB-D images for object reconstruction while predicting intermediate placements for constructing regrasping graphs simultaneously. To optimize the regrasping graph construction, we propose a Truncated Signed Distance Function (TSDF)-based placement evaluation algorithm that selects the next-best-placement by estimating the information gain and retention of the placements. Additionally, we adopt a novel data fusion method based on predicted intermediate placements to enhance object reconstruction and TSDF updates, effectively mitigating in-hand errors. Our system surpasses baselines in accuracy and diversity of placement predictions. It demonstrates superior performance in object reconstruction, excelling in quantitative analysis and visualization. Real-world experiments validate the effectiveness of our system in reorientation and reconstruction tasks on real industrial parts, demonstrating its robustness and applicability in practical scenarios. Note to Practitioners—This paper was motivated by the problem of object reorientation in robot assembly tasks in the industrial sector. Robots often struggle to reorient objects to the desired orientation through a single pick-and-place due to workspace limitations. Our solution involves capturing RGB-D images of objects and implementing multiple pick-and-place steps, known as regrasping, to reorient the objects effectively. During this process, the point clouds of the objects are reconstructed. A key innovation of our approach is the method for selecting the next-best-placement for each step. This is achieved by estimating the information gain and retention and fusing data to improve the accuracy of both regrasping and reconstruction. Our approach has shown superior performance in both simulated and real-world experiments with industrial parts. Practitioners can use this system to enhance the efficiency and reliability of robotic reorientation tasks. Its practical value lies in its ability to manage complex practical scenarios, making it a valuable tool for automation processes. The system’s dependence on predefined parameters restricts its adaptability to dynamic tasks. Future work will explore the integration of multi-modal language models to improve adaptability and accuracy in handling diverse scenarios.}, archive = {J_TASE}, author = {Zhiyuan Chen and Jiangshan Liu and Ronghao Chen and Jiankun Wang}, doi = {10.1109/TASE.2025.3558473}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14084--14095}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Closed-loop placement planning for regrasping and reconstruction with single-view RGB-D images}, volume = {22}, year = {2025}, }
• (2025). Practical prescribed-time control for underactuated marine surface vessels: Theory and experiment. TASE, 22, 14074--14083. (www)

  @article{ , abstract = {This paper proposes a practical prescribed-time control strategy for underactuated marine surface vessels (UMSVs) with fore-aft asymmetry. A sufficient condition for practical prescribed-time stability (PPTS) is established through a novel time-varying gain function that maps the settling time into a single user-defined parameter. First, the nonholonomic constraint of the UMSV is transformed into an integral cascade system via the hand position approach, which addresses the design challenge from underactuation. Subsequently, input saturation nonlinearity is approximated by a sigmoid function, while model uncertainties are compensated through an adaptive parametric design paradigm. A barrier function-integrated adaptive law is further developed to bound the adaptive parameter without overestimation. Rigorous stability analysis confirms that the tracking error converges to an adjustable bounded domain within the prescribed time. Experimental results validate the efficacy of the proposed controller. Note to Practitioners—Controlling underactuated marine surface vessels (UMSVs) with asymmetric hull designs is challenging for time-sensitive operations such as autonomous navigation. Existing methods often require manual tuning to meet strict deadlines, which limits their reliability. To address this issue, we propose a practical prescribed-time control strategy that guarantees the tracking error converges to a tunable region within a user-defined time. By introducing a practical time-varying gain function, the controller achieves precise control over the settling time, thereby enhancing better reliability for real-time applications. This approach improves the efficiency and performance of UMSVs in dynamic environments. It can be extended to other autonomous systems requiring time-critical control. The experimental dataset is available at: https://github.com/zdh58324488/Experimental-Dataset-for-UMSV-Control}, archive = {J_TASE}, author = {Daohui Zeng and Chengtao Cai and Yongchao Liu and Jie Zhao and Chenming Li}, doi = {10.1109/TASE.2025.3557510}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14074--14083}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical prescribed-time control for underactuated marine surface vessels: Theory and experiment}, volume = {22}, year = {2025}, }
• (2025). Adaptive terminal control of n-DOF robot manipulators with enhanced nonlinear model-based synergetic manifold. TASE, 22, 14062--14073. (www)

  @article{ , abstract = {This paper presents an Adaptive Terminal Model-Based Synergetic Control (ATMB-SC) framework for n-DOF robot manipulators, addressing critical limitations of conventional synergetic control. The proposed strategy integrates a nonlinear model-based manifold function for robotic manipulators, ensuring robust trajectory tracking while eliminating the coupling effect inherent in the conventional synergetic control method. By establishing a linear relationship between the torque input and both position error and its rate of change, the method simplifies control design without compromising robustness or accuracy. An adaptive strategy based on the MIT rule dynamically adjusts the convergence matrix in real-time is implemented, enhancing precision and responsiveness to uncertainties and external disturbances. Additionally, a terminal term is incorporated into the manifold, guaranteeing finite-time convergence of error trajectories, which is critical for high-precision, time-sensitive applications. Experimental validation on a 4-DOF robot demonstrates the strategy’s effectiveness, with input torque variation improved by 10% to 112.67% and the least mean square error achieved across all axes, ranging from $1.63e-4$ to $9.15e-5$ , compared to other controllers. Note to Practitioners—This paper addresses the challenges faced by practicing engineers in the field of robot manipulators, particularly those working in industrial automation and robotics research, such as assembly line automation, material handling, inspection and quality control, etc. The proposed control strategy, known as adaptive terminal model-based synergetic control (ATMB-SC), can be effectively used to enhance the performance of robot manipulator systems, ensuring precise trajectory tracking and efficient energy consumption. The proposed controller integrates a nonlinear manifold function with an adaptive convergence matrix to adjust the real-time performance metrics and reduce the effect of external disturbances and uncertainties. Compared to traditional synergetic control, the proposed strategy significantly improves control precision, reliability, and robustness in unpredictable environments such as automotive manufacturing, aerospace component assembly, and pharmaceutical packaging. Experimental validation using a 4-DOF robotic manipulator demonstrates superior energy efficiency and trajectory precision compared to conventional synergetic control methods. This makes the ATMB-SC method ideal for industrial applications requiring high precision and adaptability under varying conditions.}, archive = {J_TASE}, author = {Mohannad Nasir and Raouf Fareh and Saif Sinan and Sofian Khadraoui and Mammar Bettayeb}, doi = {10.1109/TASE.2025.3558961}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14062--14073}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive terminal control of n-DOF robot manipulators with enhanced nonlinear model-based synergetic manifold}, volume = {22}, year = {2025}, }
• (2025). ESPPNet: An efficient progressive spatial pyramid pooling network for real-time traffic object detection. TASE, 22, 14048--14061. (www)

  @article{ , abstract = {Traffic object detection based on computer vision (CV) can usually be deployed on the embedded computing platform of autonomous vehicles or autonomous aerial vehicles (AAVs), to provide critical information about traffic scenes for autonomous driving or traffic management. However, due to limited computing resources, there is a need for small, lightweight, and reliable object detectors. As an emerging technology, spatial pyramid pooling methods have great potential in improving the detection performance of real-time object detectors. Most of the existing works focus on the development of more complex spatial pyramid pooling methods for higher accuracy, but real-time performance is also important in the ever-changing traffic scene. Thus, to balance the tradeoff between real-time detection and accuracy, we design a solution for real-time traffic object detection: a novel real-time object detector, named ESPPNet. Specifically, we propose an efficient plug-and-play spatial pyramid pooling method (ESPP). The method consists of a progressive spatial pyramid pool structure (PSPP) and a multi-scale feature enhancement module (MFEM). We first use PSPP to capture multi-scale feature maps with richer nonlinear features. Then, MFEM is used to establish effective long-range dependencies for multi-scale features. Experimental results on the VisDrone and SODA10M public datasets demonstrate that our method can achieve better real-time performance, less resource utilization, and higher accuracy, compared with other state-of-the-art methods. Note to Practitioners—This work aims to develop a practical and efficient object detection system for deployment in resource-constrained autonomous vehicles tasked with performing traffic object detection. Traffic object detection technology plays a crucial role in providing environmental perception capabilities for autonomous vehicles and autonomous aerial vehicles. It also serves as an essential means for traffic managers to acquire road information. However, existing object detectors generally fall short of capturing several crucial aspects, such as the interactions between different depth features or the independence of features of different receptive fields. This can lead to a bottleneck in feature extraction capabilities. To address these issues, our work proposes a plug-and-play lightweight module to support real-time traffic object detection on both mobile CPU and GPU, enabling the design of an efficient and practical end-to-end object detector. We have validated the performance of our proposed method on multiple traffic scene datasets.}, archive = {J_TASE}, author = {Guotao Mao and Hongbin Liang and Yiting Yao and Lei Wang and Ning Zhang}, doi = {10.1109/TASE.2025.3558929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14048--14061}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ESPPNet: An efficient progressive spatial pyramid pooling network for real-time traffic object detection}, volume = {22}, year = {2025}, }
• (2025). Task allocation of heterogeneous robots under temporal logic specifications with inter-task constraints and variable capabilities. TASE, 22, 14030--14047. (www)

  @article{ , abstract = {Multi-Robot task allocation (MRTA) exploits different capabilities of heterogeneous robots to facilitate collaborative tasks. However, existing works are mainly built on a key assumption that the robot capabilities are invariant and few consider variable capabilities (e.g., task-dependent or time-dependent capabilities). Besides, there may also exist a variety of inter-task constraints (e.g., unrelated tasks, compatible tasks, and exclusive tasks). Motivated by this practical need, we develop a novel task allocation framework for heterogeneous multi-robot systems with variable capabilities subject to inter-task constraints and temporal logic task specifications. Specifically, we extend conventional linear temporal logic (LTL) to capability LTL, namely $\boldsymbol {\mathrm {CaLT}\mathrm {L}^{\mathcal {T}}}$ , to describe heterogeneous multi-robots systems with variable capabilities and inter-task constraints. The Task Batch Planning Decision Tree Plus (TB- $\boldsymbol {\mathrm {PDT^{+}}}$ ) is then developed, which encodes the states of Büchi automaton, the system states, and the task process into a tree structure to represent the exploration progress. Based on the TB- $\boldsymbol {\mathrm {PDT^{+}}}$ , the Variable Capability and Inter-task Constraints Search (Var-CICS) is developed to find feasible task allocations and plans. Rigorous analysis shows that Var-CICS is valid (i.e., the generated task allocation is guaranteed to satisfy the task requirements) and complete (i.e., if a feasible task allocation exists, it is ensured to be found by Var-CICS). The complexity analysis also shows that the computation time of finding a satisfactory task allocation scales only quadratically with the number of automaton states, versus the exponential growth due to the product automaton in standard model checking methods. Numerical simulations and experiments demonstrate the effectiveness of Var-CICS. Note to Practitioners—Real-world applications often require a heterogeneous multi-robot system working collaboratively on a variety of tasks. Within such applications, robots can have diverse capabilities which may vary depending on the task at hand or over time, and are subject to inter-task constraints. Thus, in this work, we propose a new temporal logic to enrich the expressiveness in describing heterogeneous multi-robots systems with variable capabilities and inter-task constraints. We then develop the task batch planning decision tree plus and the variable capability and inter-task constraints search (Var-CICS) for the task allocation of heterogeneous multi-robot system. In contrast to automata-based methods, our method does not require sophisticated product automaton, which enables efficient and effective search of feasible task allocations. Furthermore, theoretical analyses show that Var-CICS is both complete and valid, while experimental results demonstrate its validity, efficiency and scalability.}, archive = {J_TASE}, author = {Lin Li and Ziyang Chen and Hao Wang and Zhen Kan}, doi = {10.1109/TASE.2025.3558977}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14030--14047}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Task allocation of heterogeneous robots under temporal logic specifications with inter-task constraints and variable capabilities}, volume = {22}, year = {2025}, }
• (2025). Coprime factorization-based encryption and attack detection for nonlinear cyber-physical systems using deep learning approach. TASE, 22, 14020--14029. (www)

  @article{ , abstract = {This paper presents a data-driven framework for integrating encryption transmission and attack detection in cyber-physical systems (CPS) with nonlinear physical plants. The main focus of this research is to use deep neural networks to realize the coprime factorization (CF) of nonlinear systems. The definition of the CF guides the network training and designing process, and the model’s topology is designed in the state-space form, which improves the interpretability of the data-driven CF. Based on the CF-aided neural networks, an encrypted transmission module is designed that projects information related to system dynamics into a perpendicular data space, which complements existing encryption methods from a control theory perspective. Subsequently, an anomaly detector are designed using the same CF pairs. This detector not only provides high-accuracy detection of attacks but also distinguishes between attacks and faults, thereby reducing the false positive rate and enhancing the reliability of the attack detection. The proposed method has been validated in a real CPS using a mecanum-wheeled vehicle as the physical plant, demonstrating its effectiveness and applicability. Note to Practitioners—Given new security challenges posed by eavesdropping and stealthy attacks in CPS, this paper presents a data-driven framework that integrates encrypted transmission and attack detection for nonlinear CPS based solely on input and output data. The CF-aided neural networks enable an encrypted transmission module that maintains accuracy while projecting system dynamics into a perpendicular data space, which enhances existing encryption from a control theory perspective. Furthermore, the framework includes an attack detection mechanism that effectively differentiates between attacks and system faults, which is rarely considered in existing work. The framework has been validated in a real-world CPS application using a mecanum-wheeled vehicle, showcasing its practical applicability. Its data-driven and model-agnostic design makes it adaptable to various critical CPS applications. For instance, the proposed deep learning-based method shows promise in smart grid systems, effectively addressing complex sensor signal processing challenges. Additionally, the experimental validation provides a foundational reference for extending this methodology to other unmanned vehicle systems. However, implementation may require specific computational resources and deep learning expertise, which could hinder immediate adoption. Further research is needed to explore its scalability across different CPS applications, focusing on optimizing computational efficiency and adapting to domain-specific requirements.}, archive = {J_TASE}, author = {Shimeng Wu and Hao Luo and Jiusi Zhang and Xinyu Qiao and Jilun Tian and Yuchen Jiang}, doi = {10.1109/TASE.2025.3557962}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14020--14029}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Coprime factorization-based encryption and attack detection for nonlinear cyber-physical systems using deep learning approach}, volume = {22}, year = {2025}, }
• (2025). Physical intrusion attack detection in fieldbus network with passive fail-safe biasing. TASE, 22, 14008--14019. (www)

  @article{ , abstract = {Fieldbus is widely used for real-time distributed control in Industrial Control Systems (ICSs) due to its simplicity and stability. The real-world fieldbus network contains hundreds of interconnected devices, presenting a widespread network layout. Attackers can attach external intrusion devices to these communication lines to launch various attacks. In this paper, we model the fieldbus network’s channel fingerprint based on the signal’s amplitude and propose a detection method to identify potential attackers (silent intrusion devices that are eavesdropping) via channel fingerprint differences. Leveraging the passive fail-safe biasing voltage in the fieldbus network such as RS485, we can still detect the intrusion device when the fieldbus is idle (i.e., no devices are transmitting commands), which can significantly reduce the detection delay with lower sampling costs. Moreover, our method can adapt to environmental changes with little computational overhead by generating dynamic thresholds. Using a monitoring unit with stored channel fingerprints, our method can be easily deployed in fieldbus networks without occupying communication resources. The effectiveness and robustness of the proposed method have been demonstrated via extensive experiments on two real-world scenarios and one simulation scenario, where we can achieve 100% accuracy and 0% false alarm rates against various intrusion devices. Note to Practitioners—This paper is motivated by a practical need for detecting unauthorized intrusion devices in fieldbus networks. Existing detection methods face several challenges: active detection methods based on traffic analysis may disrupt normal bus communication, and it is hard to identify silent intrusion devices that are eavesdropping. Moreover, adapting these methods to changing environments is still challenging and costly. To address these issues, we leverage the inevitable amplitude differences in fail-safe biasing voltage signals and benign devices’ communication voltage signals to detect intrusion devices passively. Furthermore, to adapt to rapidly changing environments, we generate the detection thresholds dynamically based on the hypothesis testing theory. Extensive physical and simulation experiments demonstrate that the detection method against physical intrusion attacks is accurate and robust.}, archive = {J_TASE}, author = {Xiangming Wang and Yang Liu and Nanpeng Yu and Pengfei Liu and Nanyi Deng and Yanan Li and Ting Liu}, doi = {10.1109/TASE.2025.3556720}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {14008--14019}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Physical intrusion attack detection in fieldbus network with passive fail-safe biasing}, volume = {22}, year = {2025}, }
• (2025). Appointed-time fuzzy fault-tolerant control of hypersonic flight vehicles with flexible predefined behaviors. TASE, 22, 13995--14007. (www)

  @article{ , abstract = {This paper explores a novel fuzzy fault-tolerant control scheme for hypersonic flight vehicles (HFVs) that accounts for parameter perturbations and system disturbances, thereby achieving flexible prescribed properties and designated-time convergence. In contrast to existing studies, this framework incorporates the more detrimental elevator stuck fault within the context of prescribed performance control (PPC). To mitigate the fragility issues associated with PPC in fault-tolerant control, we propose an innovative approach that integrates a fault sensing system along with a readjustment prescribed envelope. This is accomplished by transforming the original HFVs model into an imprecise pure feedback model, allowing for the design of low-complexity fuzzy controllers with minimal model dependence through the utilization of higher-order error functions and fuzzy logic systems. Finally, numerical simulations validate both the effectiveness and superiority of the proposed method. Note to Practitioners—The motivation for this article arises from the need for fault-tolerant control in the context of the more challenging actuator stuck faults, taking into account model parameter perturbations and disturbances. The PPC framework ensures good transient performance of the controlled system; however, it also introduces fragility issues. To address the potential fragility issue in PPC-based fault-tolerant control, we propose a flexible prescribed performance control scheme that effectively senses faults and appropriately scales the prescribed performance envelope, thereby circumventing control singularity. In consideration of the real-time control, the methodology in this paper skillfully transforms the altitude subsystem of HFVs into a pure feedback form and constructs a higher-order error function, thus resulting in a concise controller structure. In addition, the minimum learning parameter technique is employed to further reduce the online learning burden. The results presented in this paper offer significant insights into addressing the fragility issues associated with PPC and extending the PPC-based control strategy to the domain of fault-tolerant control.}, archive = {J_TASE}, author = {Ruining Luo and Guangjun He and Yulun Li and Xiangwei Bu and Qiuni Li}, doi = {10.1109/TASE.2025.3556989}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13995--14007}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Appointed-time fuzzy fault-tolerant control of hypersonic flight vehicles with flexible predefined behaviors}, volume = {22}, year = {2025}, }
• (2025). Event-based probability-guaranteed set-membership secure fusion estimation for energy-constrained multi-sensor systems with asynchronous samplings. TASE, 22, 13983--13994. (www)

  @article{ , abstract = {This paper addresses the problem of designing probability-guaranteed set-membership secure estimation algorithms for energy-constrained multi-sensor systems with multi-rate asynchronous samplings. An event-triggered strategy (ETS) is employed to minimize data transmission overhead while maintaining estimation accuracy by transmitting only essential data. A novel measurement model is proposed to accurately characterize the operation of the multi-sensor system under ETS, taking into account both high- and low-energy transmission (HLET) modes and random denial-of-service (DoS) attacks, which impact communication energy consumption and data security. To cope with the challenges posed by uncertain sampling periods, a new fusion estimation model is established, including a redefined fusion estimation weight matrix and the formulation of a probability-guaranteed set-membership secure fusion estimation algorithm. Furthermore, a recursive optimization algorithm based on linear matrix inequalities is utilized to determine the minimum ellipsoid of the design parameters. The effectiveness of the proposed algorithm is validated through simulation studies. Note to Practitioners—This paper is dedicated to addressing the state estimation problem in practical applications such as autonomous systems, smart grids, and industrial automation. In these fields, ensuring that the true state remains bounded within a certain probability is crucial, especially in applications such as weapon firing tests and drone flights. To meet this need, probability-guaranteed set-membership filtering is introduced, which not only provides state estimation intervals but also ensures that these intervals contain the true state with a certain probability. However, in practical systems, sensor sampling rates are typically asynchronous, which increases the complexity of the design algorithm. Moreover, wireless networks often face bandwidth limitations, and different communication protocols have varying energy consumption, which affect the security and reliability of data transmission. When subjected to cyber-attacks, these varying energy consumption protocols can introduce different security risks, further complicating system performance. Therefore, this paper introduces an ETS to reduce communication load, while designing a fusion estimation algorithm that accounts for both HLET modes and DoS attacks on communication channels. Preliminary simulation results demonstrate the feasibility of the proposed algorithm. Considering the potential complexities such as hybrid attacks, incomplete measurements, and time-delays in real-world systems, future research will focus on improving the interference resilience and robustness of the fusion estimation system, ensuring efficient and accurate state estimation even under harsh conditions.}, archive = {J_TASE}, author = {Haiyu Song and Meichen Lai and Zhen Hong and Bo Chen and Wen-An Zhang and Li Yu}, doi = {10.1109/TASE.2025.3558663}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13983--13994}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based probability-guaranteed set-membership secure fusion estimation for energy-constrained multi-sensor systems with asynchronous samplings}, volume = {22}, year = {2025}, }
• (2025). A physics-enhanced separation framework for spatiotemporal modeling of distributed parameter systems with multi-fidelity data. TASE, 22, 13968--13982. (www)

  @article{ , abstract = {Many industrial processes are typically distributed parameter systems (DPSs) described by partial differential equations. Data-driven methods have become popular for spatiotemporal modeling of DPSs, which is crucial for system understanding, simulation, and control improvement. However, current data-driven methods rely heavily on data volume and fidelity. With limited high-fidelity data, they exhibit unsatisfactory long-term predictions for out-of-sample scenarios. To remedy this issue, a novel physics-enhanced separation framework (called PhysiT/S) is proposed. PhysiT/S is composed of a physics-enhanced spatiotemporal unit and a coefficient network. By enhancing physics utilization through the physics-enhanced spatiotemporal unit, PhysiT/S becomes less data-dependent while computationally efficient. Multi-fidelity learning of the coefficient network further alleviates the request for a large volume of high-fidelity data. As a result, PhysiT/S can accurately predict out-of-sample distributions, even when only limited high-fidelity data is available. PhysiT/S introduces a novel way of combining physics with multi-fidelity data for spatiotemporal modeling of DPSs. Extensive simulations on two benchmark DPSs and the thermal process of lithium-ion batteries demonstrate the merits of PhysiT/S. Note to Practitioners—This paper is motivated by the problem of data-driven spatiotemporal modeling of DPSs, which is also applicable to other prediction tasks for complex spatiotemporal systems. Existing data-driven approaches rely heavily on high-fidelity data, while physics-informed approaches struggle to extract coupled spatiotemporal features. These drawbacks limit their practical applications to engineering modeling. In this paper, we propose a new generic modeling framework that integrates physics with data through the theory of time/space separation. It aims to achieve accurate long-term predictions for out-of-sample scenarios by constructing spatiotemporal enhancement units and using multi-fidelity data to decrease reliance on high-fidelity datasets. We have applied the proposed method to the thermal processes of a catalytic rod and a cylindrical lithium-ion battery. Preliminary results show that this method is feasible with better predictions than others, but it is untested in production. Future research will address online modeling in practical engineering.}, archive = {J_TASE}, author = {Bing-Chuan Wang and Yan-Bo He and Yong Wang and Han-Xiong Li}, doi = {10.1109/TASE.2025.3549932}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13968--13982}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A physics-enhanced separation framework for spatiotemporal modeling of distributed parameter systems with multi-fidelity data}, volume = {22}, year = {2025}, }
• (2025). Air shepherd: Trajectory prediction-based target localization and circumnavigation in cluttered environments. TASE, 22, 13953--13967. (www)

  @article{ , abstract = {This paper proposes a trajectory prediction-based target localization and circumnavigation pattern for cluttered three-dimensional environments, which is more realistic and suitable for more complex environments than traditional patterns. The main work of the paper consists of two parts: tracking based on trajectory prediction and circumnavigation based on broadcast information. On the one hand, the tracking Autonomous Aerial vehicle (AAV) obtains target trajectory prediction based on the B-spline curve, and then achieves target localization and tracking through front-end search and back-end optimization. On the other hand, without communicating with each other, a distributed control strategy is presented so that the multiple circumnavigation AAVs can achieve target circumnavigation and reciprocal avoidance by only observing the status of adjacent AAVs. In the simulation, obstacle avoidance vehicles moving freely at different speeds are selected as targets in two scenarios and the simulation results are given to verify the effectiveness of the proposed approach. Furthermore, a hardware-in-the-loop experiment and a overall system validation experiment are designed to verify the feasibility of the algorithm. Note to Practitioners—Practical tracking scenarios often involve numerous obstacles, and it is difficult to determine the location of the target being tracked and surrounded. Many existing methods assume that the trajectory or position of the tracked target is already known, but in reality, this assumption is difficult to fulfill due to the agility of the target and the complexity of the environment. Compared to existing work, this paper proposes an ‘air shepherd’ mechanism, in which a tracking AAV plays the role of a ‘sheepdog’, and based on trajectory prediction, the target can still be tracked even when it is out of sight. The circumnavigation AAVs play the role of the ‘sheep’, achieving pursuit and encirclement without requiring communication between each other while ensuring reciprocal avoidance.}, archive = {J_TASE}, author = {Kai Rao and Huaicheng Yan and Hongliang Ren and Tan Chen and Youmin Zhang}, doi = {10.1109/TASE.2025.3557701}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13953--13967}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Air shepherd: Trajectory prediction-based target localization and circumnavigation in cluttered environments}, volume = {22}, year = {2025}, }
• (2025). Real-time smooth trajectory generation for five-axis hybrid machining robots with velocity profile blending. TASE, 22, 13932--13952. (www)

  @article{ , abstract = {Five-axis toolpaths are typically formatted as G01 commands by breaking curves into linear segments before being inputted into the computer numerical control systems. Despite the availability of various trajectory generation methods, real-time generation of smooth trajectories with limited smoothing error and constrained high-order kinematics remains a challenge. This paper proposes a velocity profile blending-based interpolation method for G01 paths of five-axis hybrid robots to obtain jerk-limited trajectories with constrained smoothing error. Each linear segment is represented by a 13-phase velocity profile instead of the widely adopted S-shape curve. The velocity profiles are locally blended to construct a smooth trajectory. For adjacent velocity profiles, the first three phases of the following profile strictly align with the last three phases of the preceding one. In the implementation, a bidirectional scanning algorithm is adopted to generate the velocity profile of each linear segment with the time constants of the overlapping phases pre-optimized. In the simulations and experiments, the effectiveness and benefits of the proposed method on G01 paths comprised of both long segments and short segments are validated through a comparative analysis with several representative trajectory generation methods. Note to Practitioners—This work aims to generate error-constrained and jerk-limited trajectories with high machining efficiency from G01 commands in real time for five-axis hybrid robots. Computer-aided manufacturing software typically exports five-axis toolpaths as G01 commands. Next, computer numerical control systems read the G01 commands and produce trajectories with limited error and constrained kinematics as reference commands to the drivers. To obtain smooth trajectories from G01 commands for five-axis mechanisms more efficiently, one-step trajectory generation methods based on finite impulse response (FIR) filtering are increasingly studied. Nevertheless, due to their simplistic velocity profiles, existing FIR filtering-based one-step trajectory generation approaches struggle to simultaneously enhance machining efficiency, limit blending errors, and constrain the mechanism’s high-order kinematics. Thus, we propose a more general velocity profile blending-based interpolation method that expresses the movements along linear segments as 13-phase velocity profiles rather than the S-shape velocity profiles generated by the FIR filters. Accordingly, a novel strategy is developed to efficiently determine the velocity profiles. The proposed method, implemented on a specific hybrid machining robot, was compared with a spline-based trajectory generation approach and three representative FIR filtering-based approaches. Simulations and experiments were performed, in which G01 paths constituted by both long and short segments were used to validate our method. The results suggest that our method offers several advantages at the same time: low computational cost, significant improvements in machining efficiency, strictly constrained blending error, and precisely limited robot kinematics. Furthermore, this method does not make any assumptions about the structure of the five-axis mechanism, allowing for its application to general five-axis mechanisms. In industrial applications, the proposed method can be embedded in the computer numerical control system for real-time implementation.}, archive = {J_TASE}, author = {Weijia Zhang and Zikang Shi and Xinxue Chai and Ye Ding}, doi = {10.1109/TASE.2025.3556759}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13932--13952}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time smooth trajectory generation for five-axis hybrid machining robots with velocity profile blending}, volume = {22}, year = {2025}, }
• (2025). Lag output consensus for second-order nonlinear multiagent systems under PID control. TASE, 22, 13920--13931. (www)

  @article{ , abstract = {Based on PD and PI control methods, the lag output consensus problem for second-order nonlinear multiagent systems (MASs) without and with external disturbances is addressed, which generalizes the existing lag consensus results for MASs. Firstly, a lag output consensus condition is derived for the MAS by designing a proper PD control scheme. Furthermore, a PI control strategy is developed to ensure that the MAS realizes the lag output consensus. In addition, two lag ${\mathcal {H}}_{\infty } $ output consensus criteria are also established on the basis of PD and PI control schemes for the MAS with external disturbances. Lastly, two examples are utilized to verify the correctness and effectiveness of the devised control schemes. Note to Practitioners—Recently, virtual leader-based approach has been widely utilized to address the output consensus control problem for various MASs. Practically, each agent receiving its own output information may be later than the virtual leader. Therefore, it is more reasonable and meaningful to discuss the lag output consensus for MASs. Furthermore, considering that PID control is very simple and easy to implement, some PD and PI control strategies are developed to tackle the lag output consensus for MASs. In addition, since external disturbances are ubiquitous and may destroy the lag output consensus, the lag ${\mathcal {H}}_{\infty } $ output consensus for MASs is also investigated. Multiple single-link robot systems experiments verify that the approaches given in this paper are feasible and valid.}, archive = {J_TASE}, author = {Jin-Liang Wang and Xiao Han and Shun-Yan Ren and Tingwen Huang}, doi = {10.1109/TASE.2025.3551091}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13920--13931}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lag output consensus for second-order nonlinear multiagent systems under PID control}, volume = {22}, year = {2025}, }
• (2025). Dynamic disassembly planning of end-of-life products for Human–Robot collaboration enabled by multi-agent deep reinforcement learning. TASE, 22, 13907--13919. (www)

  @article{ , abstract = {Disassembly is a critical step in the remanufacturing of end-of-life products. High labor costs and the limited ability of robots to perform intricate disassembly tasks have led to the increasing use of human-robot collaboration (HRC) for disassembly. This paper addresses a challenge in HRC-based disassembly, i.e., the inherent human uncertainty during disassembly. The uncertainty is that the disassembly time for a task and the task sequence selection by a human during execution might differ from the pre-defined disassembly plan so that dynamic disassembly planning for subsequent tasks is necessary. Stackelberg equilibrium-enabled disassembly task assignment policies are designed to meet the above purpose efficiently and safely. The human leader’s policy is to choose tasks that maximize the efficiency-related return value based on the robot’s optimal response to the human’s choice. As the follower, the robot selects tasks that maximize the safety-related return value for each human task choice. To identify the optimal values of the policies to ensure the safety and efficiency of the entire HRC-based disassembly process, an improved multi-agent proximal policy optimization (i-MAPPO) algorithm is designed. Finally, a case study for disassembling an electric vehicle battery is used to verify that the proposed approach can adapt to human uncertainty with a high success rate while ensuring that the disassembly time remains short and the human-robot distance remains within the safety threshold throughout the disassembly process. Note to Practitioners—HRC-based disassembly approaches can enhance the flexibility of disassembly lines. However, a significant challenge in HRC-based disassembly planning is the inherent uncertainty of human behavior, including instability in task completion times and unpredictability in task selection. To address this issue, this study combines the Stackelberg game theory with a multi-agent deep reinforcement learning optimization algorithm to design a dynamic disassembly task assignment approach. This approach can adapt to human uncertainties and quickly provide task assignment solutions based on changes in the disassembly scenario. Using the disassembly of retired electric vehicle batteries as a representative case, this approach offers a feasible solution for safe and efficient disassembly processes.}, archive = {J_TASE}, author = {Yiqun Peng and Weidong Li and Yong Zhou and Duc Truong Pham}, doi = {10.1109/TASE.2025.3557155}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13907--13919}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic disassembly planning of end-of-life products for Human–Robot collaboration enabled by multi-agent deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Bioinspired sensing of undulatory flow fields generated by leg kicks in swimming. TASE, 22, 13895--13906. (www)

  @article{ , abstract = {The artificial lateral line (ALL) is a bioinspired flow sensing system for underwater robots, comprising of distributed flow sensors. The ALL has been successfully applied to detect the undulatory flow fields generated by body undulation and tail-flapping of bioinspired robotic fish. However, its feasibility and performance in sensing the undulatory flow fields produced by human leg kicks during swimming have not been systematically tested and studied. This paper presents a novel sensing framework to investigate the undulatory flow field generated by swimmer’s leg kicks, leveraging bioinspired ALL sensing. To evaluate the feasibility of using the ALL system for sensing the undulatory flow fields generated by swimmer leg kicks, this paper designs an experimental platform integrating an ALL system and a lab-fabricated human leg model. To enhance the accuracy of flow sensing, this paper proposes a feature extraction method that dynamically fuses time-domain and time-frequency characteristics. Specifically, time-domain features are extracted using one-dimensional convolutional neural networks and bidirectional long short-term memory networks (1DCNN-BiLSTM), while time-frequency features are extracted using short-term Fourier transform and two-dimensional convolutional neural networks (STFT-2DCNN). These features are then dynamically fused based on attention mechanisms to achieve accurate sensing of the undulatory flow field. Furthermore, extensive experiments are conducted to test various scenarios inspired by human swimming, such as leg kick pattern recognition and kicking leg localization, achieving satisfactory results. Note to Practitioners—This paper tackles the challenge of sensing the flow fields created by human leg kicks during swimming using a bioinspired artificial lateral line (ALL) system. Unlike the undulating movement of a fish’s body or tail, human leg kicks vary in frequency and phase, and very limited research has been done on using ALL systems to sense these movements. To address this, we propose a new method for detecting complex flow patterns with the ALL system. We integrate a lab-fabricated human leg model with the ALL system to show how it captures the features of leg kicks. This method improves flow sensing accuracy by combining time-domain and time-frequency features using advanced neural networks. Key findings include the successful recognition of leg kick patterns and precise localization of the kicking leg, demonstrating the ALL system’s potential for sensing the undulatory flow fields. This study aims to provide theoretical and technical insights for human-robot interaction and formation using ALL systems, enhancing the capabilities of underwater robots and swimmers in complex environments. Additionally, this technology could help develop assistive robots for swimmers, boosting human abilities underwater and ensuring safety during aquatic activities.}, archive = {J_TASE}, author = {Jun Wang and Tongsheng Shen and Dexin Zhao and Feitian Zhang}, doi = {10.1109/TASE.2025.3556994}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13895--13906}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bioinspired sensing of undulatory flow fields generated by leg kicks in swimming}, volume = {22}, year = {2025}, }
• (2025). MagScanner: A magnetic planar XY stage for the scanning probe-based surface profilometer. TASE, 22, 13884--13894. (www)

  @article{ , abstract = {Planar XY scanners are essential for scanning probe microscopes. To overcome limitations in current piezo-actuated scanners, a novel normal-stressed electromagnetic scanner named MagScanner is introduced to scan a chromatic confocal sensor probe and construct a non-contact scanning probe-based surface profilometer. The MagScanner has a monolithic armature to provide dual-axial decoupled driving forces and a set of dual-axial compliant bearings to guarantee the decoupled planar scans. To properly select the system parameters for satisfying the design target, analytical models and finite element simulations are combined for the electromagnetic and mechanical components of the MagScanner. The experiment test on the prototype shows a stroke of $ \pm 200~ \mu $ m and a natural frequency of around 220 Hz for each axis, and the mutual dynamic cross-talks are smaller than 36.34 dB. Furthermore, by adopting the internal model control principle, the prototype precisely follows the Lissajous scanning trajectory with a peak-to-valley tracking error of less than 0.37%. After calibrating the error motion-induced measurement errors, the single-shot sub-aperture measurement of a typical micro-structured surface is demonstrated to have a peak-to-valley form error within $ \pm ~60 $ nm, which is further verified by applying for the measurement of a large-area micro-structured surface. Note to Practitioners—This work was motivated by developing a probe-based surface profilometer for measuring micro-structured surfaces. Although a large scanning area with high frequency was essential to achieve efficient surface scanning, it was challenging for conventional piezoelectric scanners owing to the inherently limited strain rate of piezoelectric ceramics. Hence, we developed a novel planar XY scanner based on normal-stressed electromagnetic actuation, which is promising for implementing dual-axial direct-drive scanning in hundreds of micrometers with an outperforming natural frequency.}, archive = {J_TASE}, author = {Rongjing Zhou and Yan-Ning Fang and Lingbao Kong and Li-Min Zhu and Zhiwei Zhu}, doi = {10.1109/TASE.2025.3532341}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13884--13894}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MagScanner: A magnetic planar XY stage for the scanning probe-based surface profilometer}, volume = {22}, year = {2025}, }
• (2025). A two-stage payload dynamic parameter identification method for interactive industrial robots with large components. TASE, 22, 13871--13883. (www)

  @article{ , abstract = {Taking human-robot collaborative assembly as an example, the methods based on contact forces can improve the assembly efficiency of industrial robots with large components in industrial manufacturing. However, due to the large size, high payload, assembly accuracy and dynamic changes in grip position, accurately estimating the contact forces between the payload and the operator becomes challenging when handling these large components. In this paper, a two-stage method is proposed for payload dynamic parameter identification. The parameter identification equation in the sensor coordinate system is initially established. Furthermore, the identification model of recursive restricted total least squares (RRTLS) based on total least squares (TLS) is constructed to achieve low-consumption online identification. According to the assembly requirements and payload characteristics, the posture coordinate system is designed for safety, including the feasible workspace for the robot. Subsequently, the static identification postures and dynamic excitation trajectory are planned to obtain static values and dynamic inertial parameters. In the end, a high-payload human-robot collaborative assembly system is built to validate the proposed method. Experimental results show that compared with the existing methods, the proposed approach can effectively identify and compensate the payload, leading to more accurate external force sensing. Note to Practitioners—Accurate payload parameter identification is essential for contact interactive applications of industrial robots. This research aims to solve two problems affecting parameter identification: the incompleteness of payload identification parameter consideration, and the real-time and security of large payload identification in industrial scenarios. In this paper, a two-stage payload dynamic parameter identification method is proposed, which can meet the application of large components in industrial scenarios. Firstly, a complete parameter identification model is established by decoupling the dynamic and static parameters. Then, a low-cost recursive identification is realized in a set feasible workspace. Finally, the external force sensing after parameter identification shows potential in but not limited to human-robot collaboration applications.}, archive = {J_TASE}, author = {Mingxuan Liu and Pengcheng Li and Jinjun Duan and Lunqian Liu and Ye Shen and Wei Tian and Yuqi Ji}, doi = {10.1109/TASE.2025.3557064}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13871--13883}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A two-stage payload dynamic parameter identification method for interactive industrial robots with large components}, volume = {22}, year = {2025}, }
• (2025). Supervisory control of discrete event systems modeled with labeled petri nets for diagnosability enforcement. TASE, 22, 13855--13870. (www)

  @article{ , abstract = {This paper proposes an active fault diagnosis method to enforce the diagnosability of discrete event systems using labeled Petri nets by constructing a diagnostic supervisor. For a non-diagnosable net model, its diagnosability is viewed as a control specification and addressed by using supervisory control techniques. First, an event-based monitor and a Petri net structure, referred to as data isolation arcs, are introduced to apply control specifications for labeled Petri nets. Then, a diagnostic supervisor reachability graph is generated to estimate the current state of the diagnoser. By analyzing the diagnostic supervisor reachability graph, an integer linear programming model is formulated to design a diagnostic supervisor, which can prevent the system from entering into any indeterminate cycle only by observing the occurrence of events. Finally, by the obtained diagnostic supervisor, the resulting net model is shown to be diagnosable. Some examples are presented to demonstrate the proposed method. Note to Practitioners—Faults have a significant impact on the normal operation of a system, and timely detection and isolation are crucial to ensuring system stability and production efficiency. Nevertheless, in certain systems, the occurrence of faults cannot be determined by a finite number of observations. To address this problem, this work introduces an active fault diagnosis method in the framework of labeled Petri nets, aiming to enforce the diagnosability of a system. The diagnosability condition is treated as a control specification such that it is readily accessible for practitioners to construct a diagnostic supervisor with Petri nets by following a typical and traditional control paradigm, which ensures that the controlled system is diagnosable.}, archive = {J_TASE}, author = {Chengzong Li and Yufeng Chen and Almetwally M. Mostafa and Mohamed Khalgui}, doi = {10.1109/TASE.2025.3554353}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13855--13870}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Supervisory control of discrete event systems modeled with labeled petri nets for diagnosability enforcement}, volume = {22}, year = {2025}, }
• (2025). Sim2Real learning with domain randomization for autonomous guidewire navigation in robotic-assisted endovascular procedures. TASE, 22, 13842--13854. (www)

  @article{ , abstract = {Over the past decade, significant advancements have been made in the research and industrialization of robotic systems for endovascular procedures, yet their clinical application remains relatively limited. Physicians commonly report that these robots lack certain intelligent assistive capabilities during procedures. There has been increasing interest and attempts to apply learning-centered algorithms to the training and enhancement of surgical robot skills. This paper proposes an autonomous navigation algorithm for interventional guidewires that is initially trained solely in a virtual simulation environment and subsequently deployed to a real-world robot. Experimental results demonstrate the feasibility of this approach for real-world applications. The proposed approach can help physicians reduce the learning curve for guidewire manipulation and elevate the robot to a higher level of autonomous operation, thereby breaking through the current bottleneck in the level of intelligence for clinical applications of interventional robots. It also holds promise for bringing intelligent transformation to future interventional procedures. Note to Practitioners—This work is motivated by the emerging need to increase the level of autonomy in robotic-assisted endovascular procedures, which has the potential to improve procedural efficiency, standardize procedures, and broaden the adoption of robotic systems in clinical practice. The proposed simulation-based reinforcement learning provides a safe and efficient method for training robotic systems, enabling them to master complex tasks in simulation environments prior to real-world application. The successful deployment of models trained in simulation onto physical robotic platforms demonstrates the feasibility of this method for real-world applications. The proposed simulation-based reinforcement learning method offers a promising and viable pathway for enhancing skill acquisition in endovascular interventional robots.}, archive = {J_TASE}, author = {Tianliang Yao and Haoyu Wang and Bo Lu and Jiajia Ge and Zhiqiang Pei and Markus Kowarschik and Lining Sun and Lakmal Seneviratne and Peng Qi}, doi = {10.1109/TASE.2025.3555559}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13842--13854}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sim2Real learning with domain randomization for autonomous guidewire navigation in robotic-assisted endovascular procedures}, volume = {22}, year = {2025}, }
• (2025). Conditional disturbance compensation-based fault-tolerant group consensus control for MASs: An event-based switching method. TASE, 22, 13828--13841. (www)

  @article{ , abstract = {With respect to the group consensus control for multi-agent systems (MASs) with disturbances and actuator nonidentical and unknown direction faults (NUDFs), this paper proposes a novel fault-tolerant control scheme by incorporating an event-based switching function and a saturation function. The event-based switching function is implemented to directly switch to the desired working mode, thereby reducing invalid switching after a reverse fault. Concurrently, the saturation function, acting at the software level, can not only prevent the instantaneous impact of excessive reverse input caused by a reverse fault, but also eliminate input peaking and state chattering after switching the working mode. It is worth mentioning that the proposed switching-based fault-tolerant control method can avoid the control chocks and excessive control gains as in the existing methods by using Nussbaum functions and thus improve the transient performance. Additionally, for disturbances, the designed conditional disturbance compensation mechanism ensures a smaller steady-state error, and less energy consumption by utilizing beneficial disturbances and compensating for harmful ones. Besides, the conditional compensation mechanism also reduces the saturation error. Finally, the effectiveness of the proposed control scheme are verified through a simulation example of uncrewed aerial vehicles (UAVs). Note to Practitioners—Actuator faults, especially reversed faults such as rudder reversals in aircraft or reverse currents in circuits and sensors, can severely disrupt industrial systems. With regard to NUDFs, this paper focuses on the switching function-based fault-tolerant method for high-order high-dimensional MASs, which are applicable to various practical systems such as UAV swarms, multi-robotic manipulators, multi-ship dynamic, multi-manned submersible vehicles, multi-quarter-car suspension, etc. Additionally, the switching function designed in this paper can be easily extended to solve the problems with unknown control directions. On the other hand, the estimator-controller framework provides a solution for achieving group consensus and fault-tolerant control separately, which simplifies the design of the fully distributed controller. Simulations of UAVs confirm the effectiveness of the proposed control approach.}, archive = {J_TASE}, author = {Pei-Ming Liu and Xiang-Gui Guo and Jian-Liang Wang and Wei-Wei Che and Zheng-Guang Wu}, doi = {10.1109/TASE.2025.3557448}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13828--13841}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Conditional disturbance compensation-based fault-tolerant group consensus control for MASs: An event-based switching method}, volume = {22}, year = {2025}, }
• (2025). A bayesian spike-and-slab sensor selection approach for high-dimensional prognostics. TASE, 22, 13814--13827. (www)

  @article{ , abstract = {With recent advances in sensor technology, more and more sensors are being used to simultaneously monitor the degradation of a system. As the number of sensors increases, it becomes increasingly difficult to distinguish informative sensors from uninformative sensors when performing prognostics, especially under the presence of different sensor correlations, signal-to-noise ratios, measurement units, and data characteristics. Existing methods for sensor selection typically rely on penalized-likelihood methods, which are known to provide biased estimates and poor sensor selection results in such high-dimensional settings. To overcome this challenge, we propose a novel data-fusion method that simultaneously selects informative sensors using Bayesian spike-and-slab priors and fuses the informative sensors into a 1-D health index (HI) to better characterize the degradation process for further prognostic analysis. Compared to the existing literature, the proposed Bayesian spike-and-slab sensor selection approach provides several unique advantages: 1) superior sensor selection performance in high-dimensional scenarios; 2) consistent sensor selection results with correlated sensors; 3) guaranteeing weak and strong selection consistency under mild assumptions; and 4) higher RUL prediction accuracy in a wide range of simulation and case studies. Note to Practitioners—This paper is motivated by the practical challenge of selecting informative sensors from high-dimensional multisensor systems (i.e., when there are many sensors relative to the number of training units) when conducting prognostics. Informative sensors not only provide valuable insights on the system’s degradation process, but also can be used to construct degradation indicators for practitioners to monitor and interpret the system status. In order to select informative sensors, this paper proposes a novel approach involving Bayesian spike-and-slab priors. After we select the informative sensors, they are then fused into a 1-D HI to better characterize the degradation process. This approach is particularly useful for selecting informative sensors under high-dimensional scenarios with possibly correlated sensors, when all units degrade under a single failure mode and operating condition.}, archive = {J_TASE}, author = {Ye Kwon Huh and Ying Fu and Kaibo Liu}, doi = {10.1109/TASE.2025.3555165}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13814--13827}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A bayesian spike-and-slab sensor selection approach for high-dimensional prognostics}, volume = {22}, year = {2025}, }
• (2025). Graph representation learning for infrared and visible image fusion. TASE, 22, 13801--13813. (www)

  @article{ , abstract = {Infrared and visible image fusion aims to extract complementary features to synthesize a single fused image. In our method, we covert the regular image format into the graph space and conduct graph convolutional networks (GCNs) to extract NLss for the reliable infrared and visible image fusion. More specifically, GCNs are first performed on each intra-modal set to aggregate the features and propagate the inherent information, thereby extracting independent intra-modal NLss. Then, such intra-modal non-local self-similarity (NLss) features of infrared and visible images are concatenated to explore cross-domain NLss inter-modally and reconstruct the fused images. Extensive experiments show the superior performance of our method with the qualitative and quantitative analysis on the TNO, RoadScene and M3FD datasets, respectively, outperforming many state-of-the-art (SOTA) methods for the robust and effective infrared and visible image fusion. Note to Practitioners—This paper was motivated by the problem that the most existing methods that employ convolutional neural networks (CNNs) and transformer-based frameworks mainly extract local features and long-range dependence. However, they often cause overlooking the image’s NLss or information redundancy, resulting in poor infrared and visible image fusion. To address these problems, graph-based data representations can construct relationships among spatially repeatable details or textures with far-space distances, which are more suitable for handling irregular objects. Therefore, it is significant to covert the regular image format into the graph space and conduct graph convolutional networks (GCNs) to extract NLss for the reliable infrared and visible image fusion. In this paper, we develop an infrared and visible image fusion method based on graph representation learning strategy.}, archive = {J_TASE}, author = {Jing Li and Lu Bai and Bin Yang and Chang Li and Lingfei Ma}, doi = {10.1109/TASE.2025.3557234}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13801--13813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Graph representation learning for infrared and visible image fusion}, volume = {22}, year = {2025}, }
• (2025). Adaptive semi-global bipartite average tracking of nonlinear multi-agent systems with input saturation via observer-based approach. TASE, 22, 13788--13800. (www)

  @article{ , abstract = {This paper addresses adaptive semi-global bipartite average tracking (SGBAT) for nonlinear multi-agent systems (MASs) with input saturation under two different low-gain feedback based control protocols via observer-based approach. A state observer and an average signal estimator (ASE) are designed for estimating actual states and obtaining auxiliary outputs. By using the estimated states and the auxiliary outputs, a continuous distributed control protocol is first proposed with adaptive gains. By constructing appropriate Lyapunov function, the adaptive SGBAT can be guaranteed for saturated nonlinear MASs by using low-gain feedback technique. Then, for reducing the communication load, an adaptive dynamic event-triggered control (DETC) protocol is developed with two dynamic event-triggered mechanisms (DETMs). Finally, for verifying the effectiveness of the two control protocols, an application simulation example of multiple satellite systems is displayed. Note to Practitioners—The control of nonlinear MASs with input saturation remains a significant challenge in engineering applications, such as formation of multiple satellite systems, synchronization of robot systems, etc. In practical systems, input saturation can lead to performance degradation if not properly addressed. Meanwhile, considering that actual system states are often unmeasured in these scenarios, further complexity arises. To solve the problem, this paper introduces an adaptive observer-based control protocol with a boundary layer function firstly. However, continuous communication between agents can lead to excessive communication overhead in systems with limited resources. To address this, an observer-based DETC protocol is further developed to reduce the communication load without sacrificing control performance. The proposed control schemes, which account for control inputs constrained by maximum and minimum values, are particularly applicable to scenarios like multiple satellite systems, where utilization rate of communication resources and tracking performance under input saturation are critical. The effectiveness of the control protocols is demonstrated through simulation, showcasing its potential for broader application in similar engineering challenges.}, archive = {J_TASE}, author = {Haichuan Xu and Fanglai Zhu}, doi = {10.1109/TASE.2025.3556707}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13788--13800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive semi-global bipartite average tracking of nonlinear multi-agent systems with input saturation via observer-based approach}, volume = {22}, year = {2025}, }
• (2025). Multi-context aggregation network with foreground correction for automated few-shot defect segmentation. TASE, 22, 13777--13787. (www)

  @article{ , abstract = {State-of-the-art defect segmentation methods rely on sufficient training data and struggle to generalize to unseen categories. Few-Shot Semantic Segmentation (FSS) is introduced to specifically address these issues. However, existing FSS models still face two challenges in the industry. 1) Defects usually present as weak features, resulting in incomplete segmentation; 2) Severe background interference often leads to incorrect segmentation. To tackle these problems, we propose the Multi-Context Aggregation Network (MCANet). Specifically, we design a Cross-Layer Multi-Level Feature Aggregation Module (CMAM). CMAM effectively aggregates discretely distributed multi-level defect features across different layers and guides the query image to perceive defects from the pixel level, which avoids incomplete segmentation caused by weak features. Additionally, a Foreground Correction Module (FCM) is developed, which is equipped with a dedicated background predictor (BP) and a foreground corrector (FC). BP places more emphasis on learning features from backgrounds rather than defects. FC achieves efficient feature ensemble and further suppresses the backgrounds misidentified as defects in CMAM. They collaborate to prevent incorrect segmentation caused by background interference. Extensive experiments demonstrate the effectiveness of our method. We achieve state-of-the-art results on both FSSD-12, a public benchmark FSS dataset for strip steel, and FSS-AEB, an FSS dataset for aero-engine blades. Specifically, with 1/5 support images, we achieve 64.6%/65.6% mIoU on FSSD-12 and 55.0%/57.8% mIoU on FSS-AEB. Note to Practitioners—Surface defect segmentation has always been a hot topic in the industry. However, existing methods rely on sufficient training data and struggle to generalize to unseen categories, which significantly hinders the automation of defect segmentation. To address this problem, we propose MCANet for automated few-shot defect segmentation. It achieves effective segmentation for surface defects with limited data, even for unseen categories. Furthermore, MCANet achieves state-of-the-art results on two datasets from real-world industrial scenarios and also delivers significant improvements over the widely concerned large vision models. Finally, we integrate MCANet into an automated surface defect inspection platform consisting of an imaging system and a high-performance computing server for real-world performance validation.}, archive = {J_TASE}, author = {Yunfeng Ma and Min Liu and Shuai Jiang and Xueping Wang and Yuan Bian and Yaonan Wang}, doi = {10.1109/TASE.2025.3554987}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13777--13787}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-context aggregation network with foreground correction for automated few-shot defect segmentation}, volume = {22}, year = {2025}, }
• (2025). Adaptive reinforcement learning tracking control of vehicle based on threshold band event-triggered. TASE, 22, 13764--13776. (www)

  @article{ , abstract = {In this paper, an adaptive neural network control algorithm based on event-triggered reinforcement learning is proposed for a four-wheel independent steering and four-wheel independent driving (4WS4WD) mobile robot. A kinematic model is established based on the kinematic relationship between the robot wheels and the body under the consideration of the effect of slip-turn perturbation. The dynamics model is established using the Lagrangian dynamics equations. An improved performance metric function is designed and approximated using the Critic neural network and the Actor neural network to approximate the unknown long-term performance metric function and controller respectively. A threshold band event triggering is proposed for reducing the consumption of communication and computational resources. It is rigorously demonstrated using Lyapunov analysis that both the neural network error and the system error are up to the final consistent bound. As well as proved that the proposed event-triggered mechanism can eliminate the Zeno phenomenon. Finally, comparative experiments demonstrated the effectiveness of the proposed algorithm. Note to Practitioners—4WS4WD mobile robots are able to perform inspection tasks in more scenarios due to their excellent characteristics. However, due to its complex structure, it is difficult to control it accurately when there are slipping, sliding and input disturbances. In this paper, the model of the robot is established by considering the effects of slip, slip and disturbance. And an adaptive neural network control algorithm based on threshold band event-triggered reinforcement learning is proposed. Simulation results show that the proposed algorithm enables the robot to effectively reduce computational resources while maintaining good tracking results.}, archive = {J_TASE}, author = {Yan-Jun Liu and Xiaosheng Sun and Yi Liao and Shu Li and Lei Liu and Jason J. R. Liu}, doi = {10.1109/TASE.2025.3555765}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13764--13776}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive reinforcement learning tracking control of vehicle based on threshold band event-triggered}, volume = {22}, year = {2025}, }
• (2025). Trajectory optimization for tooth preparation robot based on P-MRSD algorithm. TASE, 22, 13752--13763. (www)

  @article{ , abstract = {Oral diseases, including dental caries and cracked teeth, are leading causes of tooth loss and can pose significant health risks if left untreated. Manual tooth preparation by dentists is often prone to visual bias and positioning errors. To address these issues, tooth preparation robots, driven by automation and intelligence, have been proposed as replacements for repetitive tasks. However, existing tooth preparation robots with serial systems suffer from low preparation accuracy and poor safety when preparing real teeth of hard and brittle materials. This includes the problems of low accuracy of theoretical preparation trajectory planning and deformation of the end of the serial system due to force. In this study, we propose a preparation trajectory optimization strategy that includes methods for surface morphology optimization and end deformation optimization to enhance preparation accuracy and safety. Our approach utilizes the proposed prediction of material residue and stiffness deformation (P-MRSD) method to optimize tooth morphology based on five key parameters, such as the bur pose and preparation trajectory. Additionally, the extrusion force during preparation is optimized by considering factors such as the material removal rate, the Euclidean distance field of tool contacts (TC), and the system stiffness of the tooth preparation robot. The accuracy and safety of the execution trajectory are ensured by minimizing stiffness deformation. Finally, a tooth preparation robot hardware system is developed to verify the correlation between predicted and observed tooth morphology, with end deformation optimization based on the optimized trajectory. The feasibility of the robot for preparing hard and brittle teeth is demonstrated, and the proposed trajectory optimization method improves both accuracy and safety in preparation process. This provides a theoretical foundation and technical support for advancing automated robotic technology, particularly in the development of more accurate and safer serial robotic arm systems. Note to Practitioners—The motivation of this paper is to address the challenge of trajectory optimization for grinding hard and brittle materials using a serial robotic system, with potential applications in machining. The interaction between tool pose, trajectory, and material (specifically a cracked hard and brittle tooth) is analyzed to investigate its impact on preparation. Two optimization parameters are proposed to refine the theoretical preparation trajectory based on the surface morphology indexes. Real preparation experiments are then conducted using the proposed extrusion force optimization model, which enhances the motion accuracy of the theoretical trajectory. This improvement boosts the performance of the serial robotic arm system when preparing hard and brittle materials. The findings also suggest that computer-aided design and manufacturing systems could autonomously generate preparation trajectory optimization plans that consider both surface morphology and safety, offering insights and support for machining in other fields. Preliminary physical experiments demonstrate the feasibility of the proposed trajectory planning and kinematic parameter optimization methods, improving both preparation accuracy and safety. However, the research is still in the laboratory phase and has not yet been applied clinically. Future studies will focus on balancing safety and accuracy when the serial system operates intraorally, advancing the potential clinical application of tooth preparation robots.}, archive = {J_TASE}, author = {Jianpeng Sun and Jingang Jiang and Chunrui Wang and Zhonghao Xue and Ao Li and Jie Pan}, doi = {10.1109/TASE.2025.3555227}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13752--13763}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory optimization for tooth preparation robot based on P-MRSD algorithm}, volume = {22}, year = {2025}, }
• (2025). Asynchronous switching fuzzy control for stochastic highly nonlinear systems with markov jump and periodic time-varying delay: By dupire’s functional itô formula. TASE, 22, 13741--13751. (www)

  @article{ , abstract = {This paper predominantly studies how to achieve the mean-square exponential stability (MSES) of stochastic highly nonlinear fuzzy systems with Markov jump and periodic time-varying delay (SHNFSMPTD), using asynchronous switching fuzzy control. Considering the time-varying delay and the drift and diffusion terms satisfying polynomial growth conditions, we construct two different functionals that explicitly depend on time t in the monotonically increasing and decreasing intervals, respectively. Combined with Dupire’s functional Itô formula, graph theory, and the properties of periodic time-varying delay, we can show that the negative definiteness of the functionals holds in both intervals, which leads to the stability of systems. In addition, by combining the generator matrix of Markov jump and the conditional probability matrix of asynchronous switching, we provide additional MSES criteria for SHNFSMPTD under asynchronous switching fuzzy control. Ultimately, the theoretical results obtained are utilized in a class of van der Pol-Duffing oscillators (VDPDO), and the six-robot systems with center pattern generator networks, and the feasibility of the theoretical results is illustrated through numerical simulations. Note to Practitioners—Stochastic highly nonlinear fuzzy systems with Markov jump and periodic time-varying delays are widely found in many practical systems, such as communication networks, control systems, and biological rhythm systems. In this paper, we propose for the first time to deal with the exponential stability problem of such systems using Dupire’s functional Itô formula. By constructing suitable functionals and applying Dupire’s functional Itô formula, we obtain the negative definiteness of the operator in both intervals. This paper provides a new method for the construction of Lyapunov functional and offers some new ideas for the stability study of this kind of systems. In the future, we attempt to study the noise stabilization of such systems with various attacks and intermittent communication under intermittent control.}, archive = {J_TASE}, author = {Ning Zhang and Zhenhao Zhang and Ju H. Park and Wenxue Li}, doi = {10.1109/TASE.2025.3553682}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13741--13751}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous switching fuzzy control for stochastic highly nonlinear systems with markov jump and periodic time-varying delay: By dupire’s functional itô formula}, volume = {22}, year = {2025}, }
• (2025). Auxiliary reward generation with transition distance representation learning. TASE, 22, 13728--13740. (www)

  @article{ , abstract = {Reinforcement learning (RL) has shown strengths in challenging sequential decision-making problems. The reward function in RL is crucial to the learning performance, as it quantifies the degree of task completion. In real-world problems, the rewards are predominantly human-designed, which requires laborious tuning, and is susceptible to human cognitive biases. To achieve automatic auxiliary reward generation, we propose a novel representation learning approach that can measure the “transition distance” between states. Building upon these representations, we introduce an auxiliary reward generation technique for both single-task and skill-chaining scenarios without the need for human knowledge. Furthermore, we theoretically show that the proposed auxiliary rewards maintain the policy invariance property, i.e., the generated rewards will not hurt the policy optimality under the original rewards. In the experiment section, we evaluate the proposed approach in both online and offline learning settings in a wide range of tasks, including robot manipulation and locomotion. The experiment results demonstrate the effectiveness of measuring the transition distance and the induced improvement by auxiliary rewards, which promotes better learning efficiency and increases convergent stability. Beyond that, we demonstrate that the learned manipulation policy with the auxiliary rewards in a simulator can be transferred to the real robot, as shown in https://sites.google.com/view/transition-distance-rp/tdrp. Note to Practitioners—The motivation for this paper arises from the need for a technique that enhances robot skill-learning efficiency and performance in both single-task and skill-chaining scenarios. Our research primarily focuses on robot arm manipulation tasks. To accelerate the policy learning process and improve policy performance for executing these tasks, we introduce an auxiliary reward generation technique for both single-task and skill-chaining scenarios without requiring human expertise. This technique leverages the proposed novel representation learning approach, which can measure the “transition distance” between states. During each policy training round, the robot receives a dense reshaped reward created by our approach. Using the policy trained by our method, we successfully control a real Franka Panda robot arm to complete various manipulation tasks.}, archive = {J_TASE}, author = {Siyuan Li and Shijie Han and Yingnan Zhao and Yiqin Yang and Qianchuan Zhao and Peng Liu}, doi = {10.1109/TASE.2025.3554861}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13728--13740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Auxiliary reward generation with transition distance representation learning}, volume = {22}, year = {2025}, }
• (2025). Robotic fluorescent lighting method based on dynamic fluorescence imaging modeling for in vivo cell manipulation. TASE, 22, 13716--13727. (www)

  @article{ , abstract = {In vivo cell manipulation is crucial for understanding organ functions and dysfunctions at the cellular level. Due to its low optical transparency, the in vivo environment needs to be lighted up first to visualize the targets for manipulation. At present, operators usually carefully blow a micropipette filled with fluorescent dye solution to create a fluorescence field around the micropipette opening, lighting up the in vivo environment. This manual dynamic control of the fluorescence field is usually a low-accuracy, labor-intensive, and high-skill requirement task, becoming even more challenging when the micropipette moves to locate or approach target cells. A large number of in vivo lighting-control tests may improve manual control efficiency of the fluorescent field, but they usually cost many precise animal samples, making them costly and often unaffordable. To enhance the dynamic control efficiency of the fluorescent field while reducing experimental costs, a robotic fluorescent lighting control method is proposed based on an in vivo dynamic fluorescent imaging simulator. First, a dynamic fluorescence imaging model composing an injection model, a diffusion model, and a luminescence model of fluorescent molecules is established to simulate the fluorescence field in vivo. The fluorescent intensity distribution obtained through the model is highly in accordance with experimental results. Based on this model, an adaptive sliding mode controller is employed to achieve the desired fluorescent intensity in the brain tissue. Both the simulation results and the experimental results demonstrate significant advantages of the proposed controller in terms of control accuracy and stability compared to the traditional PID controller and manual operation. Note to Practitioners—Forming a stable fluorescent light field with a specified fluorescent intensity is of vital importance operation for in vivo micromanipulations. At present, manual adjustment of the fluorescent field by blowing the micropipette with mouth is time-consuming and energy-draining. In this paper, we first propose a dynamic fluorescence imaging simulator to address the economic and time costs of live experiments. We then design an adaptive sliding mode controller to dynamically control the fluorescent intensity at a specified position in the fluorescent field. Simulations based on the established model, along with experiments on brain tissues, demonstrate the controller’s effectiveness. Our system is expected to reduce human involvement and enable high-precision lighting for in vivo cell manipulation in the future.}, archive = {J_TASE}, author = {Ruimin Li and Jie Li and Zizhen Li and Jinyu Qiu and Bo Hu and Minghui Li and Mingzhu Sun and Xin Zhao and Qili Zhao}, doi = {10.1109/TASE.2025.3556926}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13716--13727}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic fluorescent lighting method based on dynamic fluorescence imaging modeling for in vivo cell manipulation}, volume = {22}, year = {2025}, }
• (2025). Dynamic protocol-based SMC for networked continuous stirred tank reactor system: Switched stochastic semi-markov theory. TASE, 22, 13706--13715. (www)

  @article{ , abstract = {In this paper, the sliding mode control (SMC) problem is investigated for continuous stirred tank reactor (CSTR) via dynamic self-triggered protocol subject to external disturbance. By introducing the switched stochastic system with semi-Markov parameter, the fluctuations in product concentration and the variations in reaction temperature within the CSTR are modeled. Without the sojourn-time constraint, novel strategy about asymptotic mean-square stability is adopted to design the SMC scheme based on the Lyapunov function and the stationary distribution of embedded Markov chain. Furthermore, to enhance the transmission efficiency and the dynamic performance, novel self-triggered protocol is developed in combination with SMC technology, predicting the subsequent triggered moment under the current sampling data and providing the lower bound of internal execution time. Under the dynamic self-triggered protocol, an appropriate SMC law is proposed to ensure the finite-time reachability of the specified sliding surface. Sufficient conditions are established for asymptotic mean-square stability of the associated switched stochastic systems. Finally, the proposed method is verified the simulation. Note to Practitioners—The continuous stirred tank reactor (CSTR) system is widely used in the industrial processes, in which the performance and control are critical to efficiency and quality. Under the key parameters of temperature and concentration, it is necessary to strictly control the temperature and monitor the concentration in real-time. Modern manufacturing and industrial processes have become increasingly complex, and single system can no longer accurately describe the dynamic characteristics. As a special hybrid system with continuous evolution, instantaneous change, and stochastic effect, switched stochastic semi-Markov system can describe the concentration fluctuations and reaction temperature changes in the CSTR, showing significant potential for applications. In recent years, networked control system has attracted attention due to their advantages in remote task execution. The self-triggered protocols conserve bandwidth and reduce dependence on external devices by predicting the next triggered time. In this work, switched stochastic system with semi-Markov parameter is adopted to investigate the self-triggered sliding mode control (SMC) problem for the CSTR. This research provides a novel perspective of SMC strategy based on switched stochastic semi-Markov system and self-triggered algorithm in the CSTR.}, archive = {J_TASE}, author = {Wenhai Qi and Yongbo Yang and Guangdeng Zong and Zheng-Guang Wu and Huaicheng Yan}, doi = {10.1109/TASE.2025.3555385}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13706--13715}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic protocol-based SMC for networked continuous stirred tank reactor system: Switched stochastic semi-markov theory}, volume = {22}, year = {2025}, }
• (2025). Dual graph attention networks for multi-view visual manipulation relationship detection and robotic grasping. TASE, 22, 13694--13705. (www)

  @article{ , abstract = {Visual manipulation relationship detection facilitates robots to achieve safe, orderly, and efficient grasping tasks. However, most existing algorithms only model object-level or relational-level dependency individually, lacking sufficient global information, which is difficult to handle different types of reasoning errors, especially in complex environments with multi-object stacking and occlusion. To solve the above problems, we propose Dual Graph Attention Networks (Dual-GAT) for visual manipulation relationship detection, with an object-level graph network for capturing object-level dependencies and a relational-level graph network for capturing relational triplets-level interactions. The attention mechanism assigns different weights to different dependencies, obtains more accurate global context information for reasoning, and gets a manipulation relationship graph. In addition, we use multi-view feature fusion to improve the occluded object features, then enhance the relationship detection performance in multi-object scenes. Finally, our method is deployed on the robot to construct a multi-object grasping system, which can be well applied to stacking environments. Experimental results on the datasets VMRD and REGRAD show that our method significantly outperforms others. Note to Practitioners—The motivation of this research paper is to design efficient and accurate visual manipulation relationship reasoning methods to accomplish relevant grasping tasks in complex stacked scenes with multiple objects. The problem requires judging the positional relationship between objects in a stacked scene and determining the appropriate grasping order. In order to accurately grasp the target objects, this paper proposes a Dual Graph Attention Network for visual manipulation relationship detection, which utilizes object-level and relationship-level dependencies to obtain accurate global information for better performance. Meanwhile, multi-view feature fusion effectively improves the object occlusion problem. Our method can be combined with grasping detection to realize the robot grasping task in a real environment with multiple objects stacked in cluttered scenes. Practitioners can apply our method to real-time robot operating systems.}, archive = {J_TASE}, author = {Mengyuan Ding and Yaxin Liu and Yaorui Shi and Xuguang Lan and Nanning Zheng}, doi = {10.1109/TASE.2025.3555206}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13694--13705}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual graph attention networks for multi-view visual manipulation relationship detection and robotic grasping}, volume = {22}, year = {2025}, }
• (2025). Peak force torsional resonance microscopy for accurate nanoscale characterization of in-plane and out-of-plane mechanical properties. TASE, 22, 13684--13693. (www)

  @article{ , abstract = {Accurately measuring both in-plane and out-of-plane mechanical properties is essential for understanding material behavior at the nanoscale. Conventional atomic force microscopy (AFM) techniques often struggle with the coupling between vertical and lateral forces, which can distort lateral mechanical property measurements. To address this limitation, we propose a novel Peak Force Torsional Resonance (PFTR) method that combines Peak Force Tapping for precise height control with torsional resonance for in-plane property extraction. This approach effectively decouples vertical and lateral forces, significantly reducing the influence of topography and vertical interactions on lateral measurements. Additionally, we introduce a mathematical model to accurately quantify lateral properties, such as shear modulus and lateral viscosity, independent of height variations. Experimental results confirm that the PFTR method achieves high-resolution imaging of both in-plane and out-of-plane mechanical properties, offering a more accurate and reliable solution for nanoscale mechanical characterization compared to conventional AFM-based methods. Note to Practitioners—This work addresses a common challenge in nanoscale mechanical property measurement: the interference from surface topography that affects accurate imaging of in-plane (lateral) properties. Traditional AFM methods struggle with this coupling effect, making it difficult to reliably measure both in-plane and out-of-plane mechanical characteristics—measurements that are vital in fields such as semiconductor manufacturing and materials development. This paper introduces PFTR Microscopy, a new AFM-based approach designed to decouple these forces and accurately resolve nanoscale mechanical properties. PFTR combines PFT for height control with torsional resonance to extract in-plane mechanical data without distortions from surface topography. By providing clear imaging of lateral properties and reducing topographical artifacts, PFTR enables more reliable assessments of materials such as semiconductor coatings, thin films, and composite compounds used in high-precision industries. PFTR is validated through experimental comparisons with conventional AFM methods, demonstrating that it achieves higher resolution and accuracy in in-plane measurements. While this approach provides robust data for characterizing homogeneous materials, further research is needed to adapt it for complex, heterogeneous surfaces and to simplify calibration for broader industrial application. The ability to distinguish mechanical properties at the nanoscale also suggests potential for applications in MEMS design, failure analysis, and the characterization of soft or biological materials.}, archive = {J_TASE}, author = {Jialin Shi and Huiyao Shi and Yongliang Yang and Peng Yu and Tie Yang and Yang Yang and Chanmin Su and Lianqing Liu}, doi = {10.1109/TASE.2025.3555896}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13684--13693}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Peak force torsional resonance microscopy for accurate nanoscale characterization of in-plane and out-of-plane mechanical properties}, volume = {22}, year = {2025}, }
• (2025). High-fidelity integrated aerial platform simulation for control, perception, and learning. TASE, 22, 13662--13683. (www)

  @article{ , abstract = {This paper presents a simulator framework tailored Integrated Aerial Platforms (IAPs) using multiple quadrotors. Our framework prioritizes photo and contact fidelity, achieved through a modular design that balances rendering and dynamics computation. Key features include: i) support for diverse IAP configurations; ii) a customizable physics engine for realistic motion and contact simulation for aerial manipulation; and iii) Unreal Engine 5 for lifelike rendering, with sensor designs for visual-inertial SLAM positioning simulation. We showcase our framework’s versatility through a range of scenarios, including trajectory tracking for both fully and under-actuated IAPs, peg-in-hole and direct wrench control tasks under external wrench influence, tightly-coupled SLAM positioning with physical constraints, and air docking task training and testing using offline-to-online reinforcement learning. Furthermore, we validate our simulator framework’s fidelity by comparing results with real flight data for trajectory tracking and direct wrench control tasks. Our simulator framework promises to be valuable for developing and testing integrated aerial platform systems for aerial manipulation. Note to Practitioners—Motivated by the demand for effective simulation tools for Integrated Aerial Platforms (IAPs), this research addresses a significant gap in the availability of comprehensive simulation platforms designed to meet their unique challenges. This paper presents a high-fidelity simulation platform tailored specifically for IAPs, supporting a variety of configurations and capabilities. The platform not only generates high-fidelity image data and facilitates contact simulation but also serves as a vital resource for advancing perception, control, and learning for IAPs. By offering a robust simulation environment, this work aims to bridge the divide between theoretical research and practical applications, ultimately driving advancements in the field of aerial robotics.}, archive = {J_TASE}, author = {Jianrui Du and Kaidi Wang and Yingjun Fan and Ganghua Lai and Yushu Yu}, doi = {10.1109/TASE.2025.3555014}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13662--13683}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-fidelity integrated aerial platform simulation for control, perception, and learning}, volume = {22}, year = {2025}, }
• (2025). DartBot: Overhand throwing of deformable objects with tactile sensing and reinforcement learning. TASE, 22, 13644--13661. (www)

  @article{ , abstract = {Object transfer through throwing is a classic dynamic manipulation task that necessitates precise control and perception capabilities. However, developing dynamic models for unstructured environments using analytical methods presents challenges. In this study, we present DartBot, a robot that integrates tactile exploration and reinforcement learning to achieve robust throwing skills for nonrigid relatively small objects under the influence of moment of inertia which cause the object to spin in the air. Unlike traditional sim-to-real transfer methods, our approach involves direct training of the agent on a real hardware robot equipped with a high-resolution tactile sensor, enabling reinforced learning in a realistic and dynamic environment. By leveraging tactile perception, we incorporate pseudo-embeddings of the physical properties of objects into the learning process through tilting actions at two distinct angles. This tactile information enables the agent to infer and adapt its throwing strategy, resulting in improved accuracy when handling various objects and targeting distant locations. Furthermore, we demonstrate that the quality of a grasp significantly impacts the success rate of the throwing task. We evaluate the effectiveness of our method through extensive experiments, demonstrating superior performance and generalization capabilities in real-world throwing scenarios. We achieved a success rate of 95% for unseen objects with a mean error of 3.15 cm from the goal. A high-resolution video demo of our work is available at https://youtu.be/KNFgDeLt-0g. Note to Practitioners—The industrial demand for precise and accurate object transfer beyond the robot’s maximum kinematic range is rapidly growing, necessitating advancements in robotic throwing manipulation to efficiently utilize resources. In this context, this paper contributes to the field by presenting a method for the transfer of deformable objects through overhand throwing, addressing the challenges of achieving precise control and perception capabilities by learning the complex physics of the task using raw data. The approach offers valuable insights and techniques for enhancing throwing manipulation tasks. The integration of high-resolution tactile sensing and reinforcement learning, while considering the influence of moment of inertia, opens up new possibilities for handling deformable objects in throwing tasks and developing dynamic models for such unstructured environments. The learned throwing policy is applied to a variety of previously unseen objects, demonstrating its effectiveness and ability to generalize in real-world throwing scenarios. The presented work holds applicability in various domains such as medical rehabilitation, logistic warehouses for packaging, and handling of urban waste. It shows promise in improving the precision, accuracy and efficiency of the throwing task. In the future, we aim to expand the current overhand throwing framework by incorporating rapid grasping in cluttered environments and enabling throwing to far distances with diverse target locations.}, archive = {J_TASE}, author = {Shoaib Aslam and Krish Kumar and Pokuang Zhou and Hongyu Yu and Michael Yu Wang and Yu She}, doi = {10.1109/TASE.2025.3556875}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13644--13661}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DartBot: Overhand throwing of deformable objects with tactile sensing and reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered resilient control of nonlinear multi-agent systems against asynchronous DoS attacks. TASE, 22, 13633--13643. (www)

  @article{ , abstract = {This paper investigates dynamic event-triggered resilient formation control for nonlinear multi-agent systems under multi-channel Denial-of-Service (DoS) attacks. Unlike existing resilient formation control strategies, which assume a unified attack model across all communication channels among followers under known system dynamics, which is not realistic. We propose a novel architecture for secure formation control to tackle the challenges posed by heterogeneous and uncertain dynamics, as well as distributed and asynchronous DoS attacks. Specifically, each communication channel, whether between the leader and followers, or among followers, may be independently and asynchronously attacked. In addition, a novel dynamic event-triggered mechanism that incorporates attack parameters is designed to mitigate the overuse of network bandwidth while avoiding the Zeno behavior. Notably, in comparison with conventional methods for heterogeneous cooperative systems, our approach eliminates the need for distributed observers to reconstruct the leader’s information, thus significantly reducing the transmission of additional variables and simplifying the system architecture. Finally, the effectiveness of our proposed algorithms is verified through a practical example involving a multi-robot system. Note to Practitioners—This paper investigates dynamic event-triggered resilient formation control for nonlinear multi-agent systems under multi-channel Denial-of-Service (DoS) attacks. The proposed distributed resilient control algorithms can be applied to multiple ground vehicles, air vehicles, and underwater vehicles. Unlike existing resilient formation control strategies, which assume a unified attack model across all communication channels among followers under known system dynamics, which is not realistic. We propose a novel architecture for secure formation control to tackle the challenges posed by heterogeneous and uncertain dynamics, as well as distributed and asynchronous DoS attacks. Specifically, each communication channel, whether between the leader and followers, or among followers, may be independently and asynchronously attacked. In addition, a novel dynamic event-triggered mechanism that incorporates attack parameters is designed to mitigate the overuse of network bandwidth while avoiding the Zeno behavior. Notably, in comparison with conventional methods for heterogeneous cooperative systems, our approach eliminates the need for distributed observers to reconstruct the leaders information, thus significantly reducing the transmission of additional variables and simplifying the system architecture. Finally, the effectiveness of our proposed algorithms is verified through a practical example involving a multi-robot system.}, archive = {J_TASE}, author = {Meng-Ying Wan and Yong Xu and Zheng-Guang Wu}, doi = {10.1109/TASE.2025.3552133}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13633--13643}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered resilient control of nonlinear multi-agent systems against asynchronous DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time fault-tolerant containment control of fully actuated heterogeneous multiagent systems without estimations of fault parameters. TASE, 22, 13621--13632. (www)

  @article{ , abstract = {This paper studies practical prescribed-time control of fully actuated heterogeneous multiagent systems subjected to actuator faults. A new observer-based containment control structure is given in the scenario that only a small number of followers can access the knowledge of leaders. Firstly, prescribed-time distributed input, velocity, and position observers are designed to estimate the information of the convex hull that leaders have spanned. Then, the original practical prescribed-time fault-tolerant control issue is transformed into one with a deferred constraint on tracking errors by introducing a time-varying constraining function, and decentralized tracking controllers are designed based on the system’s fully actuated system model. With the prescribed-time prescribed performance function, the proposed method allows for the advance determination of the settling time and the final tracking accuracy as needed, as well as the reduction of the impact of actuator faults on the system without the need for actuator fault parameter estimations. The efficacy of the suggested approach is shown through simulation results. Note to Practitioners—The containment control issue of heterogeneous multiagent systems, as a popular topic in the control field, is crucial to practical engineering. It is worth noting that actuator faults often exist and they have the potential to spread throughout networks in practical applications. A novel observer-based containment control structure is developed and a robust fault-tolerant control algorithm is given to reduce the impact of actuator faults on the system without the need for actuator fault parameter estimations. What’s more, achieving containment control in a brief amount of time is highly desirable. Therefore, a practical prescribed-time control protocol is given to ensure containment errors converge to the predefined region within an assignable prescribed settling time interval. In summary, a practical prescribed-time fault-tolerant containment control technique is proposed, which promotes the advancement of containment control for MASs in practical applications.}, archive = {J_TASE}, author = {Yonghao Ma and Ke Zhang and Bin Jiang}, doi = {10.1109/TASE.2025.3555239}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13621--13632}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time fault-tolerant containment control of fully actuated heterogeneous multiagent systems without estimations of fault parameters}, volume = {22}, year = {2025}, }
• (2025). Smooth Human–Robot shared control for autonomous orchard monitoring with UGVs. TASE, 22, 13603--13620. (www)

  @article{ , abstract = {Precision agriculture offers the opportunity to automate routine or difficult tasks in orchards and vineyards, such as spraying or inspection, with Uncrewed Ground Vehicles (UGV). In this context, human operators should be kept in the closed-loop control of the robot for safety and reliability. This work is motivated by the challenges of effectively deploying human-robot shared control in the field. First, an asymptotically stable controller keeps the robot on the desired trajectory between rows of trees, whose distance is on the order of the robot’s width. Second, the robot must efficiently avoid static and moving obstacles on its path. Third, the control inputs must not exceed the actuator limits, which can degrade trajectory tracking performance, cause instability, or damage critical hardware. Finally, in real-life scenarios, user intervention is sometimes required to manage unpredictable situations. To overcome these challenges, we propose and deploy a shared controller that continuously and smoothly varies the ratio of human and automatic control inputs depending on the human’s intent, geometrically rescales trajectory inputs to maintain bounded control, and incorporates obstacle avoidance capabilities – all while preserving asymptotic stability of the closed-loop system. Additionally, we introduce a time re-scaling strategy that modifies trajectory evolution, ensuring target positions remain within a defined vicinity of the robot. The system performance was assessed in simulation and in 26 field trials inside an apple orchard using different obstacle configurations, weather, and terrain conditions, with a success rate of 100% and an average tracking error of 0.1 m. Note to Practitioners—The proposed shared control approach is developed for use with a differentially steered Uncrewed Ground Vehicle (UGV) with first order kinematic constraints and can be adapted to different indoor and outdoor scenarios. The environment in which the UGV is deployed should be mapped in advance to create a reference trajectory for the UGV to follow. If the location of obstacles is not known in advance, an obstacle detection and tracking system, as well as an online mapping system, must be developed. A simulated model of the UGV and the environment are useful for determining initial values of the shared control gains that can be further tuned when the physical platform is first deployed. In GPS-denied scenarios, a Simultaneous Localization and Mapping (SLAM) system must be implemented; a lidar-inertial based SLAM system is recommended. A force-reflexive joystick is ideal for sensitive human input. In addition, the communication between the UGV and the base station (where the human operator supervises the UGV and provides commands to it) should have minimal time delays, not exceeding typical human reaction time (ca. 0.25 s).}, archive = {J_TASE}, author = {Cheikh Melainine El Bou and Michele Focchi and Michael R. Chang and Marco Camurri and Karl D. von Ellenrieder}, doi = {10.1109/TASE.2025.3554368}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13603--13620}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Smooth Human–Robot shared control for autonomous orchard monitoring with UGVs}, volume = {22}, year = {2025}, }
• (2025). A replay-attack-resilient power system state estimation scheme. TASE, 22, 13588--13602. (www)

  @article{ , abstract = {Replaying outdated or uncorrelated data can lead to a distorted view of the actual system state, resulting in suboptimal control decisions that potentially compromise grid stability, reliability, and efficiency. Replay Attacks (RA) are a kind of stealthy attacks which can mar one of the very key application of the Energy Management System (EMS), viz., Power System State Estimation (PSSE). This paper, therefore, proposes a novel scheme to make the PSSE resilient against RAs. To this end, first, a Power Transfer Distribution Factor (PTDF)-assisted vulnerability analysis is carried out to identify the critical SCADA measurements which may become a preferred choice of the attacker to launch RAs. Next, an optimal number of secured phasor measurements are exploited to detect and correct any RA in the SCADA measurement set, followed by a simple hybrid estimation scheme to reconstruct the falsified set of measurements. The proposed strategy’s effectiveness is validated through testing on three standard IEEE test systems, namely IEEE 14, New England (NE) 39, and IEEE 118 using simulated data generated from Real-time Digital Simulators (RTDS) and MATPOWER. Finally, the efficacy, feasibility and robustness of the proposed method under different variant of RA scenarios are compared with two existing Kalman Filter-based distributed state estimation methods associated with the three widely used statistical cyber attack detectors. Note to Practitioners—Power System State Estimation (PSSE) is a key application in the EMS, which further caters to many important real-time grid applications. If the SCADA measurements, which are fed to the PSSE program get corrupted with the Replay Attacks (RAs), it can jeopardize the smooth functioning of the power grids. Being motivated by this, this article presents a simple approach to detect and mitigate RAs in PSSE. The proposed scheme exploits a set of limited secured phasor measurements, along with the vulnerable SCADA measurements, in the form of Hybrid State Estimation (HYB-SE) to detect and correct any presence of RAs. The proposal is validated on the IEEE 14, NE 39, and IEEE 18 bus systems, modelled in Real Time Digital Simulator with a maximum detection accuracy of 94.6%, 94.2%, and 84.47% respectively. The superior performance of the proposed detection approach is also validated against two existing distributed state estimation methods accompanied with three statistical cyber attack detectors.}, archive = {J_TASE}, author = {Sourav De and Ranjana Sodhi}, doi = {10.1109/TASE.2025.3555488}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13588--13602}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A replay-attack-resilient power system state estimation scheme}, volume = {22}, year = {2025}, }
• (2025). Strategic resource allocation in dual-channel healthcare systems: The role of telemedicine in physician assignment. TASE, 22, 13569--13587. (www)

  @article{ , abstract = {Asynchronous telemedicine offers flexibility and accessibility to address the problem of insufficient and unevenly distributed healthcare resources. However, hospitals lack managerial guidance for resource allocation decisions across online and offline services. This study considers a hospital operating both offline and online channels, which aims to achieve economic objectives through channel capacity allocation and workload distribution between working- and after-hours. Patient medical processes are formulated as a queuing network, integrating a multi-type patient choice model. We collected related datasets from our collaborative hospital, performed simulations to validate the modeling assumptions and conducted an empirical study. Empirical results indicate that e-visits crowd out offline first-visits and substitute for offline return-visits, and the strength of the substitution effect decreases with an increase in treatment complexity. Our results suggest that increasing offline channel capacity or enhancing compensation for after-hours telemedicine services would bring a heightened preference for in-person visits during working-hours and e-visit during after-hours, which might help with economic objectives. Our analysis indicates that indiscriminate expansion of online channels may be ineffective. Instead, departmental treatment complexity should be considered, and we propose three scenario-dependent allocation strategies. This study helps hospitals make scientific decisions about developing online channels and guiding physicians’ workload allocation. Note to Practitioners—Our work focuses on improving the management of new asynchronous telemedicine, such as Internet hospitals, with a primary focus on both channel capacity planning and dual-channel workload allocation. In this paper, we propose a dual-channel queuing network model, a channel capacity allocation strategy and an adjustment rule for workload allocation. This paper provides a case study of a representative general hospital. The results by department quantify the impact of Internet hospitals on outpatient healthcare utilization, make recommendations on workload adjustment, and show how to use the modeling tools in this paper to obtain an allocation strategy for channel capacity.}, archive = {J_TASE}, author = {Ziyang Wang and Yipei Wang and Yunzhe Qiu and Changxiao Jin}, doi = {10.1109/TASE.2025.3549822}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13569--13587}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Strategic resource allocation in dual-channel healthcare systems: The role of telemedicine in physician assignment}, volume = {22}, year = {2025}, }
• (2025). Explainable and safety aware deep reinforcement learning-based control of nonlinear discrete-time systems using neural network gradient decomposition. TASE, 22, 13556--13568. (www)

  @article{ , abstract = {This paper presents an explainable deep-reinforcement learning (DRL)-based safety-aware optimal adaptive tracking (SOAT) scheme for a class of nonlinear discrete-time (DT) affine systems subject to state inequality constraints. The DRL-based SOAT utilizes a multilayer neural network (MNN)-based actor-critic to estimate the cost function and optimal policy while the MNN update laws are tuned both using the singular value decomposition (SVD) of activation function gradient in order to mitigate the vanishing gradient issue and safety-aware Bellman error at each layer. An approximate safety-aware optimal policy is developed using Karush-Kuhn–Tucker (KKT) conditions by incorporating the higher-order control barrier function (HOCBF) into the Hamiltonian through the Lagrangian multiplier. The resulting safety-aware Bellman error helps with safe exploration both during online learning phase and at steady state without any explicit actor-critic MNN update law changes. To study the explainability and gain insights, we employ the Shapley Additive Explanations (SHAP) method to construct an explainer model for the DRL-based SOAT scheme in order to identify the important features in determining the optimal policy. The overall stability is established. Finally, the effectiveness of the proposed method is demonstrated on Shipboard Power Systems (SPS), achieving over a 35% reduction in cumulative cost compared to the existing actor-critic MNN optimal control policy. Note to Practitioners—In practical control systems, meeting safety constraints is often critical since ignoring constraints can lead to degraded performance or damage to equipment. This paper addresses the challenge of a safe DRL-based control approach that not only optimizes performance but also integrates robust safety assurances. Our DRL-based SOAT scheme specifically targets nonlinear discrete-time systems that must satisfy state inequality constraints. The successful proposed control performance in simulations on a Shipboard Power System demonstrates the potential for practical applications. DRL-based SOAT employs an MNN with an actor-critic framework for continuous learning and policy adaptation. Integrating HOCBFs directly into the optimization ensures safe operation, even during online learning, which is critical for real-time applications. The addition of SHAP enhances transparency by identifying key features that influence control decisions. Future work could adapt this framework to other constrained environments, such as autonomous vehicles, robotics, and industrial automation, where safety, optimality, and explainability are essential.}, archive = {J_TASE}, author = {Behzad Farzanegan and Sarangapani Jagannathan}, doi = {10.1109/TASE.2025.3554431}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13556--13568}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Explainable and safety aware deep reinforcement learning-based control of nonlinear discrete-time systems using neural network gradient decomposition}, volume = {22}, year = {2025}, }
• (2025). Extended receptive field UDA semantic segmentation based on spatial alignment and knowledge distillation. TASE, 22, 13544--13555. (www)

  @article{ , abstract = {In recent years, unsupervised domain adaptation (UDA) has significantly advanced, addressing the issue of requiring large amounts of labeled data in deep learning. Some UDA strategies have effectively alleviated domain shift, but they are still struggling to tackle challenges such as the model’s inability to continuously extract features, the difficulty to achieve better segmentation boundaries in the target domain, and tend to neglect previously acquired knowledge. To address these issues, we propose an extended receptive field UDA semantic segmentation based on spatial alignment and knowledge distillation (ERF). Firstly, based on the idea of combining serial and parallel, we design a novel large continuous receptive field decoder (largeCF) to extract large continuous receptive field features. This approach alleviates the bias of model in feature extraction between objects of different sizes and simultaneously reduces model complexity. Secondly, we propose an edge consistency strategy that aligns edge features and matches the spatial arrangement between predicted and ground truth labels, improving edges segmentation accuracy of the target domain. Finally, we employ a knowledge distillation module to achieve an optimized student-teacher framework, where the teacher effectively guides the student to retain previously learned information, resulting in more accurate segmentation of the target domain. Experimental results demonstrated the effectiveness of the proposed approach, which achieved mIoU of 76.5% and 68.8% on UDA benchmark tasks GTA $\rightarrow $ CityScapes and SYNTHIA $\rightarrow $ CityScapes, respectively. The code is available at: https://github.com/fz-ss/ERF. Note to Practitioners—This paper focuses on the challenges of semantic segmentation in autonomous driving, particularly facing the issue of extensive manual annotation required for dense semantic labels. We propose a novel extended receptive field UDA semantic segmentation based on spatial alignment and knowledge distillation. The article begins by outlining the initial implementation process of UDA, laying the foundation for subsequent in-depth discussions. Subsequently, we conduct a theoretical analysis of the Large Continuous Decoder, Boundary Consistency Strategy, and Knowledge Distillation Scheme, which constitute the core components of our method. Finally, experiments on two UDA benchmark tasks demonstrate the feasibility of our approach. However, the performance gap between UDA and supervised semantic segmentation still exists. In future research, we will focus on reducing the feature gap between different domains. Additionally, we will strive to fully use image features and distilled features to make greater progress, thereby driving advancements in semantic segmentation for autonomous driving.}, archive = {J_TASE}, author = {Yunna Song and Jinlong Shi and Caisheng Liu and Suqin Bai and Yu Yuan and Xin Shu and Qiang Qian and Dan Xu and Yunhan Sun}, doi = {10.1109/TASE.2025.3554780}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13544--13555}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Extended receptive field UDA semantic segmentation based on spatial alignment and knowledge distillation}, volume = {22}, year = {2025}, }
• (2025). A novel GA-CP method for fixed-type multi-robot collaborative scheduling in flexible job shop. TASE, 22, 13531--13543. (www)

  @article{ , abstract = {With the rapid development of intelligent manufacturing, multi-robot collaborative systems are increasingly integrated into various production processes. In the flexible job shop environment of automotive stamping, achieving smooth operation and efficient manufacturing of production lines hinges on solving the critical issues of multi-robot task allocation and scheduling. However, for such fixed-type multi-robot collaboration problems, robots are constrained by specific areas or predetermined trajectories, and processing times can only be adjusted by varying the number of available robots. Therefore, the scheduling problem in multi-robot collaborative flexible job shop problems (MCFJSP) is divided into two sub-problems: FJSP with controllable processing times and multi-robot collaborative task balancing. To address these, we propose three distinct methods: mixed integer linear programming (MILP), constraint programming (CP), and a hybrid genetic algorithm-constraint programming (GA-CP). Finally, a set of 48 benchmark cases and two real-world cases are developed to test these methods. Comparative experiments demonstrate that the MILP model is superior in small-scale cases, while the GA-CP model exhibits the best overall performance in medium to large-scale cases. Furthermore, through comparisons with two advanced algorithms, the effectiveness and superiority of the GA-CP method in addressing real-world cases are confirmed.[8pt]Note to Practitioners—In modern manufacturing environments, particularly in industries like automotive manufacturing, multiple robots working together on complex tasks are increasingly common. This paper addresses the practical challenge of effectively scheduling these robots to maximize efficiency while reducing the number of robots assigned to each task. This study introduces and compares different methods, including MILP, CP, and GA-CP methods, that can help practitioners determine the best way to allocate tasks among robots and schedule them efficiently. For example, in small-scale tasks, the MILP model can quickly provide the best solution. However, as the complexity and scale of the task increase, the GA-CP method becomes more practical, offering high-quality solutions within a reasonable timeframe. The study provides actionable insights that can be applied directly to real-world production scenarios, helping practitioners in industries like automotive stamping to maximize job shop productivity while reducing energy consumption losses in robot processing.}, archive = {J_TASE}, author = {Jin Huang and Xinyu Li and Liang Gao}, doi = {10.1109/TASE.2025.3554019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13531--13543}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel GA-CP method for fixed-type multi-robot collaborative scheduling in flexible job shop}, volume = {22}, year = {2025}, }
• (2025). Real-time obstacle detection for a biped robot climbing on exoskeleton-structure glass windows. TASE, 22, 13519--13530. (www)

  @article{ , abstract = {This paper presents a novel real-time obstacle detection method for a biped climbing robot. It is composed of a modified UNet neural network combined with a large-field separable contextual integration (LSCI) module and machine vision techniques for line detection. Such an approach allows the biped robot to accurately detect and locate metal exoskeleton-structure obstacles on glass curtain walls. Moreover, the algorithm adopts point cloud data collected from a depth camera to ensure precise identification of target exoskeleton-structure frames including the positions and tilt angles, facilitating effective real-time collision-free motion planning for the cleaning robot system. Comparative experiments with other neural networks used for image segmentation demonstrate that the proposed LSCI-UNet architecture significantly improves the frame obstacle detection accuracy on exoskeleton-structure glass windows with limited data. The reported method enables the biped cleaning robots to detect obstacles in real-time and effectively navigate on glass walls with exoskeleton-structure beams. Note to Practitioners—Cleaning glass windows of high-altitude buildings imposes significant risks to human workers due to potential falls, which inspires the deployment of window-cleaning robots for such tasks. Bipedal robots exhibit an ability to navigate inclined surfaces of complex exoskeleton structures. Meanwhile, accurate detection and positioning of metal beam obstacles on the windows is crucial for efficient robotic operations in complex environments. However, detecting such obstacles on transparent glass curtain walls remains a considerable challenge. Particularly, the objects behind the glass curtain cause confusion in obstacle detection. To overcome such issue, this paper proposes a new approach of real-time obstacle detection of a biped climbing robot for cleaning glass windows on an exoskeleton-structure building. The effectiveness of the reported methodology has been validated by conducting experimental investigations.}, archive = {J_TASE}, author = {Zikang Li and Weijian Zhang and Yonghao Huang and Lap Mou Tam and Qingsong Xu}, doi = {10.1109/TASE.2025.3554653}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13519--13530}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time obstacle detection for a biped robot climbing on exoskeleton-structure glass windows}, volume = {22}, year = {2025}, }
• (2025). An adversarial branched network for degradation modeling under multiple failure modes. TASE, 22, 13501--13518. (www)

  @article{ , abstract = {Failure Mode (FM) diagnosis and Remaining Useful Life (RUL) prediction are two major tasks in prognostics health management. Numerous data fusion-based methods have been employed to construct the Health Index (HI) derived from multiple sensors that provide a holistic view of system degradation status to make prognosis, allowing for early detection of anomalies and predictive maintenance actions. However, few studies have focused on constructing the HI to capture diverse degradation characteristics in the context of multiple FMs. To address this issue, this paper proposes an adversarial branched model to achieve FM diagnosis and RUL prediction. We first establish an unsupervised branched Deep Neural Network (DNN) to construct the HIs of units under multiple FMs by fully considering degradation properties. Then, an adversarial training strategy is developed based on data of historical units to capture diverse degradation properties of different FMs. Finally, we diagnose the FM of an in-service unit via a similarity measurement method based on the constructed HIs. Given the diagnosed FM and corresponding HI, the RUL of the in-service unit is predicted. A simulation study and a case study on the degradation of aircraft gas turbine engines are presented to evaluate the performance of the proposed method. Note to Practitioners—The paper aims to develop an adversarial branched model to construct the HI under multiple FMs for failure diagnosis and RUL prediction of operating units. Specifically, the developed method addresses a challenging issue in practice, i.e., how to construct an HI from multiple sensor signals that can effectively extract distinguishable degradation features under multiple FMs. To implement this method in practice, four steps are included as follows: First, collect multiple run-to-failure sensor signals and FMs of historical units. Second, establish the branched DNN to construct the HI for multiple FMs. Third, train the branched DNN via the adversarial algorithm using data of historical units. Fourth, diagnose the FM and predict the RUL based on the HI for in-service units. As a deep learning model, the branched DNN is expected to be applicable to a large number of scenarios and practical cases, especially for manufacturing systems with complex structures and multiple FMs.}, archive = {J_TASE}, author = {Di Wang and Xinyu Pan and Ershun Pan and Yanting Li}, doi = {10.1109/TASE.2025.3549061}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13501--13518}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adversarial branched network for degradation modeling under multiple failure modes}, volume = {22}, year = {2025}, }
• (2025). Distributed and reactive controller synthesis for multi-agent systems under finite horizon temporal logic tasks. TASE, 22, 13485--13500. (www)

  @article{ , abstract = {Automated synthesis of local controllers for multi-agent systems to satisfy complex task specifications has attracted extensive attention. However, it remains unclear how to formally guarantee that their composed behavior satisfies the specified global task. In this paper, we aim to synthesize an automated and distributed control strategy for fleet-wise tasks specified as linear temporal logic formulas, such that the agents act asynchronously and synchronize only on shared actions via local coordination. The proposed method consists of three main steps. First, the set of satisfying global control strategies is computed via parallel composition. Among these strategies the conditions for decomposability are evaluated, based on which the global strategy over a maximum synchronization scheme is found. Then the synchronization scheme is further refined to obtain more efficient local control strategies. It is formally proven that the resulting global behavior along with the synchronization scheme satisfies the specified global task. Also, the synthesized local controllers are reactive to changes in the workspace or in the fleet during online execution, without the need for replanning. Thus, collaborative relations among the agents are adaptive as needed. Numerical simulations and hardware experiments are conducted for nontrivial scenarios. Note to Practitioners—This paper was motivated by the problem of coordinating a fleet of autonomous robots in collaborative search and delivery processes, where local controllers are synthesized for each robot such that the specified global task specification is satisfied. Existing approaches to address such problems often rely on a fully-connected communication topology and fixed collaborative relations among the robots, thus limiting efficiency of the multi-robot execution and yielding difficulty of coordination. Also, most of existing approaches cannot directly deal with the changes in the workspace or in the fleet during task execution unless a task replanning is performed, which leads to longer time of system-wide communication and task completion, yielding failures during online execution. In this paper, we propose a novel distributed control architecture to tackle these issues, where the robots are controlled to operate asynchronously and achieve online local coordination by synchronization on collaborative actions. Moreover, compared with the common solutions, the collaborative relations among the robots are formed and removed dynamically as needed and the robots are robust to uncertainty during task execution. It is formally proven that the resulted global behavior of the robot fleet along with the synchronization scheme is consistent with the specified global task. We have shown that it is particularly useful for complex and coupled multi-robot applications, where the inter-robot collaborations are feasible and local. Experimental results suggest that this approach is applicable to multi-robot systems which greatly improves the concurrency and efficiency of task execution. In the future research, we will draw inspiration from decentralized approaches for task decomposition to alleviate the computational burden as the number of robots increases.}, archive = {J_TASE}, author = {Yuchong Huang and Ning Xu and Meng Guo and Jie Li and Lincheng Shen}, doi = {10.1109/TASE.2025.3553498}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13485--13500}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed and reactive controller synthesis for multi-agent systems under finite horizon temporal logic tasks}, volume = {22}, year = {2025}, }
• (2025). Concept and strategies: Equivalent predictive control and handle point control for bipedal-vehicle transformable robots under various disturbances. TASE, 22, 13470--13484. (www)

  @article{ , abstract = {Bipedal-vehicle transformable robots (BVTRs), equipped with driving wheels, combine the flexibility of bipedal locomotion with the speed of wheeled movement. However, maintaining balance across different formations under various external disturbances remains a significant challenge due to uncertain disturbance types and dynamic shifts between formations. To address these challenges, this paper introduces the concept of Equivalent Predictive Control (EPC), which models all disturbances as unified virtual wrenches and integrates them directly into the robot’s predictive control model, treated as an inertia-varying single rigid body. By anticipating the future impact of disturbances, EPC enhances stability and enables simultaneous handling of various disturbances. To address the challenge of dynamic changes, contact variations, and shifting constraints during formation transitions, we propose Handle Point Control (HPC). HPC simplifies multi-task tracking by reducing joint space control to a set of virtual target points, called ‘handle points’, such as knees, hips, and shoulders. This method facilitates real-time formation switching by tracking different handle points. Experiments on the BVTR platform BHR8-2 validate the effectiveness of the proposed control strategies. Note to Practitioners—This paper addresses two critical challenges for applying BVTRs in real-world industrial scenarios: 1) managing various external disturbances, and 2) overcoming the complexities associated with changing dynamics and contact situations during formation transitions. The proposed EPC strategy unifies all disturbance types and integrates them into the robot’s dynamic model, enhancing stability and adaptability across formations. The HPC method simplifies control by focusing on tracking key handle points, allowing smooth, real-time formation transitions without needing multiple optimization schemes. These methods can also be applied to other robotic systems that face similar control challenges.}, archive = {J_TASE}, author = {Chencheng Dong and Zhangguo Yu and Xuechao Chen and Junhang Lai and Jiayi Liu and Chao Li and Qiang Huang}, doi = {10.1109/TASE.2025.3552429}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13470--13484}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Concept and strategies: Equivalent predictive control and handle point control for bipedal-vehicle transformable robots under various disturbances}, volume = {22}, year = {2025}, }
• (2025). Distributed asynchronous optimization with inseparable coupled constraints and its application. TASE, 22, 13458--13469. (www)

  @article{ , abstract = {Considering the complexity of centralized plant-wide optimization and the presence of communication delay in production units, a distributed asynchronous optimization framework is developed for energy consumption optimization problem during ethylene production. First, the energy consumption optimization problem of ethylene process is formulated into a distributed asynchronous optimization problem. Then, considering multiple production units and the material transfer time involved, each unit is treated as a node, and each node is decomposed into calculation nodes, constraint nodes, and delay nodes. Specifically, each production unit can be optimized asynchronously without the necessity for synchronization. To overcome the impact of communication delay on asynchronous process, state vectors are incorporated into the distributed parameter projection algorithm. Additionally, to ensure that the inseparable coupled constraints between nodes during asynchronous operations are met, the parameter projection algorithm is employed. Numerical experiments and industrial simulations demonstrate the proposed algorithm exhibits faster convergence speeds compared to distributed synchronous algorithms, and the energy consumption is lower than the results obtained by the centralized algorithm. Note to Practitioners—As the scale of ethylene production expands, traditional centralized optimization methods struggle to meet the demands of plant-wide optimization due to their high model complexity and slow convergence rates. In this study, we develop a distributed optimization framework, in which the large-scale complex model is decomposed into small parts, and the global optimization task is accomplished cooperatively through local information exchange. Furthermore, considering the time delay due to material residence time in processing units, a distributed asynchronous parameter projection algorithm is proposed to solve the energy consumption issue in ethylene plant. Experimental results demonstrate that the proposed method is capable of converging to the feasible solution in the presence of time delay. Compared with the existing centralized methods, the proposed asynchronous distributed algorithm exhibits better performance, i.e. lower energy consumption. Besides, the proposed asynchronous distributed optimization framework holds potential for application in other process industries, such as the metallurgical industry and steel production process.}, archive = {J_TASE}, author = {Ting Wang and Weihan Liu and Zhongmei Li and Zhencheng Ye and Xin Peng and Wenli Du}, doi = {10.1109/TASE.2025.3550740}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13458--13469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed asynchronous optimization with inseparable coupled constraints and its application}, volume = {22}, year = {2025}, }
• (2025). Coordinated operation optimization of grid-interactive residential buildings based on neural network-assisted hierarchical model predictive control. TASE, 22, 13441--13457. (www)

  @article{ , abstract = {The coordinated operation of grid-interactive buildings contributes to creating a more resilient and reliable power grid. However, existing studies fail to identify the demand changes of each building resulting from their coordinated participation in providing grid services, which affects the economic compensation of each building and its willingness to coordinate. In this article, we investigate an optimal coordinated operation problem for grid-interactive residential buildings (GRBs) while considering generation capacity services and economic compensation for participating GRBs. Specifically, we first formulate two optimization problems to capture the different objectives of GRBs during non-service periods and grid-service periods, respectively. Then, we develop a physically consistent neural network (PCNN)-assisted hierarchical model predictive control (HMPC)-based GRB energy management algorithm to solve the optimization problem during non-service periods. Next, we propose a coordinated operation algorithm to solve the optimization problem during grid-service periods based on PCNN-assisted HMPC and rule-assisted binary search. By comparing the initial solutions from the proposed energy management algorithm with the final solutions generated by the proposed coordination algorithm, the demand changes of each GRB during service periods can be identified. Simulation results indicate that the proposed coordination algorithm achieves up to 37.8114% lower energy costs and 82.1459% better grid service performance than benchmarks while maintaining high thermal comfort. Note to Practitioners—Buildings with distributed energy resources (e.g., solar generation, energy storage) and flexible loads (e.g., heating, ventilation, and air conditioning (HVAC) systems) have significant potential to provide grid services, such as voltage support, frequency regulation, and power peak reduction. Since a single residential building contributes minimally to service quality, multi-building coordination through a trusted third party is necessary. However, since participation in providing grid services may affect occupant comfort and building energy costs, the demand change of each residential building should be identified so that the corresponding economic compensation can be calculated, which is a key factor for the successful deployment of such grid-interactive residential buildings (GRBs). To this end, we develop an optimal energy management algorithm for each residential building during non-service periods, which aims to minimize building energy cost while maintaining high occupant comfort. Based on the developed energy management algorithm, a coordinated operation algorithm for service periods is further proposed to limit the peak demand below a value predetermined by the system operator. By comparing the initial decisions from the energy management algorithm with the final decisions generated by the proposed coordination algorithm, we can identify the demand change of each building during service periods. Numerical results demonstrate that the proposed coordinated operation algorithm can help participating GRBs reduce energy costs and enhance service performance for power grids, with negligible sacrifice to occupant comfort.}, archive = {J_TASE}, author = {Liang Yu and Zhiqiang Chen and Dong Yue and Yujian Ye and Goran Strbac and Yi Wang}, doi = {10.1109/TASE.2025.3551649}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13441--13457}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Coordinated operation optimization of grid-interactive residential buildings based on neural network-assisted hierarchical model predictive control}, volume = {22}, year = {2025}, }
• (2025). Improved intermediate estimator-based fault tolerant control for HFVs with heterogeneous actuators. TASE, 22, 13429--13440. (www)

  @article{ , abstract = {This study proposes a compound fault-tolerant control for re-entry hypersonic flight vehicles (HFVs) equipped with heterogeneous actuators in the presence of unknown disturbances, sensor faults, and aerodynamic surface constraints. The pitch rate sensor faults and external disturbances are estimated by an improved intermediate estimator (IIE), and then can update a sliding mode controller (SMC). Through a quadratic programming method to optimize the virtual control torque calculated by the adaptive law and super-twisting reaching law, the aerodynamic surfaces (AS) can achieve optimal deflections under rudder limits. Employing a fuzzy optimization algorithm with Aquila Optimizer (AO), reaction control systems (RCS) can compensate for the insufficient control torque produced by AS owing to the low dynamic pressure in the re-entry phase of HFVs. Comparative simulation results demonstrate that the proposed scheme can restore the accurate track of the desired attitude. Note to Practitioners—Compared to the conventional flight vehicles, the HFVs have some salient characteristics such as high nonlinearity, strong coupling, and complex uncertainty, which bring great challenges to their flight controller designs. Among these challenges, the sensor faults and external disturbances may trigger inaccurate attitude tracking of the HFVs and the aerodynamic surface constraints would let the actuators cannot produce enough supporting control torque. To address these problems, this study introduces a compound fault-tolerant control strategy designed for re-entry HFVs with heterogeneous actuators. An intermediate estimator is designed to handle the pitch sensor faults and external disturbances simultaneously, where the estimation information is compensated in a sliding mode controller updated by adaptive law and super-twisting reaching law. By separately designing the control allocation scheme to distribute torque to the heterogeneous actuators, RCS can offset the negative effect where AS cannot provide the sufficient control torque because of the low dynamic pressure in the re-entry phase. The simulation results show that the proposed scheme achieves satisfactory attitude accuracy.}, archive = {J_TASE}, author = {Zikun Xiang and Lifu Du and Wei Chen and Fuyang Chen}, doi = {10.1109/TASE.2025.3548666}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13429--13440}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Improved intermediate estimator-based fault tolerant control for HFVs with heterogeneous actuators}, volume = {22}, year = {2025}, }
• (2025). DBAN: Double bias adjustment network for domain shift defect detection in photovoltaic intelligent manufacturing. TASE, 22, 13411--13428. (www)

  @article{ , abstract = {Developing reliable, generalized, and accurate defect detection technology for photovoltaic (PV) manufacturers is particularly critical with the demand for production line expansion. Existing technologies perform well when handling independently and identically distributed (IID) data. However, their performance significantly reduces when they encounter domain shift problems, including style and instance bias, prompted by production line expansion. In this paper, we propose a novel Dual Bias Adjustment Network (DBAN) to enhance the generalization and reliability of PV defect detection. Specifically, we construct a Global Style-Generalized Contrastive Learning (GSCL), which uses nonlinear transformation functions and contrastive learning strategies to enhance model adaptability for different style changes and global discriminative ability, effectively overcoming the style bias problem. We design a Test-Time Prototype Adjustment (TTPA) that employs graph methods and prototype learning to adjust feature representations accurately. TTPA enhances prediction reliability during testing via dynamic prototype repositories and memory mechanisms, effectively addressing instance deviations. We conduct comprehensive experiments proving that DBAN achieves optimal performance, surpassing other advanced algorithms. Moreover, GSCL and TTPA add little inference time to the model, making them suitable for practical industrial applications. Finally, we conduct extensive experiments on the public domain-shifted PV dataset ELES, where our model achieves state-of-the-art performance in Single-domain generalized object detection. Note to Practitioners—This work proposes a practical defect detection solution, DBAN, enabling PV manufacturers to maintain reliable quality control across multiple production lines. The model can be directly integrated into existing inspection systems by deploying DBAN software on a central processing server that receives EL images from production lines. Trained with historical defect data from one production line, the model can automatically inspect products from expanded lines and continuously update its detection capabilities using real-time test data without retraining. DBAN enables practitioners to monitor multiple lines through a unified interface while maintaining high-quality standards, thus reducing equipment and labor costs. The system continuously enhances detection accuracy under varying manufacturing conditions, making it especially valuable for manufacturers seeking efficient, scalable production while ensuring product quality.}, archive = {J_TASE}, author = {Shenshen Zhao and Haiyong Chen and Chuhan Wang and Kun Liu}, doi = {10.1109/TASE.2025.3554328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13411--13428}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DBAN: Double bias adjustment network for domain shift defect detection in photovoltaic intelligent manufacturing}, volume = {22}, year = {2025}, }
• (2025). Modeling of asynchronous mode-dependent delays in stochastic markovian jumping modes based on static neural networks for robotic manipulators. TASE, 22, 13398--13410. (www)

  @article{ , abstract = {Over the past 20 years, many specialists in the business have become interested in managing flexible joint robotics. Considering flexibility in joints in the control scheme of inflexible robots is challenging because it eliminates certain structural properties that facilitate control, such as full activation, distinct control for every socket, and passivity of the torque generated by the motor for coupling velocity. In this research, the authors analyze asymptotic stability for neural models of fixed stochastic process that have time-varying delays, which depend on jumping modes; their derivatives no longer need to be smaller than one. Secondly, by applying a stochastic setup, the researchers explore fixed neutral models with Markovian jump variables based upon mode-dependent time-varying delays for modeling flexible joint robots. Thirdly, this research develops various stability circumstances with delay-dependent conditions based on linear matrix inequalities (LMIs) by combining the novel Lyapunov theory with the convex polyhedron technique, concluding with two numerical examples illustrating their use. As a result, an optimization problem can be addressed to calculate a control law created for implementing the stochastically stable platform. As a final step, the proposed method is verified using a quadruple-tank model and for modeling flexible joint robots. Note to Practitioners—Neural network controllers are frequently employed in the medical field, automation industry, robotics sector, and other industries where certain jobs cannot be completed by autonomous or human robotics. By increasing flexibility, refining control technique, relaxing stability conditions, and promoting human-robot collaboration, this research seeks to create novel solutions to important problems in industrial robotic manipulators. In order to maintain the system’s stability in both motion states-coupling velocity and torque motion and meet the associated position and velocity synchronization goals, this paper suggests a NN-based control technique that can manage random process and variable time-delays signals. More straightforward structures, fewer tuning factors, and improved control behaviors are some of the benefits of this approach that make engineering applications easier. Lastly, a simulation using a realistic strategy is demonstrated to validate the theoretical assertions.}, archive = {J_TASE}, author = {Summera Shamrooz and Muhammad Shamrooz Aslam and Houguang Liu and Hazrat Bilal and Athanasios V. Vasilakos}, doi = {10.1109/TASE.2025.3552645}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13398--13410}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling of asynchronous mode-dependent delays in stochastic markovian jumping modes based on static neural networks for robotic manipulators}, volume = {22}, year = {2025}, }
• (2025). Disturbance compensation control for humanoid robot hand driven by tendon-sheath based on disturbance observer. TASE, 22, 13387--13397. (www)

  @article{ , abstract = {The operation accuracy of humanoid robot hands driven by the tendon sheath will be reduced due to the influence of friction torque during rotation, which is not conducive to the dexterous operations of humanoid robot hands. In order to improve the control accuracy of humanoid robot hands, this paper proposes a control strategy based on the disturbance observer compensation, which eliminates the external disturbance torque by compensating the friction torque. Firstly, this article proposes the mechanical structure of humanoid robot hands driven by the tendon sheath with 19 degrees of freedom (DOF). This humanoid robot hands can grasp most irregular objects. Next, the dynamic model of humanoid robot hands’ drive systems is established based on the tendon sheath transmission theory. The driving system’s dynamic model reveals the influence of friction torque on the motion accuracy of the humanoid robot hand. Then, the disturbance observer (DOB) is designed based on the robust stability theorem. The DOB is used to improve the control accuracy of the driving system, thereby improving the operational accuracy of humanoid robot hands. Finally, this article conducts simulation rotation tracking control and prototype grasping control experiments on humanoid robot hands. The experimental results show that the proposed control strategy based on disturbance compensation can effectively improve the operational accuracy of humanoid robot hands. Note to Practitioners—This paper proposes a humanoid hand with 19 degrees of freedom based on the tendon-driven theory and applies the cable theory for its dynamic modeling. To address the issue of decreased precision caused by friction in practical operations, a friction compensation control strategy based on disturbance observer is proposed in this study. This control strategy improves the motion accuracy and stability of the mechanical hand. Finally, the effectiveness of the proposed control strategy is demonstrated through numerical simulation and experimental validation.}, archive = {J_TASE}, author = {Meng Yin and Haozhe Wang and Dongyang Shang and Ming Li and Tiantian Xu and Xinyu Wu}, doi = {10.1109/TASE.2025.3540797}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13387--13397}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Disturbance compensation control for humanoid robot hand driven by tendon-sheath based on disturbance observer}, volume = {22}, year = {2025}, }
• (2025). Target-oriented environmental context modeling for pedestrian risky behavior detection. TASE, 22, 13373--13386. (www)

  @article{ , abstract = {In autonomous driving systems, detecting pedestrian risky behavior is crucial for ensuring the safety of human-vehicle interactions. The behavior of pedestrians is not only determined by their individual actions but also influenced by their interactions with the surrounding environment. Considering the impact of environmental context on pedestrian behavior, recent studies have proposed various methods for extracting contextual information. However, efficiently modeling the environmental contextual features of pedestrian behavior remains a challenge. In this paper, we propose a target-oriented environmental context modeling method that accounts for the role of target-specific features in constructing context features, achieving target-adaptive context feature extraction. Our solution, called State DEtection TRansformer (SDETR), provides an end-to-end framework for risky pedestrian behavior detection. First, we devise a Dual-level Feature Encoder that effectively decouples high-level target semantic and low-level environmental texture. Specifically, the texture encoding enables label-free environmental feature extraction. Then, we develop a Object-Environment Perception Decoder that flexibly decodes cross-feature domain contextual features using object features. Finally, a Feature Fusion Head is employed to merge object features with environmental state features for the detection output. Experiments demonstrate the outstanding performance of SDETR in two typical risky behavior detection tasks (crossing detection and intrusion detection). We report a new record of 87.3% accuracy on the JAAD dataset and 76.3% accuracy on the Cityintrusion dataset, which significantly outperforms all previously published results. Note to Practitioners—The motivation of this paper is to detect risky pedestrian behavior in traffic scenes, with a particular focus on the construction of pedestrian environmental context. Existing methods for extracting environmental context typically involve complex feature extraction or label descriptions of the background, followed by a mechanistic establishment of the interplay between pedestrians and their surroundings. This paper proposes a flexible and cost-effective method for modeling environmental context. We employ an attention mechanism to autonomously decode cross-feature domain contextual state features, using target features as query features. Moreover, we employ dual-level feature extraction methods for targets and backgrounds (target semantics and background textures), significantly reducing the labeling cost for environmental description. Preliminary experiments suggest that this method is feasible in the detection of two risky pedestrian behaviors: pedestrian crossing and pedestrian intrusion. However, it has not yet been extended to other pedestrian dangerous behavior tasks. In future research, we intend to delve deeper into the recognition of unlabeled pedestrian abnormal and risky behaviors, expanding our research beyond the current scope.}, archive = {J_TASE}, author = {Zhenyu Shi and Shibo He and Meng Zhang and Kun Shi}, doi = {10.1109/TASE.2025.3553495}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13373--13386}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Target-oriented environmental context modeling for pedestrian risky behavior detection}, volume = {22}, year = {2025}, }
• (2025). The multi-objective optimization problems of markovian jump systems closed by general controllers: A mode-separation approach. TASE, 22, 13359--13372. (www)

  @article{ , abstract = {This paper studies the multi-objective optimization problems (MOPs) of Markovian jump systems (MJSs) closed by general controllers. Firstly, the linear quadratic regulator (LQR) problem of MJSs closed by a general controller with sampled or synchronous modes is studied and estimated by a mode-separation approach, whose traditional algebraic Riccati equations (AREs) are removed. Meanwhile, two traditional situations about mode-dependent and -independent controllers are contained as special ones. Secondly, an MOP deeply depending on mode separations is proposed and solved by the non-dominated sorting whale optimization algorithm (NSWOA) such that the minimum value of LQR problem and the maximum expectation value of summed mode dwell times in the same mode separation are simultaneously optimized. Particularly, in order to further improve the estimation effect, a single-objective optimization problem (SOP) is presented and computed by applying the deep deterministic policy gradient (DDPG) technique, whose optimal controller in addition to its best mode separation is given in detail. Thirdly, similarly but more generally, another controller with its mode and state both sampled is constructed to realize the LQR problem, whose effects are better than some existing methods and can also be improved by solving some optimization problems. Finally, a simulation is used to verify the effectiveness and superiority of the proposed methods. Note to Practitioners—This paper addresses the LQR problem for MJSs, commonly used in automation and control with uncertainty and frequent mode switching. Traditional LQR methods rely on mode-dependent controllers and AREs, leading to high computational costs and real-time demands, especially in industrial settings. To overcome these challenges, this paper proposes a general controller design and an MOP, replacing AREs with a mode-separation strategy, simplifying the process and reducing costs. A controller with both mode and state sampling is also introduced to achieve the LQR objective in complex industrial environments. Although effective, further research is needed for its application in large-scale systems. Future work will focus on expanding its applicability.}, archive = {J_TASE}, author = {Guoliang Wang and Yaqiang Lyu}, doi = {10.1109/TASE.2025.3554684}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13359--13372}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The multi-objective optimization problems of markovian jump systems closed by general controllers: A mode-separation approach}, volume = {22}, year = {2025}, }
• (2025). Routing scheduling for valet-charging service with self-operating and third-party charging stations. TASE, 22, 13341--13358. (www)

  @article{ , abstract = {Many electric vehicle (EV) companies have begun providing energy replenishment services to EV owners. Valet-charging is one such service where valet-staff are dispatched to substitute for EV owners in charging their EVs. This paper considers a routing scheduling problem for the service. We explore the charging options available at both the service provider’s self-operating charging stations and third-party charging stations. We formulate the problem via a robust optimization framework to capture the uncertain waiting time at the third-party stations. A Partial-Branch-and-Price algorithm (PB&B) is designed to solve the problem to optimal efficiently for large scales. Specifically, we innovatively proposed acceleration methods to speed up our extended robust labelling algorithm within the PB&B. We conduct a case study to show the effect of uncertainty in detail and provide investment suggestions for the firm’s investment on whether to contract a third-party charging station. Note to Practitioners—The inconvenience of charging has received significant attention for EVs, and energy replenishment services offer an effective approach beyond infrastructure expansion. Valet-charging, first introduced by NIO Inc., followed by FAW-Volkswagen and NETA AUTO, represents a notable attempt in this direction. Valet-staff routing scheduling needs to consider station assignment decisions, which introduces another potential issue: the number of charging stations owned by valet-charging service providers is insufficient. Including third-party charging stations with uncertain waiting times complicates this issue further. We believe our work develops an effective tool for valet-charging service providers to handle routing scheduling with both types of stations and analyze the investment value on charging stations.}, archive = {J_TASE}, author = {Tianyi Zhao and Na Li}, doi = {10.1109/TASE.2025.3551799}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13341--13358}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Routing scheduling for valet-charging service with self-operating and third-party charging stations}, volume = {22}, year = {2025}, }
• (2025). Energy-optimal refactoring of multiple smart sensors with edge computing capability. TASE, 22, 13327--13340. (www)

  @article{ , abstract = {This paper studies the refactoring problem of multiple smart sensors (MSSs) with edge computing capability, where the program consisting of codes and driving MSSs distributes into all the smart sensors, the available energy for refactorings of MSSs is limited, and each smart sensor adopts binary refactoring mode, i.e., either performs refactoring locally by itself or fully offloads its codes to the edge server (ES) for edge refactoring. More energy strengthens MSSs to wipe out more code bugs (CBs), while corresponding to more energy consumption. Meanwhile. the refactoring mode (i.e., local or edge refactoring) also influences the CB ratio (CBR) and energy efficiency. Therefore, how to make the optimal tradeoff between CBR and energy consumption for such MSSs arises as an interesting issue. To address this issue, this paper first reveals a necessary and sufficient condition to judge whether the minimum CBR can be reached as well as its analytical expression, and then discloses the optimal refactoring time, refactoring computation rate, and refactoring mode, in terms of guaranteeing the minimum CBR under the energy constraint. An algorithm based on our discovered foundations is proposed for such MSSs to minimize the CBR within the acceptable level of energy consumption. Theoretical analysis, simulation and field experiments verify its performance. To our best knowledge, this is the initial work towards the optimal refactoring of MSSs. Note to Practitioners—For MSSs in practice, CBs lurking in their driving program often endanger their normal function. As a kind of behavior-preserving code transformation, the refactoring built in MSSs can help them to remove CBs. Such refactoring must consume energy and often suffers energy setback because the electricity and computing power of MSSs are usually very limited. To overcome this setback, the edge computing is introduced for MSSs to offload their codes to the ES for refactoring. However, the combination of MSSs and edge computing heavily challenges energy-saving refactoring, involving the invisible and unknown CBs and the curse of dimensionality regarding code-offloading. Accordingly, this work allows MSSs with edge computing capability to operate in a program-healthy and energy-efficient manner by making the optimal tradeoff between CBR and energy consumption, covering both theoretical results and algorithm. Specifically, a necessary and sufficient condition to judge whether the minimum CBR can be reached is disclosed, the analytical expression of that minimum CBR is derived out, the optimal refactoring time, refactoring computation rate, and refactoring mode for each smart sensor are discovered, and an effective algorithm for the practitioners to enjoy the minimum CBR while maintaining the energy consumption within a given range is further proposed. By controlling the refactoring time, refactoring computation rate, and refactoring mode given that the necessary and sufficient condition is satisfied, this work can be applicable for MSSs with edge computing capability to suffer the minimum CBR when they are employed to sense the physical world in the scenarios where their available energy is limited.}, archive = {J_TASE}, author = {Chen Hou and Syed Naeem Haider}, doi = {10.1109/TASE.2025.3550207}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13327--13340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy-optimal refactoring of multiple smart sensors with edge computing capability}, volume = {22}, year = {2025}, }
• (2025). Transfer learning for anomaly detection in rotating machinery using data-driven key order estimation. TASE, 22, 13310--13326. (www)

  @article{ , abstract = {The detection of anomalous behavior of an engineered system or its components is an important task for enhancing reliability, safety, and efficiency across various engineering applications. However, designing an accurate anomaly detector can be very challenging in settings where anomalous labels are sparse or, in the worst case, missing in the training data. To mitigate this issue of the lack of anomalous labels in the domain of interest, existing approaches use transfer learning by leveraging information from anomalous samples in a closely related source domain. Although previous studies have shown good results from applying transfer learning, they do not specifically address the issue of high false-positive rates from such transfer. High false-positive rates can arise from misleading information present in uninformative/irrelevant features. Inspired by this observation, this paper focuses on identifying key input features, termed as such, due to their strong predictability in anomaly detection. A transfer learning approach is introduced that leverages the optimal $f_{\beta } $ score for key feature estimation. This approach involves finding a weight vector to amplify key features and attenuate uninformative inputs during prediction. We demonstrate the capabilities of our proposed anomaly detection method as a quality check for newly manufactured automotive transmissions. Given the use of frequency domain order-based features in our use case, our proposed method is also easily extensible to the anomaly detection of other rotating machinery. Based on our findings we also find that our proposed anomaly detection algorithm, utilizing precise data-driven features, outperforms detectors based on experience/heuristics-based features currently used in automotive engineering applications. More importantly, our proposed framework can work with any downstream unsupervised anomaly detection algorithm, allowing us to freely choose the best algorithm for the anomaly detection task on hand. Note to Practitioners—Unlike legacy systems that have rich data sets for both healthy and anomalous behaviors over a broad spectrum of operating states, newly designed system or design variants of existing systems have little or no performance data. This makes the task of identifying indicators of anomalous behaviors very challenging for systems without legacy information. This work introduces a machine learning framework that uses label information from related, well-studied systems, to improve anomaly detection outcomes in a new or variant (but related) system. We demonstrate the effectiveness of our transfer approach through the End of Line (EoL) anomaly detection for quality checks on new design variants of an existing automotive transmission. Inspired by feature selection techniques, our proposed framework is easy to use and offers excellent interpretability, which is crucial for achieving a zero-failure system design.}, archive = {J_TASE}, author = {Jia Liang and Huanyi Shui and Rajesh Gupta and Devesh Upadhyay and Eric Darve}, doi = {10.1109/TASE.2025.3552009}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13310--13326}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transfer learning for anomaly detection in rotating machinery using data-driven key order estimation}, volume = {22}, year = {2025}, }
• (2025). Personalized assistance in robotic rehabilitation: Real-time adaptation via energy-based performance monitoring. TASE, 22, 13298--13309. (www)

  @article{ , abstract = {Recent studies underscore the importance of the patient’s active contribution and voluntary effort in enhancing therapy outcomes in physical rehabilitation. This paper presents an adaptive control scheme to implement active robotic rehabilitation. The primary goal is to dynamically regulate robotic assistance based on the patient’s performance and individual conditions, encouraging active participation, and effective therapy. To achieve this, a Lyapunov-based adaptive algorithm is developed that dynamically adjusts the admittance parameters by balancing the error and effort minimization. A novel performance index based on human energy input enables real-time identification of the intended human sharing role. This index is used as an adaptive rate in the proposed algorithm to enhance the control system’s dynamic responsiveness to changes in human performance. The proposed approach achieves two main rehabilitation objectives. First, it encourages active and safe human participation. Second, it enhances the therapy by providing personalized assistance, tailored to individual abilities and conditions, and thus reduces the need for therapist intervention. The performance of the proposed approach is illustrated in experimental studies. The results demonstrate the adaptability of the algorithm, ensuring compliant and safe interaction and effective task completion. Note to Practitioners—In a human-robot cooperation (HRC) framework, the automatic adaptation of the robot’s role as well as safe and stable interaction are crucial. These aspects are amplified in the context of robotic rehabilitation due to the special conditions of the human participants. Classic control methods, in shared control, lack system intelligence and automation in role allocation. However, the shared role of humans in HRC, particularly in rehabilitation applications, introduces real-time and unpredictable variations. This study addresses the shortcomings of classic control methods, by integrating intelligence into the control system through an adaptive Neural Network algorithm in shared autonomy. To emulate human-like adaptability, two crucial aspects are considered. Firstly, it incorporates safety assurance embedded in the adaptive algorithm via Lyapunov-based adaptation. Secondly, it detects the human’s role within the control loop through a novel energy-based performance index, which views the human as an active contributor to the system’s dynamic energy flow. This ensures robust behavior by dynamically adjusting the trade-off between task completion and minimal robot intervention. A standout feature of our algorithm lies in its expendability to exoskeleton systems, making it highly versatile for use in robotic rehabilitation and assistive technologies. The algorithm’s design allows for straightforward integration with exoskeletons, requiring only interaction force measurements in the joint space. It facilitates monitoring of a patient’s performance in each joint using the proposed performance index based on the human energy entry into the system. Beyond rehabilitation, the algorithm’s ability to adjust autonomy levels through adaptation makes it applicable to a wide range of Human-Robot Cooperation scenarios where automatic role allocation is necessary. Preliminary experiments underscore the adaptive algorithm’s robust responsiveness to changes in human performance. Future investigations should involve clinical experiments addressing real-life challenges associated with various movement deficiencies and responding to real-time issues that may arise during rehabilitation sessions.}, archive = {J_TASE}, author = {Leilaalsadat Pezeshki and Hamid Sadeghian and Abolfazl Mohebbi and Mehdi Keshmiri and Sami Haddadin}, doi = {10.1109/TASE.2025.3552446}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13298--13309}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Personalized assistance in robotic rehabilitation: Real-time adaptation via energy-based performance monitoring}, volume = {22}, year = {2025}, }
• (2025). DED-IM: A novel method for mapping and path planning in wire arc directed energy deposition. TASE, 22, 13286--13297. (www)

  @article{ , abstract = {This paper introduces DED-IM (https://github.com/machipanski/DED-IM), a novel image-based mapping and tool-path planning method for Wire Arc Directed Energy Deposition (WA-DED). The method automatically segments 3D models into distinct regions such as thin walls, contours, bottlenecks, and large internal areas by processing binary images. These segments enable precise, geometry-tailored tool-path generation, significantly improving control over material deposition and reducing arc interruptions. Key innovations include the application of medial axis transforms to accurately map thin walls, the introduction of oscillatory paths for bottleneck areas, and an adaptive weaving pattern that minimizes voids while enhancing geometric precision. The method also integrates user-defined parameters, offering flexibility in region mapping and filling strategies to address specific manufacturing requirements. Experimental results demonstrate the efficacy of DED-IM, showing substantial reductions in common defects such as voids, insufficient material filling, and arc interruptions. This project, provides a flexible and scalable foundation baseline for automated WA-DED processes, enabling the integration of diverse path-planning techniques, fostering innovation and customization in tool-path generation for different manufacturing requirements. Note to Practitioners—This paper introduces DED-IM, a novel path-planning method for Wire Arc Directed Energy Deposition (WA-DED) processes. Developed in Python, the method takes a 3D model as input, along with machine-specific parameters, and generates G-code instructions that can be customized for different WA-DED machines. DED-IM automatically segments 3D models into regions like thin walls, contours, bottlenecks, and large internal areas by processing binary images. These parameters are fully adjustable, allowing users to tailor the tool-path strategy to their specific needs. The modular design also makes it easy to integrate new strategies, such as custom zigzag patterns. To use DED-IM in a production workflow, practitioners will need to adjust the software according to their machine specifications.}, archive = {J_TASE}, author = {Matheus Antunes Chipanski and Tadeu Castro da Silva and Neri Volpato and Ricardo Dutra da Silva and Valdemar Rebelo R. Duarte and Telmo G. Santos and Rodrigo Minetto}, doi = {10.1109/TASE.2025.3553309}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13286--13297}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DED-IM: A novel method for mapping and path planning in wire arc directed energy deposition}, volume = {22}, year = {2025}, }
• (2025). A branch-and-price algorithm for the urban aerial delivery problem with energy constraints. TASE, 22, 13269--13285. (www)

  @article{ , abstract = {In this paper, an urban aerial delivery problem (UADP) is investigated, where the parcel transportation service is accomplished by drones in an urban setting. The aim of the problem is to minimize the total service completion time, by taking into account of the flow balance, the energy consumption, and the response time window. To fully explore the structure of the UADP, a mixed integer linear programming (MILP) model is constructed based on an arc-flow scheme. However, directly handling the UADP with commercial solvers is time consuming. In order to enhance the responsiveness of urban courier services and speed up the solving process, a set-covering model (UADP-SC) is proposed with a linear programming based relaxation. Then a branch-and-price algorithm is designed with pricing accelerating strategies based on heuristics. The computational experiments show that the proposed branch-and-price algorithm outperforms the off-the-shelf commercial solvers in terms of computation efficiency. In the mean time, the proposed algorithm can also serve to obtain optimal battery swapping and path planing decisions in face of the large-scale urban aerial delivery problem with energy constraints. Note to Practitioners—With the intensification of the aging population issue, the manual labor costs in logistics have sharply increased. Simultaneously, advancements in drone technology enable uncrewed aerial vehicles to participate in logistics distribution systems, addressing the last-mile delivery challenge. In the planning of drone delivery routes, the constraint of drone batteries cannot be ignored. This constraint not only affects delivery safety but also impacts delivery efficiency—both crucial considerations for decision-makers. Consequently, we propose a model considering the energy constraints. We introduce a branch-and-pricing algorithm to expedite the problem solving. The results show that the proposed algorithm performs well across various problem scales. Moreover, adopting a strategy of replacing batteries only when necessary can save approximately 17% to 23% of the total completion time. We also conducted performance comparisons under different ratios of orders to drones, providing decision-makers with a useful benchmark.}, archive = {J_TASE}, author = {Zhi Pei and Zhaohui Liang and Jiayan Huang and Tao Fang and Na Li}, doi = {10.1109/TASE.2025.3553200}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13269--13285}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A branch-and-price algorithm for the urban aerial delivery problem with energy constraints}, volume = {22}, year = {2025}, }
• (2025). Roll torque vibration isolation control on parallel active suspension system via impedance-based MPC for vehicular terminal. TASE, 22, 13253--13268. (www)

  @article{ , abstract = {To improve the stability of vehicle body dynamics on rugged roads, the majority of researchers have studied vertical force control at the terminal of the vehicle utilizing an active suspension system. However, terminal disturbance roll torque can also have a negative impact on the stability of the vehicle’s state. In this paper, a terminal roll torque control strategy (TRTCS) is proposed to enhance the stability of body attitude by controlling terminal roll torque in wheeled motion. The strategy includes impedance-based model predictive control (IMPC) and terminal posture return controller (TPRC). Firstly, IMPC analyzes the single-leg dynamic model embedded with torque impedance at the terminal. Then, the controller iterates through quadratic optimization to obtain an optimal impedance target torque that can efficiently mitigate terminal disturbance roll torque while stabilizing the terminal roll velocity, acceleration, and force on each electric cylinder. Then, TPRC generates additional smooth return speed for the terminal roll angle to the initial value through the Bezier curve and event-trigger mechanisms to estimate the terminal contact state to the ground. TPRC aims to compensate for low angular return speed due to insufficient measured torque on the suspending terminal during the downhill process. Finally, through simulation and experiment, TRTCS significantly diminishes the deviation of terminal roll torque and body attitude from target value to 25.2% and 35.8% respectively when a vehicle crosses multi-slopes compared with traditional torque impedance.Note to Practitioners—The motivation for this work is that the vibration isolation control on vehicles or wheeled robots often neglects the disturbance of the roll torque on their feet on rough roads. Unfortunately, this can damage the leg actuators of vehicles or robots, and even have a negative impact on the steadiness of the body posture. Therefore, this paper proposes a TRTCS, which controls the terminal roll torque by rotating its roll angle with high dynamic stability. The control framework has the following functionalities: 1). The terminal roll angular velocity is obtained based on an optimization function that can make the terminal torque converge smoothly to the target value and the leg actuator subjected to more distributed force. 2). A smooth roll angle correction additional speed is generated based on the Bezier curve to compensate for the slow roll angle correction speed caused by insufficient correction roll torque during the downhill phase. A series of simulations and experiments have verified that the control method has a small steady-state error in controlling the roll torque at the terminal. At the same time, under TRTCS, the peak force on the leg actuator is significantly reduced, which has a significant inhibitory effect on the actuator loss and unsteadiness of the body posture. In future research, we will further investigate the control of terminal pitch and yaw torque.}, archive = {J_TASE}, author = {Junfeng Xue and Zhihua Chen and Shoukun Wang and Junzheng Wang and Yongkang Xu}, doi = {10.1109/TASE.2025.3539596}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13253--13268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Roll torque vibration isolation control on parallel active suspension system via impedance-based MPC for vehicular terminal}, volume = {22}, year = {2025}, }
• (2025). A data-driven and physically constrained approach to parameter optimization for milling complex surface. TASE, 22, 13236--13252. (www)

  @article{ , abstract = {Optimizing the milling parameters for complex surface can improve the machining quality. However, existing methods are oriented toward a simple machining path and single machining process. In this paper, a milling parameter optimization method for complex surface is proposed by combining the data-driven models and physical constraints. Initially, the chatter indicator is derived from the tool vibration using Variational Mode Decomposition. Subsequently, multiple few-shot prediction models are developed based on real machining data using deep neural networks. A mathematical optimization model is then constructed by combining multiple prediction models and physical constraints, which aims to minimize the machining failure rate and maximize the material removal rate under multiple constraints. The spindle speed, feed speed, cutting depth, and path spacing are the optimization parameters. Finally, a Hypervolume-based Multi-Objective Optimization (HMOO) algorithm is proposed to solve the optimization model. The solution set produced by HMOO exhibits superior convergence and diversity compared to the S-Metric Selection Based Evolutionary Multi-Objective Algorithm (SMS-MOEA). In experiments, a five-axis machine tool is employed to mill turbine blades with complex surface. Experimental results demonstrate that the proposed prediction models achieve higher accuracy than widely used regression algorithms. Integrating the high-precision prediction models with HMOO significantly enhances blade machining quality while guaranteeing reliable machining efficiency, resulting in a 5.25% reduction in the machining failure rate. Note to Practitioners—Methods designed for simple machining paths and single machining processes are inadequate for optimizing milling parameters in complex surfaces. Given the challenging machining characteristics and stringent precision demands of complex surfaces, this paper proposes a milling parameter optimization approach for complex surface by integrating data-driven models with real physical constraints. The approach employs neural network-based prediction models to map the relationships among milling parameters, machining conditions, and machining accuracy, while accounting for actual physical constraints in the optimization process. Our approach enhances the machining accuracy of complex surfaces while ensuring reliable machining efficiency.}, archive = {J_TASE}, author = {Feng Gao and Dechang Pi and Junfu Chen}, doi = {10.1109/TASE.2025.3553115}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13236--13252}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A data-driven and physically constrained approach to parameter optimization for milling complex surface}, volume = {22}, year = {2025}, }
• (2025). Finite-time prescribed performance adaptive control for a class of switched uncertain nonlinear systems. TASE, 22, 13224--13235. (www)

  @article{ , abstract = {This paper investigates a finite-time prescribed performance adaptive control (FTPPAC) scheme for a class of full-state dependent switched uncertain nonlinear systems in the non-strict feedback form. Firstly, a model transformation is introduced to handle unknown control coefficients and unknown nonlinear functions, after which a state observer is designed to estimate the unknown state of the transformed system. Subsequently, the backstepping method and adaptive control are employed to ensure that the tracking error satisfies the pre-designed performance indices. Compared with the traditional design of prescribed performance control (PPC) for switched systems, a novel switching-dependent prescribed performance bound (SDPPB) is presented, which is specifically tailored to satisfy the requirements and characteristics of different subsystems. This design approach allows for the adjustment of the initial value of SDPPB at the switching instants based on the tracking error and subsystem change. Finally, by applying the average dwell time switching and multiple Lyapunov functions method, sufficient conditions for finite-time stability of the closed-loop system are derived. The effectiveness of the proposed control scheme is demonstrated through two simulations. Note to Practitioners—Finite-time stability analysis and quantitative evaluation of control performance are hot topics in the field of control, which play a crucial role in engineering practice. This paper proposes the finite-time prescribed performance adaptive control scheme for a class of switched uncertain nonlinear systems, which can be applied to motion tracking control of servo mechanisms, command tracking of aero-engine and attitude control of spacecraft. It is worth noting that unknown nonlinear dynamics, unknown control coefficients and system state performance constraint problems often exist in practical systems. Therefore, the model transformation is presented to address unknown control coefficients and unknown nonlinear functions. The state observer is used to estimate the unknown state of the transformed system. By designing switching-dependent prescribed performance bound, whose initial value can vary with the subsystem and tracking error, and by utilizing adaptive control method that remains continuous at the switching instant, we ensure that the tracking error satisfies the pre-designed performance indices and that all signals of the closed-loop system are bounded in finite time. In view of the growing demand for transient and steady-state performance in modern industrial production, this control scheme holds promise for practical problem-solving.}, archive = {J_TASE}, author = {Wanchen Wang and Jie Lian and Peilin Jia and Kang Chen}, doi = {10.1109/TASE.2025.3551145}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13224--13235}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time prescribed performance adaptive control for a class of switched uncertain nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). An accelerated long-term congestion assessment method for power systems with high-proportional renewable energy and energy storage systems. TASE, 22, 13212--13223. (www)

  @article{ , abstract = {Congestion analysis is critical for power system analysis and operation, but the integration of renewable energy sources (RESs) and energy storage systems (ESSs) will bring great uncertainty and computational challenges. In this study, we propose an accelerated simulation method for long-term congestion assessment in power systems characterized by a high proportion of RESs and ESSs. Our approach includes an affine adjustable robust congestion assessment model capable of identifying congestion in each sampling. The model utilizes automatic generation control to address uncertainties associated with RESs and incorporates multi-period coupling constraints for components with chronological characteristics. Additionally, the Cross-Entropy-Latin Hypercube Sampling (CE-LHS) algorithm is employed to expedite convergence during sampling generation in sequential Monte Carlo Simulation (SMCS). Numerical results from several test systems demonstrate the effectiveness and computational enhancements achieved by the proposed technique. Note to Practitioners—This paper presents an accelerated method for long-term congestion assessment in power systems with high-proportional RESs and ESSs. A robust optimization model with multi-period coupling constraints is proposed to establish operational boundaries for long-term congestion analyzes. Building on SMCS, the CE-LHS algorithm is integrated to improve convergence performance. CE produces the optimal parameter distribution of the system, while LHS effectively mitigates the truncation effect associated with CE. The application future of this method is not limited to power systems, it can also be utilized in other industrial systems characterized by significant uncertainty, such as energy systems, structural facilities, and automated control systems.}, archive = {J_TASE}, author = {Jiawei Sun and Tao Ding and Guangming Lu and Xiaojie Pan}, doi = {10.1109/TASE.2025.3548325}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13212--13223}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An accelerated long-term congestion assessment method for power systems with high-proportional renewable energy and energy storage systems}, volume = {22}, year = {2025}, }
• (2025). Robust soft constrained model predictive control and its application in wastewater treatment processes. TASE, 22, 13198--13211. (www)

  @article{ , abstract = {A robust soft constrained model predictive control (RSCMPC) method is proposed to address the effects of unknown disturbances for wastewater treatment processes (WWTPs). The disturbances involving inflow fluctuation and noises from WWTPs may result in the constraints violation of MPC due to its uncertainty of bioprocess, which may degrade the performance of the steady state. First, the artificial steady state is introduced to mimic the nearest feasible steady state when the reference steady state is not feasible. The deviation caused by disturbances between the artificial steady state and the reference steady state is also penalized to ensure that the output of MPC converges to the reference steady state. Second, the soft constraints, incorporating two slack variables and a penalty term, are designed to relax the state constraints of MPC and continuously mitigate the constraint violation, thereby ensuring its stability. Third, the input state stability (ISS) under disturbances is analyzed. Finally, the simulation tested on Benchmark simulation model 1 verifies the effectiveness of the proposed RSCMPC. The results demonstrate that RSCMPC improves the robustness of the system to maintain the stable operation of the WWTPs. Note to Practitioners—The external disturbances of wastewater treatment processes (WWTPs) will result in the constraints violation of model predictive control (MPC) and degrade the steady state performance. To overcome the influence of external disturbances, a robust soft constrained model predictive control (RSCMPC) method is designed. This method mainly includes three contributions: First, the artificial steady state is constructed with the prediction derived from fuzzy neural network, which is to mimic the nearest feasible steady state when the reference steady state is not feasible under disturbances. Second, the soft constrained method relaxes the state constraints of MPC through two slack variables to compensate the effects of disturbances and restore the feasibility of the controller. Third, the input state stability under disturbances is analyzed. Finally, the effectiveness of the proposed RSCMPC method is evaluated on a pilot platform of a real WWTPs. The experimental results show that RSCMPC can improves the control accuracy and robustness of the system under external disturbances. The proposed RSCMPC method can help practitioners improve the reliability of WWTPs operation.}, archive = {J_TASE}, author = {Xiao-Long Wu and Wen-Hai Han and Hong-Yan Yang and Xin Li and Hong-Gui Han}, doi = {10.1109/TASE.2025.3552421}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13198--13211}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust soft constrained model predictive control and its application in wastewater treatment processes}, volume = {22}, year = {2025}, }
• (2025). A multi-task learning-based approach for power system short-term voltage stability assessment with missing PMU data. TASE, 22, 13187--13197. (www)

  @article{ , abstract = {This paper proposes a novel multi-task learning approach based on spatial-temporal recurrent imputation network (SRIN) for power system short-term voltage stability (STVS) assessment with incomplete PMU measurements. The state-of-the-art data imputation methods are based on single and separated learning tasks, which lack optimality for fully exploiting the information in available data. They are also facing several challenges in practical applications, e.g., dependence on complete datasets for training, and performance degradation under continuous data missing scenarios. As a significant advantage, the proposed SRIN method jointly optimizes the objective of missing value imputation and stability prediction through a multi-task recurrent network model. In this way, the integrated model can fully learn from any available data in the incomplete historical database, and the performance of both tasks can benefit from knowledge sharing and transferring across tasks. Moreover, the proposed method has superior advantages in handling both spatial and temporal consecutive missing scenarios, where the imputations are derived by an intelligent combination of history-based and feature-based estimations. Numerical simulation results on two test systems show that, under any PMU missing condition, the proposed method can maintain a competitively high STVS assessment accuracy with a much less imputation error. Note to Practitioners—This paper addresses the challenge of incomplete system observations for power system real-time stability assessment. This problem is not unique to power systems but also extends to other sequential prediction problems facing severe data incompleteness. Existing approaches to solve the missing data problem either relay on complete historical data to train an imputation model, which may not always hold true during practical applications, or impute the missing data by simple statistics, which lacks optimality and adaptivity under diverse missing patterns. This paper proposed a novel, integrated approach to solve this problem by jointly optimizing the two tasks together through a new recurrent network model. In this way, the method can fully learn from seriously undermined datasets. Moreover, this method deals with consecutive missing in time and space, by the design of a trainable weighting component. Numerical simulation results on standard power systems shows that the proposed multi-task model improve the performance of both two tasks and have high adaptivity to different data missing scenarios. In the future research, we will try to address the learning efficiency of this approach for application to larger systems and exploring its adaptability in more extreme scenarios.}, archive = {J_TASE}, author = {Qiaoqiao Li and Chao Ren and Rui Zhang and Yan Xu}, doi = {10.1109/TASE.2025.3551593}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13187--13197}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A multi-task learning-based approach for power system short-term voltage stability assessment with missing PMU data}, volume = {22}, year = {2025}, }
• (2025). Conflict-averse multi-objective extremum seeking and its application in wastewater treatment processes. TASE, 22, 13175--13186. (www)

  @article{ , abstract = {This paper addresses multi-objective optimization problems using conflict-averse multi-objective extremum seeking (CAMOES) for unknown static mapping. As for the traditional multi-objective extremum seeking (MOES) methods, the initial conditions affect the optimal solution, which may result in over-optimizing some objectives. To overcome this issue, by minimizing the average loss function with the 2-norm of the worst individual objective, conflict-averse gradient estimator is investigated to determine the weighting factors and regularization parameter adaptively, extenuating the adverse impact of the prior parameters and initial conditions on the performance of extremum seeking. To cope with the integral windup problem for constrained inputs, penalty-function anti-windup mechanism is explored with smoothing saturation function for multiple-input multiple-output (MIMO) extremum seeking. The stability of CAMOES is theoretically analyzed. Some bi-objective optimization problems are conducted, including a numerical example and a practical problem for wastewater treatment processes (WWTPs). The results demonstrate that the proposed CAMOES exhibits competitive performance. Note to Practitioners—The initial conditions may lead to different Pareto solutions, which brings adverse impact on the multi-objective optimization performance in WWTPs. How to mitigate the influence of the initial parameters on the optimization performance and deal with the changing operation condition is the key to balancing the trade-off between the effluent quality (EQ) and energy cost (EC). This paper presents a control optimization scheme with intelligent modeling, extremum seeking optimization and the control part. The ErrCor-RBF neural network is built to predict the EC and EQ in WWTPs. For constraint conditions, the penalty-function anti-windup mechanism is exploited to avoid integral windup problems. The optimal set-points are dynamically adjusted by the proposed CAMOES to cope with the varying influent flow rate. The experimental results prove the proposed scheme has better operation and control performance. This paper proposes conflict-averse multi-objective extremum seeking (CAMOES), which is applied to the bi-objective optimization of WWTPs but can also be applied to other systems with multi-objective optimization tasks. In future research, we will address the time delay problem in WWTPs.}, archive = {J_TASE}, author = {Minghui Chu and Xin Huo and Kemao Ma}, doi = {10.1109/TASE.2025.3552046}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13175--13186}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Conflict-averse multi-objective extremum seeking and its application in wastewater treatment processes}, volume = {22}, year = {2025}, }
• (2025). Surrogate-assisted scenario-generation method for simulation-based stochastic programming problems. TASE, 22, 13161--13174. (www)

  @article{ , abstract = {Simulation is required in real-world industrial decision-making to model the complexity, such as supply chain management with uncertain future demand. Scenario generation is critical for handling inherent uncertainties as deterministic scenarios to solve stochastic programming problems (SPPs). Generating a minimal yet representative scenario set is important to solve simulation-based SPPs (SBSPPs) since SBSPPs require thousands of simulation iterations proportional to the number of scenarios, leading to significant computational time. However, conventional methods, such as Monte Carlo method and recent problem-driven approaches, are ineffective in solving SBSPPs due to time-consuming simulation evaluation. This paper proposes a surrogate-assisted scenario-generation method called Inferred Cost-Space Scenario Clustering (ICSSC) that is applicable to various SPPs including SBSPPs. ICSSC employs a scenario clustering based on a new cost-space scenario distance evaluated by the surrogate model trained with offline simulation data to quickly approximate simulation evaluations. We conducted three types of numerical experiments to validate the effectiveness: Markowitz portfolio optimization, stochastic server location, and inventory placement optimization. Empirical results revealed that ICSSC could generate an effective scenario set based on the impact of uncertainties on decision outcomes for broader SPPs, and yields better solutions with 7.2 times shorter runtime than Monte Carlo methods. Note to Practitioners—This paper introduces a scenario-generation method to reduce the computational time of simulation-based optimization under uncertainty. Such optimization problems are common in practical decision-making contexts like production planning and supply chain management, which often require highly optimized and resilient solutions. Conventional methods, such as Monte Carlo method, demand extensive computational resources due to evaluating numerous randomly sampled scenarios. This study addresses scenarios in which similar problems are solved iteratively, a common occurrence in practical applications. Our method uses a surrogate model to approximate simulation evaluations quickly and selects a set of representative scenarios on the basis of the impact of uncertainties on decision outcomes. Numerical experiments demonstrated that our method could generate representative scenarios that obtain more robust and optimal solutions with shorter calculation times than conventional methods.}, archive = {J_TASE}, author = {Issei Suemitsu and Kazuhiro Izui}, doi = {10.1109/TASE.2025.3551716}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13161--13174}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Surrogate-assisted scenario-generation method for simulation-based stochastic programming problems}, volume = {22}, year = {2025}, }
• (2025). Fast charging of lithium-ion batteries using deep bayesian optimization with recurrent neural network. TASE, 22, 13150--13160. (www)

  @article{ , abstract = {Fast charging has attracted increasing attention from the battery community for electrical vehicles (EVs) to alleviate range anxiety and reduce charging time for EVs. However, inappropriate charging strategies would cause severe degradation of batteries or even hazardous accidents. To optimize fast-charging strategies under various constraints, particularly safety limits, we propose a novel deep Bayesian optimization (BO) approach that utilizes Bayesian recurrent neural network (BRNN) as the surrogate model, given its capability in handling sequential data and providing uncertainty quantifications for the output. In addition, a combined acquisition function of expected improvement (EI) and upper confidence bound (UCB) is developed to better balance the exploitation and exploration. The effectiveness of the proposed approach is demonstrated on the PETLION, a porous electrode theory-based battery simulator. Our method is also compared with the state-of-the-art BO methods that use Gaussian process (GP) and non-recurrent network as surrogate models. The results verify the superior performance of the proposed fast charging approaches, which mainly results from that: 1) the BRNN-based surrogate model provides a more precise prediction of battery lifetime than that based on GP or non-recurrent network; and 2) the combined acquisition function outperforms traditional EI or UCB criteria in exploring the optimal charging protocol that maintains the longest battery lifetime. Note to Practitioners—This study is motivated by the need to develop fast-charging strategies for batteries to reduce the charging time while maintaining the safety and longevity of batteries. Traditional methods to optimize battery fast-charging protocols require either solving complex battery models or conducting tremendous repetitive and costly cycling experiments. In addition, the majority of works in this direction mainly focus on minimizing the charging time without considering the battery degradation caused by the designed charging protocol. To address these issues, we present a data-driven optimization technique to rapidly discover the optimal fast-charging strategies that not only decrease the charging time but also slow down the battery degradation. Our method is efficient and can find the optimal solution within tens of iterations (or repetitive experiments), thus greatly reducing the cost and efforts in optimizing battery charging protocols.}, archive = {J_TASE}, author = {Benben Jiang and Yixing Wang and Zhenghua Ma and Qiugang Lu}, doi = {10.1109/TASE.2025.3551879}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13150--13160}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast charging of lithium-ion batteries using deep bayesian optimization with recurrent neural network}, volume = {22}, year = {2025}, }
• (2025). Spatial geometry theory-based algorithm for accurate positioning with single vision beacon. TASE, 22, 13139--13149. (www)

  @article{ , abstract = {Quick response (QR) code is one of the most popular beacons for indoor positioning. Existing QR code-based algorithms typically require several QR codes for precise positioning, which may need intense QR code deployment or have limited coverage. This paper proposes a visual positioning algorithm based on feature points and triangular structures on a QR code (V-FTQ), which can achieve precise position and orientation estimation using only a monocular camera and a single QR code beacon. In the considered system, a few QR codes are placed on the ceiling, and a user holds a camera in hand to capture the QR code for positioning. In particular, we first propose an efficient matching mechanism to match the feature points of the captured QR beacon to their projections in the image plane of the camera. Then, based on the geometry relations between the feature points and their projections, V-FTQ can calculate the coordinates of the feature points in the camera coordinate. Finally, the orientation and position of the user are estimated based on linear algebra and spatial geometry. Moreover, since V-FTQ may have a few ambiguous solutions, a singular solution elimination scheme is further proposed based on plane geometry and linear programming, which can effectively eliminate the incorrect solutions. Both simulation and experimental results demonstrate that V-FTQ can achieve a positioning accuracy within 10 cm with only one captured beacon, outperforming baseline algorithms in terms of precision and robustness to noise. Note to Practitioners—Existing positioning algorithms typically require multiple beacons for accurate positioning, which may limit its coverage or require dense beacon deployment. The paper proposes a novel positioning algorithm that achieves 3D orientation and position estimation with only a single beacon and a single camera. We applied the proposed method to the mobile phone and developed an APP for actual positioning tests. Experimental results verify that the positioning scheme reduces both the number of beacons required and the cost of the positioning system, making it suitable for the automatic navigation of robots and other applications in industrial automation.}, archive = {J_TASE}, author = {Boyu Yang and Yang Yang and Caili Guo and Zhiyu Zhu}, doi = {10.1109/TASE.2025.3552107}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13139--13149}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Spatial geometry theory-based algorithm for accurate positioning with single vision beacon}, volume = {22}, year = {2025}, }
• (2025). Adaptive nonlinear recursive model updating method for quality-related process monitoring. TASE, 22, 13128--13138. (www)

  @article{ , abstract = {In data-driven nonlinear modeling, kernel partial least squares (KPLS) is a useful tool for quality-related process monitoring. However, in time-varying nonlinear processes, degraded equipment and parameter drift require nonlinear models to be updated regularly. If a model is not updated, it is difficult for the nonlinear model to track the changes in the system, and the process monitoring system has a high false alarm rate. To our knowledge, there is a lack of research on the adaptive updating of quality-related nonlinear models. To address this problem, an adaptive nonlinear recursive model updating method is proposed. In the proposed method, a nonlinear orthogonal regression algorithm, named NOKPLS, which is suitable for model updating, is first developed to construct a nonlinear model. Specifically, by constructing an orthogonal projection operator, the input space can be divided into mutually orthogonal quality-related subspaces and quality-unrelated subspaces, and false alarms for quality variables caused by quality-unrelated faults can be greatly reduced. Furthermore, on the basis of the NOKPLS model, an adaptive nonlinear model updating method is developed with low computational complexity. Finally, a quality-related process monitoring strategy for recursive nonlinear model updating is designed. A numerical simulation and the Tennessee Eastman process (TEP) are applied to demonstrate the effectiveness of the proposed method. Note to Practitioners—Monitoring and controlling the key performance indicators is crucial for ensuring the healthy operation of nonlinear processes. In practice, however, the aging of the components and the drift of the parameters often lead to slow time-varying system processes. Therefore, the nonlinear model needs to be updated regularly. Here, a recursive nonlinear PLS model adaptive updating method is proposed. The nonlinear orthogonal regression algorithm is first developed to reduce the number of false alarms caused by quality-unrelated faults and to prepare for model updating. Considering the high computational complexity of kernel-based nonlinear methods in modeling, a new model update matrix is constructed by using model parameters instead of historical data, which effectively reduces the computational complexity and improves the efficiency of model updates. Then, a complete quality-related nonlinear monitoring and adaptively updating strategy is designed. In the proposed approach, the number of false alarms induced by time-varying processes can be reduced, and unnecessary overhaul and maintenance can be avoided. In addition, the superiority and characteristics are discussed and studied experimentally.}, archive = {J_TASE}, author = {Jiayu Luo and Xiangyu Kong and Chaoli Zhang}, doi = {10.1109/TASE.2025.3550725}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13128--13138}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive nonlinear recursive model updating method for quality-related process monitoring}, volume = {22}, year = {2025}, }
• (2025). Event-triggered optimal consensus control for MASs with multiple constraints: A flexible performance approach. TASE, 22, 13117--13127. (www)

  @article{ , abstract = {This paper investigates the challenge of achieving event-triggered optimal consensus control for multiagent systems (MASs) with multiple constraints, encompassing saturation constraint at the input and performance constraint at the output. To achieve performance constraint while satisfying input saturation, a flexible prescribed performance method (FPPM) is designed. Utilizing non-negative signals generated by the improved auxiliary system to design the performance functions, the FPPM can adaptively adjust the performance constraint boundaries to ensure safe operation of the MASs with multiple constraints. Meanwhile, the proposed FPPM can achieve different performance behaviors by changing core parameters without the need to alter the control structure. Subsequently, a simplified reinforcement learning algorithm with actor-critic structure is integrated into the FPPM. By designing actor-critic neural networks and dynamic event-triggered mechanism, optimal consensus control for MASs under multiple constraint conditions is achieved cleverly while avoiding unnecessary communication transmissions. Finally, a simulation example verifies the effectiveness of the proposed method. Note to Practitioners—Considering the limitations of physical devices and the practical requirements for control performance, the input saturation constraint and performance constraint often coexist during the operation of practical systems, such as robotic systems, manipulator systems and aerospace systems. Therefore, this paper aims to design an event-triggered reinforcement learning algorithm for MASs with multiple constraints. To resolve the conflict problem caused by input saturation and performance constraint, a FPPM with adjustable performance functions is proposed. By flexibly adjusting the performance constraint boundaries, the coexistence problem of multiple constraints can be solved effectively. Meanwhile, the constructed FPPM framework can achieve various performance behaviors by adjusting parameters according to the practical application scenario without changing the controller structure. Additionally, the proposed event-triggered reinforcement learning algorithm can optimize the designed cost function and promote the utilization of communication resources.}, archive = {J_TASE}, author = {Ao Luo and Qi Zhou and Hui Ma and Hongyi Li}, doi = {10.1109/TASE.2025.3549435}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13117--13127}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered optimal consensus control for MASs with multiple constraints: A flexible performance approach}, volume = {22}, year = {2025}, }
• (2025). Sustainable operation of CCUS units under low-carbon economics. TASE, 22, 13101--13116. (www)

  @article{ , abstract = {To ensure the sustainable operation of the Carbon Capture, Utilization, and Storage (CCUS) units, a novel cooperative scheme is proposed for CCUS units and wind farm clusters (WFCs) within integrated markets of electricity, carbon emission trading (CET), and green certificate trading (GCT), incorporating tradable green certificate (TGC) offset mechanism. This scheme, utilizing the Asymmetric Nash Bargaining (ANB) theory, is divided into two sub-problems: energy trading and benefit allocation. In the energy trading problem, uncertainties in electricity prices and wind power are addressed using a data-driven distributionally robust (D-DRO) method to maximize the expected utility, and a model-oriented Benders Decomposition (BD) algorithm is proposed to ensure privacy and computational efficiency. For the benefit allocation problem, an analytical method based on Karush-Kuhn-Tucker (KKT) conditions is employed to achieve a rational and fair allocation, considering each participant’s bargaining power. Numerical experiments in small-scale and large-scale systems indicate that CCUS unit revenue has increased by 14.09% and 7.28%, respectively, with the proposed scheme. Additionally, when solving the energy trading problem considering the refined AA model for electrolyzer clusters, the computational time has accelerated by 28.56 times and 66.55 times, while ensuring solution quality. Note to Practitioners—The incorporation of CCUS units has emerged as a promising option for the energy sector to achieve low-carbon transformation. Unlocking the profitability of CCUS units amid the fluctuating conditions of power systems and within complex market environments is critical for ensuring their sustainable operation. This paper proposes a cooperative scheme between CCUS units and WFCs within multi-market settings. By sequentially addressing the energy trading problem and the benefit allocation problem, this scheme enables participating entities to mitigate potential uncertainties in transactions, effectively execute trading plans while preserving privacy, and achieve fair and rational benefits. This scheme not only provides valuable insights for designing sustainable operational strategies for CCUS units under low-carbon economics within power systems, but its applied cooperative framework and distributed algorithm also exhibit significant scalability, offering industry practitioners important practical value.}, archive = {J_TASE}, author = {Zhenzi Song and Xiuli Wang and Tianyang Zhao and Tao Qian and Libo Zhang and Buyang Qi and Yifei Wang}, doi = {10.1109/TASE.2025.3551163}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13101--13116}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sustainable operation of CCUS units under low-carbon economics}, volume = {22}, year = {2025}, }
• (2025). NAMR-RRT: Neural adaptive motion planning for mobile robots in dynamic environments. TASE, 22, 13087--13100. (www)

  @article{ , abstract = {Robots are increasingly deployed in dynamic and crowded environments, such as urban areas and shopping malls, where efficient and robust navigation is crucial. Traditional risk-based motion planning algorithms face challenges in such scenarios due to the lack of a well-defined search region, leading to inefficient exploration in irrelevant areas. While bi-directional and multi-directional search strategies can improve efficiency, they still result in significant unnecessary exploration. This article introduces the Neural Adaptive Multi-directional Risk-based Rapidly-exploring Random Tree (NAMR-RRT) to address these limitations. NAMR-RRT integrates neural network-generated heuristic regions to dynamically guide the exploration process, continuously refining the heuristic region and sampling rates during the planning process. This adaptive feature significantly enhances performance compared to neural-based methods with fixed heuristic regions and sampling rates. NAMR-RRT improves planning efficiency, reduces trajectory length, and ensures higher success by focusing the search on promising areas and continuously adjusting to environments. The experiment results from both simulations and real-world applications demonstrate the robustness and effectiveness of our proposed method in navigating dynamic environments. A website about this work is available at https://sites.google.com/view/namr-rrt. Note to Practitioners— The growing demand for autonomous robots to navigate efficiently and robustly in dynamic, crowded environments like public areas has motivated this work. Traditional risk-based motion planning algorithms often suffer from unfocused search processes, leading to inefficient exploration and performance bottlenecks. This article introduces the NAMR-RRT algorithm to address these issues by integrating neural network-generated heuristic regions to guide the search process. NAMR-RRT adaptively updates both the heuristic region and sampling rate during the planning process, allowing it to focus on more promising areas and dynamically adjust to environmental changes. Unlike conventional methods relying on random exploration, NAMR-RRT improves efficiency by focusing searches in regions more likely to lead to the feasible path, thereby reducing trajectory length and enhancing overall performance. This approach is valuable for mobile robots operating in human-robot coexisting environments, where dynamic adaptability and efficient navigation are critical. The experiment results demonstrate that NAMR-RRT provides a reliable and efficient solution for motion planning in such complex scenarios.}, archive = {J_TASE}, author = {Zhirui Sun and Bingyi Xia and Peijia Xie and Xiaoxiao Li and Jiankun Wang}, doi = {10.1109/TASE.2025.3551464}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13087--13100}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {NAMR-RRT: Neural adaptive motion planning for mobile robots in dynamic environments}, volume = {22}, year = {2025}, }
• (2025). Integrating with multimodal information for enhancing robotic grasping with vision-language models. TASE, 22, 13073--13086. (www)

  @article{ , abstract = {As robots grow increasingly intelligent and utilize data from various sensors, relying solely on unimodal data sources is becoming inadequate for their operational needs. Consequently, integrating multimodal data has emerged as a critical area of focus. However, the effective combination of different data modalities poses a considerable challenge, especially in complex and dynamic settings where accurate object recognition and manipulation are essential. In this paper, we introduce a novel framework integrating with Multimodal Information for Grasping Synthesis with vision-language models (MIG) designed to improve robotic grasping capabilities. This framework incorporates visual data, textual information, and human-derived prior knowledge. We start by creating target object masks based on this prior knowledge, which are then used to segregate the target objects from their surroundings in the image. Subsequently, we employ language cues to refine the visual representations of these objects. Finally, our system executes precise grasping actions using visual and textual data synthesis, thus facilitating more effective and contextually aware robotic grasping. We carry out experiments using the OCID-VLG dataset. We observe that our methodology surpasses current state-of-the-art (SOTA) techniques, delivering improvements of 9.91% and 5.70% for top-1 and top-5 predictions in grasp accuracy. Moreover, when apply to the reconstructed Grasp-MultiObject dataset, our approach demonstrates even more substantial enhancements, achieving gains of 17.63% and 22.76% over SOTA methods for top-1 and top-5 predictions, respectively. Note to Practitioners—As robotic systems evolve, the challenge of enabling them to function effectively in complex environments has become increasingly apparent. This paper introduces a solution that integrates multiple sources of data—visual, textual, and human knowledge—to enhance robotic grasping capabilities. The practical problems addressed include the limitations of current unimodal systems that struggle with accurate object recognition and manipulation in dynamic settings, such as warehouses or assembly lines. Our framework, MIG, demonstrates significant improvements in grasp accuracy, making it suitable for tasks where precision is critical. While our results show promise, particularly in controlled experiments, there are limitations to consider. The framework’s performance may vary in unstructured real-world environments due to factors like occlusion or varying lighting conditions. Future work should focus on refining the system for real-time application and exploring additional sensory inputs to enhance robustness. By addressing these challenges, we aim to make this approach more applicable across industries, paving the way for smarter, more adaptable robotic solutions in everyday tasks.}, archive = {J_TASE}, author = {Zhou Zhao and Dongyuan Zheng and Yizi Chen and Jing Luo and Yanjun Wang and Panfeng Huang and Chenguang Yang}, doi = {10.1109/TASE.2025.3550360}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13073--13086}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrating with multimodal information for enhancing robotic grasping with vision-language models}, volume = {22}, year = {2025}, }
• (2025). A robotic micromanipulation system for homogeneous organoid culture. TASE, 22, 13061--13072. (www)

  @article{ , abstract = {Organoids are cell clusters cultured in vitro that maintain the structure and function of the donor organs. They have found important applications in biomedicine, such as drug screening and personalized therapy. However, conventional organoid culture methods lack control of physical properties like size and distribution, leading to increased heterogeneity and very low batch-to-batch reproducibility, which significantly limits their widespread use. Controlling these properties at the microscale is challenging, particularly for fragile fragments, which are the main source for culturing organoids. To address this issue, we present a robotic micromanipulation system that allows operators to select fragments of particular sizes and automatically transfer them into a customized in-situ organoid chip (IOC) for culture. The chip was designed with microwell arrays to uniform the culture environment and facilitate imaging analysis. The transfer of fragments is modeled based on computational fluid dynamics (CFD) and is enabled by designing a robust model predictive control (RMPC) framework. Simulation and experiment results demonstrated the effectiveness of the model and controller. In colorectal cancer organoid culture experiments, our system significantly improved the morphological homogeneity of organoids. Note to Practitioners—Organoids have been demonstrated to be one of the most promising in vitro models. Lacking control of its size and distribution results in significant heterogeneity and low batch-to-batch reproducibility, which limits its wide uses. Here, we report a robotic micromanipulation system that allows operators to select fragments of particular sizes and morphologies and automatically transfer them into a customized organoid chip for culture. The results of colorectal cancer organoids culture experiments verified the effectiveness of our system in reducing the morphological heterogeneity among organoids.}, archive = {J_TASE}, author = {Xiaofei Wang and Xiaotian Lin and Xinghu Yu and Qiong Mo and Mingsi Tong and Meng Jiang and Songlin Zhuang and Huijun Gao}, doi = {10.1109/TASE.2025.3551297}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13061--13072}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A robotic micromanipulation system for homogeneous organoid culture}, volume = {22}, year = {2025}, }
• (2025). Proactive contingency-aware task allocation and scheduling in multi-robot multi-human cells via hindsight optimization. TASE, 22, 13046--13060. (www)

  @article{ , abstract = {Multi-robot systems are becoming more common in various real-world applications, such as manufacturing and warehouse logistics. However, task allocation and scheduling for a multi-agent team face complex challenges due to the need to simultaneously consider time-extended tasks, task constraints, and uncertainties in execution. Potential task failures or contingencies can add additional tasks to recover from the failures, and reactively addressing contingencies can decrease teaming efficiency. To efficiently and proactively consider contingencies, this paper proposes treating the problem as a multi-robot task allocation under uncertainty problem. We suggest a hierarchical approach that divides the problem into two layers. We use mathematical program formulation for the lower layer to find the optimal solution for a deterministic multi-robot task allocation problem with known task outcomes. The higher-layer search intelligently generates more likely combinations of contingency scenarios and calls the inner-level search repeatedly to find the optimal task allocation sequence for the given scenario. We validate our results in simulation for manufacturing applications and demonstrate that our method can reduce the effect of potential delays from contingencies.Note to Practitioners—Automation engineers interested in deploying robotic cells in low-volume applications need to consider contingency handling. When the occurrence of contingencies can be characterized as probability distributions, it is often useful to consider using a proactive approach for task allocation and scheduling. To implement our algorithm, automation engineers will need to develop a hierarchical task network specified by domain experts that models task constraints and a task-agent duration model, which may be generated from simulation environments. Furthermore, they must identify tasks that can result in contingencies and describe them with a probabilistic model. This model can be generated from historical data and/or real-world experiments. Lastly, for addressing the contingency, the practitioner will need to specify a task procedure to recover from a specific contingency type. To run the algorithm, we found that repeatedly approximating the best proactive task allocation for a fixed computation budget and dispatching the best tasks worked well. The computation budget required to approximate the best task allocation is directly affected by the number of contingency scenarios that can be sampled. Therefore, the practitioner must determine a suitable computational budget empirically based on the number of contingencies that can occur.}, archive = {J_TASE}, author = {Neel Dhanaraj and Heramb Nemlekar and Stefanos Nikolaidis and Satyandra K. Gupta}, doi = {10.1109/TASE.2025.3546281}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13046--13060}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Proactive contingency-aware task allocation and scheduling in multi-robot multi-human cells via hindsight optimization}, volume = {22}, year = {2025}, }
• (2025). Universal trajectory optimization framework for differential drive robot class. TASE, 22, 13030--13045. (www)

  @article{ , abstract = {Differential drive robots are widely used in various scenarios thanks to their straightforward principle, from household service robots to disaster response field robots. The nonholonomic dynamics and possible lateral slip of these robots lead to difficulty in getting feasible and high-quality trajectories. Although there are several types of driving mechanisms for real-world applications, they all share a similar driving principle, which involves controlling the relative motion of independently actuated tracks or wheels to achieve both linear and angular movement. Therefore, a comprehensive trajectory optimization to compute trajectories efficiently for various kinds of differential drive robots is highly desirable. In this paper, we propose a universal trajectory optimization framework, enabling the generation of high-quality trajectories within a restricted computational timeframe for these robots. We introduce a novel trajectory representation based on polynomial parameterization of motion states or their integrals, such as angular and linear velocities, which inherently matches the robots’ motion to the control principle. The trajectory optimization problem is formulated to minimize computation complexity while prioritizing safety and operational efficiency. We then build a full-stack autonomous planning and control system to demonstrate its feasibility and robustness. We conduct extensive simulations and real-world testing in crowded environments with three kinds of differential drive robots to validate the effectiveness of our approach.Note to Practitioners—The Differential drive robot, known for its simple mechanics and high maneuverability, is widely used in many applications. However, current methods have limitations in practice when high-performance motion is needed. Due to the state representation in Cartesian space, path planning makes it difficult to consider nonholonomic constraints directly. The existing trajectory optimization cannot effectively constrain the angular velocity and it is difficult to model forward and backward motion into a continuous trajectory. This paper provides a novel trajectory representation that inherently utilizes the motion performance of differential drive robots, which ensures its universality for different platforms, and reduces the time required to generate trajectories to ensure real-time performance. Based on this, we propose a robust planning and control framework to achieve efficient navigation. We release the source code at https://zju-fast-lab.github.io/DDR-opt/ facilitating expansion and deployment for practitioners. We validate this framework through extensive experiments, demonstrating its capability to navigate challenging environments.}, archive = {J_TASE}, author = {Mengke Zhang and Nanhe Chen and Hu Wang and Jianxiong Qiu and Zhichao Han and Qiuyu Ren and Chao Xu and Fei Gao and Yanjun Cao}, doi = {10.1109/TASE.2025.3550676}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13030--13045}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Universal trajectory optimization framework for differential drive robot class}, volume = {22}, year = {2025}, }
• (2025). Byzantine-resilient distributed state estimation: A distance-based multivariable filtering mechanism. TASE, 22, 13015--13029. (www)

  @article{ , abstract = {This paper studies the problem of resilient distributed state estimation for a linear system using a network of agents, some of which are subject to the Byzantine attacks. First, by introducing the distance function to quantify the difference between the estimates of the neighboring agents, a multivariable filtering mechanism is designed such that the regular agents can extract the reliable information from the vectors sent by their in-neighbors. Then, using the properties of the distance-based multivariable filtering mechanism and the detectability decomposition, resilient distributed observers are designed for the regular agents to asymptotically estimate the state vector of the system despite the adversarial influence of the Byzantine attacks. Furthermore, a graph-dependent Lyapunov function is proposed to analyze the convergence of the proposed method. In contrast to the existing scalar filtering mechanism-based methods, the proposed method can reduce the complexity of the reliable information extraction, and does not require the existence of multiple individual agents to detect each unstable eigenvalue of the system matrix. Finally, an example is given to demonstrate the effectiveness of the proposed method. Note to Practitioners—This paper is motivated by the problem of collaboratively estimating the state vector of a dynamical system using a network of agents in an attack-prone environment. Existing methods for regular agents to extract the reliable information from the received vectors generally rely on the scalar filtering mechanism, which increase in complexity as the dimension of the system increases and require relatively conservative observability assumption. In contrast, we propose a multivariable filtering mechanism-based resilient distributed state estimation method that has lower complexity and works under milder observability assumption. In future research, we will consider the applications of the proposed method to address the related problems such as target tracking, environmental monitoring, surveillance and patrolling.}, archive = {J_TASE}, author = {Rui Gao and Guang-Hong Yang}, doi = {10.1109/TASE.2025.3549054}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13015--13029}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Byzantine-resilient distributed state estimation: A distance-based multivariable filtering mechanism}, volume = {22}, year = {2025}, }
• (2025). Heterogeneous attention-based graph convolutional network for solving asymmetric pickup and delivery problem. TASE, 22, 13000--13014. (www)

  @article{ , abstract = {In recent years, there has been a notable increase in demand for pickup and delivery services, mainly driven by the expansion of e-commerce. These services can be conceptualized as pickup and delivery problems (PDPs), which are important variants of vehicle routing problems (VRPs). While neural methods based on deep reinforcement learning (DRL) have demonstrated success in solving VRPs, current neural methods for PDP predominantly depend on customer coordinates and Euclidean distances. This heavy reliance can diminish their practical value, given the intricacies of real-world road networks and inherent asymmetries. This paper presents a novel learning-based method, i.e., HA-GCN, that couples heterogeneous attention (HA) and a graph convolutional network (GCN) to tackle the asymmetric pickup and delivery problem (APDP). The HA-GCN model utilizes HA to comprehend the innate pairing and precedence constraints between nodes in APDP. Concurrently, it employs GCN to amalgamate features of nodes and edges, facilitating the extraction of asymmetric attributes. Additionally, to augment the real-world relevance of HA-GCN, we train the model grounded in a dataset drawn from real geographic information. Comprehensive experiments suggest that our proposed method performs favorably against both traditional and learning-based state-of-the-art approaches, while demonstrating robust generalization capabilities. Note to Practitioners—Research on the pickup and delivery problem (PDP) addresses challenges in the rapidly growing field of end-to-end delivery services, such as food and parcel delivery. This study focuses on the asymmetric pickup and delivery problem (APDP), a variant of PDP reflecting real-world road networks with one-way streets and varying distances between locations. However, due to the exponential complexity and asymmetric nature of APDP, existing traditional methods and deep reinforcement learning (DRL) approaches struggle to effectively address this challenge. To address this, we propose HA-GCN, a DRL-based model compling heterogeneous attention and a graph convolutional network. HA-GCN effectively distinguishes pickup and delivery nodes in APDP, capturing differences in asymmetric distances in real-world road networks. Experimental results based on instances of different scales show that HA-GCN outperforms learning-based and traditional baselines, exhibiting satisfactory generalization capabilities. Given these confirmed advantages, our HA-GCN has great potential to not only offer the practitioners effective alternative approaches, but also improve the solution quality for the pickup and delivery in the operation of express hubs or logistics centers in specific regions of the real world.}, archive = {J_TASE}, author = {Jiayi Li and Guohua Wu and Mingfeng Fan and Zhiguang Cao and Yalin Wang}, doi = {10.1109/TASE.2025.3548141}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {13000--13014}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Heterogeneous attention-based graph convolutional network for solving asymmetric pickup and delivery problem}, volume = {22}, year = {2025}, }
• (2025). FCF-CSM: A fuzzy clustering framework based on chromaticity statistical model for automatic segmentation of port wine stains. TASE, 22, 12986--12999. (www)

  @article{ , abstract = {Extracting lesions with accurate boundaries from clinical images is crucial for the clinical diagnosis, progression monitoring, and efficacy evaluation of port wine stains (PWS). However, accurately delineating lesion boundaries remains challenging due to complex boundary structures and unreliable annotations. In this paper, we propose a fuzzy clustering framework, FCF-CSM, for automated segmentation of PWS lesions. It combines prior knowledge of PWS color distribution with superpixels’ boundary characterization capability. Firstly, a chromaticity statistical model (CSM) is established based on 1000 collected PWS images, providing prior probabilities of PWS lesions to guide the improvement of the fuzzy clustering framework. Secondly, a superpixel method incorporating CSM is applied to PWS images, generating superpixels with improved boundary characterization. Thirdly, these superpixels are finely clustered using color features related to erythema index, color statistics, and color volume, improving PWS lesion distinction from complex backgrounds. Finally, a CSM-based automatic decision method distinguishes lesions from the background, achieving fully automated PWS segmentation within a fuzzy clustering framework. In addition, a boundary local fitting (BLF) metric is proposed to evaluate the segmentation precision of the PWS boundaries. Comparative experiments are conducted to verify the superiority of FCF-CSM. It achieves comparable overall segmentation performance with Jaccard and Dice metrics of 84.14% and 91.11%, respectively, compared to state-of-the-art methods. In terms of boundary segmentation, FCF-CSM outperforms other methods with an 81.52% BLF metric. FCF-CSM has proven to be effective for PWS segmentation and is promising to improve boundary delineation. The code is available at https://github.com/JinrongMu/FCF-CSM. Note to Practitioners—The motivation of this study was to construct a statistical model of port wine stain (PWS) color to quantify prior knowledge of PWS lesion color in RGB images. Existing methods for automatic segmentation of PWS rely heavily on annotated data, but the lack of publicly available datasets hinders the development of such algorithms due to the privacy of clinical data. This paper proposes a fuzzy clustering framework based on the chromaticity statistical model, which can achieve high-precision and fine-grained delineation of the boundaries of PWS lesions without annotating data. In this study, we describe the prior of PWS color distribution based on colorimetry theory and apply this knowledge to the PWS automatic segmentation task, thereby realizing knowledge sharing while protecting patient privacy from being leaked. Comprehensive comparative experiments demonstrate the effectiveness and reliability of the chromaticity statistical model. However, the dataset used to build this model only includes populations with yellow skin tones. In future research, we will address the automatic identification of PWS lesions applicable to other skin tones.}, archive = {J_TASE}, author = {Jinrong Mu and Yuanyuan Wang and Hong Song and Xianqi Meng and Yunqi Li and Jingfan Fan and Danni Ai and Defu Chen and Haixia Qiu and Jian Yang and Ying Gu}, doi = {10.1109/TASE.2025.3550681}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12986--12999}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FCF-CSM: A fuzzy clustering framework based on chromaticity statistical model for automatic segmentation of port wine stains}, volume = {22}, year = {2025}, }
• (2025). Variational-based geometric nonlinear model predictive control for robust locomotion of quadruped robots. TASE, 22, 12975--12985. (www)

  @article{ , abstract = {This paper proposes a novel nonlinear model predictive control (NMPC) method based on geometric variational calculus for high-dynamic and complex motion control of quadruped robots. By approximating system trajectory tracking error dynamics on the Special Euclidean group (SE(3)), the method avoids the singularities of Euler angles and the challenges of quaternion representation while capturing the coupling between rotational and translational dynamics for a more comprehensive motion description. Leveraging variational calculus, the resulting Geometric Nonlinear Model Predictive Controller (GNMPC) enables high-frequency updates while preserving essential nonlinear system characteristics. Experimental results across various scenarios validate the effectiveness and advantages of the proposed controller. Note to Practitioners—The primary motivation of this paper is to investigate the application of geometric methods in Model Predictive Control (MPC) and to validate their effectiveness in the context of quadruped robots, which exhibit nonlinear dynamics. In this work, the authors model the robot’s motion on a nonlinear manifold and linearize the system using variational methods. Sequential Quadratic Programming (SQP) is then applied to approximate the globally optimal solution. Experimental results demonstrate that this approach significantly improves the performance of quadruped robots, particularly in handling highly dynamic and robust motions.}, archive = {J_TASE}, author = {Botao Liu and Fei Meng and Sai Gu and Xuechao Chen and Zhangguo Yu and Qiang Huang}, doi = {10.1109/TASE.2025.3546734}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12975--12985}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Variational-based geometric nonlinear model predictive control for robust locomotion of quadruped robots}, volume = {22}, year = {2025}, }
• (2025). Robust multiple autonomous underwater vehicle cooperative localization based on the principle of maximum entropy. TASE, 22, 12960--12974. (www)

  @article{ , abstract = {Cooperative localization aims to continuously provide position estimates for multiple-autonomous underwater vehicle (multi-AUV) systems during task execution such as marine monitoring; it is preferable over noncooperative schemes due to its high accuracy and strong robustness. However, various uncertain factors underwater, including model mismatches, accumulated errors, measurement noises and biases, time-varying communication channels, etc., still challenge the accuracy and robustness of cooperative localization. When such uncertainties arise, the performances of traditional methods degrade significantly. Therefore, this paper proposes a robust multi-AUV cooperative localization method that is able to combat these uncertainties by leveraging the principle of maximum entropy. To be explicit, a message-passing scheme is established using factor graphs, over which a distributed position estimation strategy for AUVs is designed based on belief propagation. To reduce the damages of uncertainties, maximum-entropy distributions are designed respectively for the prediction and correction processes of localization and are realized by particles. Specifically, through enlarging the particle coverage, uncertainty-induced misleading in position estimation is alleviated, and hence higher robustness is achieved. Simulations and field experiments show the advantages of the proposed algorithm over the state-of-the-art methods in terms of localization accuracy, robustness, and scalability.Note to Practitioners—In practical applications, due to the unavailability of global positioning systems underwater, AUV localization still lacks mature and stable solutions. In harsh underwater environments, most theoretical models often fail to accurately describe the practical conditions, leading to widespread mismatches, which severely degrade localization performance. Moreover, uncertainties such as accumulated errors, measurement noises, and position deviations would further reduce the position estimation accuracy. Existing methods usually consider these uncertainties independently, while, in practical uses, uncertainties often emerge in combination, making it challenging to maintain localization accuracy. What is worse is that most sophisticatedly designed algorithms pursue high accuracy, and their adaptability in engineering applications is difficult to guarantee. Hence, this work provides a robust solution for multi-AUV cooperative localization that can handle uncertainties simultaneously. By reducing the sensitivity to different uncertainties, the proposed algorithm can provide sustained high-accuracy position estimates for AUVs in complex underwater environments. The proposed method is experimentally verified and suitable for multi-AUV applications such as oceanic rescue, resource development, and marine monitoring.}, archive = {J_TASE}, author = {Yichen Li and Wenbin Yu and Haotian Xu and Xinping Guan}, doi = {10.1109/TASE.2025.3546674}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12960--12974}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust multiple autonomous underwater vehicle cooperative localization based on the principle of maximum entropy}, volume = {22}, year = {2025}, }
• (2025). Fractional-order lyapunov-based backstepping-like feedback control of N-DOF mechanical systems. TASE, 22, 12949--12959. (www)

  @article{ , abstract = {This paper studies a fractional-order control and stability scheme for a class of n-DOF mechanical systems that are subject to actuator saturation and unknown disturbances. By the fractional-order Leibniz rule, a generalized fractional-order Lyapunov stability (FOLS) method is developed to support the design of a class of fractional-order robust controllers, which can guarantee the uniform ultimate boundedness and faster convergence. Based on this, a fractional-order backstepping-like feedback controller (FOBFC) is raised to realize an effective and stable control performance of mechanical systems, which can ameliorate the shortages of the traditional backstepping control effectively. Moreover, a set of fractional-order methods, including actuator compensation (FOAC) and disturbance observer (FODO), are proposed and involved in FOBFC to handle the actuator saturation and unknown time-varying disturbances, making it possible to achieve a more robust and stable control performance. Finally, numerical simulations and comparative experiments are carried out to demonstrate the effectiveness and superiority of the proposed fractional-order control scheme, in n-DOF mechanical systems. Note to Practitioners—A fractional-order control and stability scheme is developed to regulate a class of practical n-DOF mechanical systems, which is composed of FOBFC, FOAC and FODO. It can be applied to deal with the actuator saturation and time-varying disturbance of n-DOF mechanical systems, contributing to realizing a more robust and stable control performance. In the application, a Euler-Lagrange dynamics of mechanical systems is required for constructing the fractional-order tracking errors. Subsequently, FOBFC law can be raised based on the generalized FOLS, to guarantee the asymptotic convergence and stability of mechanical systems. To address the practical issues of mechanical systems, FOAC and FODO are necessary for the proposed control scheme to compensate for the actuator saturation adaptively, and handle the unknown and time-varying disturbances effectively.}, archive = {J_TASE}, author = {Xiaolei Li and Xiaogang Wang and Qixin Kui and Weiran Yao and Guanghui Sun}, doi = {10.1109/TASE.2025.3548986}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12949--12959}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fractional-order lyapunov-based backstepping-like feedback control of N-DOF mechanical systems}, volume = {22}, year = {2025}, }
• (2025). Global event-triggered adaptive stabilization of nonlinear time-delay systems with unknown measurement sensitivity. TASE, 22, 12937--12948. (www)

  @article{ , abstract = {This paper addresses the problem of global stabilization in nonlinear time-delay systems with unknown measurement sensitivity. Notably, our system allows for the unknown measurement sensitivity to be non-differentiable, coupled with unmeasurable states, which necessitates the development of observers and control input strategies based on dynamic gain. To optimize resource usage and mitigate network congestion, we introduce an event-triggering mechanism based on two events. This mechanism evaluates dynamic gain and control signals, ensuring a guaranteed positive lower bound on execution time. Moreover, the dynamic gain is designed to compensate for the impact of execution errors. The introduction of a relational sensitivity error allows the unknown measurement sensitivity to converge to a small range. By selecting appropriate Lyapunov-Krasovskii functionals, we eliminate the influence of time-delay, ultimately proving global stability of the closed-loop system. Consequently, comparative simulation results validate the effectiveness of the proposed scheme. Note to Practitioners—This paper explores event-triggered control of nonlinear systems with applications in intelligent transportation, robotics, and aerospace. Notably, we address three critical issues. Firstly, we propose methods to manage unmeasurable states, which is essential for systems like autonomous vehicles and drones where full state measurement is often not feasible. Secondly, we improve existing event-triggering mechanisms to optimize network resource usage, significantly reducing communication frequency, which is particularly beneficial in networked control systems, such as smart grids and industrial automation. Thirdly, we tackle the challenge of unknown measurement sensitivity, relevant for systems operating in uncertain environments like industrial robots and aerospace applications where sensor accuracy can vary. Comparative simulations show that our dynamic event-triggered control method effectively reduces network burden, proving its practical value in real-world applications where network bandwidth and reliability are crucial. In future research, we will also consider reducing the transmission frequency from sensor to controller to further improve system performance. Additionally, we aim to explore adaptive mechanisms to better manage uncertainties in measurement sensitivity, thereby expanding the practical applicability of our approach.}, archive = {J_TASE}, author = {Cheng Tan and Xinrui Ma and Yuzhe Li and Xiangpeng Xie}, doi = {10.1109/TASE.2025.3550186}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12937--12948}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Global event-triggered adaptive stabilization of nonlinear time-delay systems with unknown measurement sensitivity}, volume = {22}, year = {2025}, }
• (2025). A data-driven control algorithm with time delay compensation based on optimized recursive adaptive identification for magnetic levitation systems. TASE, 22, 12922--12936. (www)

  @article{ , abstract = {This study addresses the challenge of precise control of the maglev system under the condition of rapid time-varying dynamics and large time delay. Maglev technology has been applied in many fields, such as maglev train and magnetic bearing, for its non-contact operation, low maintenance cost and high energy efficiency. However, the inherent nonlinear characteristics and delayed response characteristics of the magnetic levitation system make it very difficult to control. In practical scenarios, it is difficult to establish an accurate magnetic suspension system model, which limits the use of model-based control methods. PID control method is also difficult to guarantee the control performance when dealing with unmeasurable load disturbance and system delay scenarios. This study proposes a data-driven control solution, which integrates an optimized recursive identification algorithm and an adaptive delay compensation algorithm. Uniquely, the control solution achieves real-time iterative optimization with only a few input and output data, reducing the data storage resource requirements, avoiding offline training, and ensuring real-time performance and efficient resource utilization in practical applications. Practitioners can use this new method to enhance the performance and reliability of maglev applications, such as maglev trains, industrial processes and precision instruments, so as to achieve more robust and adaptive control in the face of rapid dynamic changes and large time delays. Note to Practitioners—This study addresses the challenge of precise control of the maglev system under the condition of rapid time-varying dynamics and large time delay. Maglev technology has been applied in many fields, such as maglev train and magnetic bearing, for its non-contact operation, low maintenance cost and high energy efficiency. However, the inherent nonlinear characteristics and delayed response characteristics of the meglev system make it very difficult to control. In practical scenarios, it is difficult to establish an accurate magnetic suspension system model, which limits the use of model-based control methods. PID control method is also difficult to guarantee the control performance when dealing with unbearable load disturbance and system delay scenarios. This study proposes a data-driven control solution, which integrates an optimized recursive identification algorithm and an adaptive delay compensation algorithm. Uniquely, the control solution achieves real-time iterative optimization with only a few input and output data, reducing the data storage resource requirements, avoiding offline training, and ensuring real-time performance and efficient resource utilization in practical applications. Practitioners can use this new method to enhance the performance and reliability of maglev applications, such as maglev trains, industrial processes and precision instruments, so as to achieve more robust and adaptive control in the Phase of rapid dynamic changes and large time delays.}, archive = {J_TASE}, author = {Zhen Li and Shangtai Jin and Yuzhou Wei and Honghai Ji and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3543411}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12922--12936}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A data-driven control algorithm with time delay compensation based on optimized recursive adaptive identification for magnetic levitation systems}, volume = {22}, year = {2025}, }
• (2025). Efficient RGB-D co-salient object detection via modality-aware prompting. TASE, 22, 12911--12921. (www)

  @article{ , abstract = {RGB-D co-salient object detection (Co-SOD) aims to identify and segment co-occurring salient objects in a set of correlated images and depth maps. Most existing RGB-D Co-SOD methods fully fine-tune the dual-stream encoder-decoder architecture and fuse the RGB and depth features using a complex feature fusion strategy, which is expensive to train owing to the large number of parameters that need to be updated during the feature extraction and fusion process. In addition, current methods do not pay sufficient attention to differentiate co- salient information from non-co- salient information effectively. This interfering information affects the localization of co-salient targets. Therefore, this study proposes a simple and effective modality-aware prompting network (MAPNet) for efficient RGB-D Co-SOD. MAPNet mainly performs RGB-D Co-SOD through two approaches, namely modal fusion and consensus feature extraction, using a multimodal prompt generator (MPG) and consensus feature extraction module (CFEM), respectively. Specifically, the MPG module guides the depth features in the fine-tuned backbone network from the RGB features obtained in the frozen backbone network for fusion in hyperbolic spaces to generate multilevel modal cues that are subsequently injected into the fine-tuned backbone network for efficient modal fusion. The CFEM uses RGB features to generate an image salient prior, combines the salient prior with the highest level of fusion features to obtain the central point, and uses the salient features closer to the central point as the consensus features of the image group. In addition, contrast loss is introduced to separate the synergistic and non-synergistic salient features to obtain pure co-salient features. The trained MAPNet delivered state-of-the-art performance on three benchmark datasets (RGB-D CoSal1k, RGB-D CoSal150, and RGB-D CoSeg183), with the structure-measure improved by 2.1% on the RGB-D CoSeg183 dataset. The codes are available at https://github.com/trumpetor/MAPNet. Note to Practitioners—This study presents a straightforward and effective modality-aware prompting network (MAPNet) designed for efficient RGB-D Co-SOD. Initially, the MPG module of MAPNet enables the effective integration of RGB and depth modalities through prompt learning. Subsequently, the CFEM employs the pixel group centroid proxy and top-k selection mechanism to extract high-level integrated features and salient prior consensus features, which serve as coordinated saliency features for image groups. Finally, the coordinated obtained salient and integrated features are input into the decoder to generate predictions.}, archive = {J_TASE}, author = {Zhangping Tu and Xiaohong Qian and Wujie Zhou}, doi = {10.1109/TASE.2025.3543586}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12911--12921}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient RGB-D co-salient object detection via modality-aware prompting}, volume = {22}, year = {2025}, }
• (2025). Deep comparison convolutional neural network for industrial localization. TASE, 22, 12899--12910. (www)

  @article{ , abstract = {Accurate visual localization is essential for industrial applications, e.g., automated workpiece localization and manipulation. Traditional template matching methods are limited by lighting conditions and lack kinematic flexibility, while Convolutional Neural Network (CNN)-based approaches necessitate retraining for each specific application, hindering broader use. This article presents UniShot, a deep comparison CNN tool with a triple-CNN architecture that extracts features and predicts observer coordinates through image comparison. We also derived view-frame kinematics (VFK) element transfer models, which facilitate visual data generation. Trained on diverse scenes, UniShot is universally applicable without needing application-specific retraining, providing a significant advantage over existing methods. Experimental results show that UniShot outperforms conventional techniques and OneShot in dynamic environments, achieving superior accuracy and operational range compared to state-of-the-art methods like YOLOv8. Saliency map analyses further validate UniShot’s comparative localization mechanism. We announce that UniShot is now open-source. Note to Practitioners—Accurate object positioning, such as aligning a workpiece on a fixture, is crucial in manufacturing. Traditional methods like template matching rely on controlled lighting and specific poses, limiting their effectiveness in varied environments, such as those encountered by autonomous mobile robots. This study introduces UniShot, a deep learning-based model that enhances adaptability to lighting and kinematic variations. Utilizing a dual-input neural network, UniShot compares baseline and query images to output the observer’s relative coordinates. Trained on millions of images, it requires no retraining for new applications. Users only need to provide a baseline image while the camera monitors the scene in real time. UniShot achieves accuracy above 90% and processes images at over ten frames per second. Experimental results show that UniShot significantly outperforms traditional template matching, YOLOv8, and OneShot methods.}, archive = {J_TASE}, author = {Bo-Xu Wen and Qinxin Zhan and Hongxiang Zhuang and Chih-Hung G. Li}, doi = {10.1109/TASE.2025.3548279}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12899--12910}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep comparison convolutional neural network for industrial localization}, volume = {22}, year = {2025}, }
• (2025). Skeleton-aware representation of spatio-temporal kinematics for 3D human motion prediction. TASE, 22, 12884--12898. (www)

  @article{ , abstract = {3D human motion prediction, which attempts to foresee the behaviors of human, is an issue of great significance in computer vision. Attention-based neural networks and graph convolution networks (GCNs) have recently shown great promise in 3D skeleton-based human motion prediction for their attractive performance in learning spatial and temporal kinematics. However, existing methods have several critical issues: 1) Spatial dependencies for distal joints in each independent frame are hard to learn; 2) The GCN ignores hierarchical structure and diverse motion patterns of different body parts; 3) Existing methods disregard the statistical interdependence inherent in time series data. To address these issues, this paper proposes a skeleton-aware representation of spatio-temporal kinematics for 3D human motion prediction. The proposed method makes three key contributions: a learnable temporal aggregation, a skeleton-aware spatio-temporal attention, and an upper/lower decoupling GCN. The learnable temporal aggregation selectively obtains past information by leveraging the dependencies between each time step and its historical moments. The skeleton-aware spatio-temporal attention method leverages the self-attention mechanism and a designed adjacency matrix to model the skeleton constraints of distal joints. The upper/lower decoupling GCN introduces a grouping strategy to learn the dynamics of various body parts separately. Experimental results on three publicly available datasets demonstrate that the proposed method achieves state-of-the-art performances for both short-term prediction and long-term prediction. Note to Practitioners—3D human motion prediction forms a fundamental component of human-centered automation systems by enabling safer, more efficient, and more natural interactions between humans and machines. This paper was motivated by the challenges of predicting human motion: 1) Explicitly capturing the complex spatial patterns of distal joints is challenging; 2) Neglecting the inter-part variations of motion dynamics is problematic; 3) Neglecting the statistical interdependence inherent in time series data of human motion leads to poor performance. This paper suggests a skeleton-aware representation of spatio-temporal kinematics for 3D human motion prediction through three innovations: a learnable temporal aggregation, a skeleton-aware spatio-temporal attention, and an upper/lower decoupling GCN. The three contributions overcome the weaknesses of existing works and made a pioneering attempt of skeleton-aware representation of spatio-temporal human kinematics. It will significatively advance the development of many automation systems relevant to human motion prediction such as human-robot interaction and teleoperation.}, archive = {J_TASE}, author = {Songlin Du and Zhihan Zhuang and Zenghui Wang and Yuan Li and Takeshi Ikenaga}, doi = {10.1109/TASE.2025.3548846}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12884--12898}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Skeleton-aware representation of spatio-temporal kinematics for 3D human motion prediction}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance robust approximate optimal tracking control via stackelberg game. TASE, 22, 12871--12883. (www)

  @article{ , abstract = {Real-world applications of nonlinear systems tracking control are always challenging due to the existence of uncertainties and disturbances. To design a robust optimal tracking controller for uncertain nonlinear systems with disturbances and actuator saturation, this paper investigates the prescribed performance robust optimal tracking control problem. A prescribed performance mechanism is constructed to convert the dynamics of tracking error into transformed error dynamics, which keeps the system’s operating states within specific bounds, ensuring tracking with predefined error constraints. For the optimal tracking controller design, an optimal index is established to optimize the performance of tracking control, and a robust optimal index is established to optimize the disturbance effect on the tracking error. To achieve robust optimal tracking control that minimizes both optimal and robust optimal indexes, a Stackelberg game is constructed, which provides a hierarchical game structure for the optimal controller and the worst disturbance. The robust optimal controller is approximated online using reinforcement learning techniques. An actor-critic-identifier algorithm is designed to approximate the optimal value function, optimal controller, and drifted system parameters. Lyapunov theory is utilized to analyze the closed-loop system’s stability. To demonstrate the effectiveness of the proposed robust optimal control method, two numerical simulations and a hardware experiment on a quadcopter system are conducted. The experiment results demonstrate that our method successfully achieves prescribed performance tracking control when actuators are saturated and disturbances are present. Note to Practitioners—In this paper, the probelm of mixed $H_{2}/H_{\infty }$ prescribed-performance optimal tracking control for nonlinear systems with input saturation is investigated. To constrain the operating states of the system within certain bounds, the prescribed performance transformation is designed to achieve tracking with predefined error constraints. For the optimal controller design, the $H_{2}$ index is established to minimize the optimal tracking performance, and the $H_{\infty }$ index is designed to minimize the disturbance effect on the tracking error. A Stackelberg-based non-zero sum game between the optimal controller and the worst disturbance is established to design the mixed $H_{2}/H_{\infty }$ optimal tracking controller. The designed optimal controller is approximated online using reinforcement learning. Effectiveness of the proposed method is demonstrated by two numerical simulations and a hardware experiment on a quadcopter system. Based on the proposed high-performance controller, engineers can design a high-performance robust optimal tracking controller for uncertain nonlinear systems with extreme conditions of disturbances and actuator saturation.}, archive = {J_TASE}, author = {Junkai Tan and Shuangsi Xue and Huan Li and Zihang Guo and Hui Cao and Dongyu Li}, doi = {10.1109/TASE.2025.3549114}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12871--12883}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance robust approximate optimal tracking control via stackelberg game}, volume = {22}, year = {2025}, }
• (2025). An adaptive wind power prediction model based on cross-attention mechanism with incremental alignment. TASE, 22, 12857--12870. (www)

  @article{ , abstract = {Wind energy is a clean energy source that can effectively reduce pollution. Efficient and accurate wind power forecasting helps optimize the operation and scheduling of wind farms, thereby enhancing energy utilization. The intensity of wind energy is closely related to weather information, and these data are in the form of time series. Currently, most time series forecasting models use multi-head self-attention mechanisms and gated recurrent units. However, the self-attention mechanism has disorderly characteristics, which may disrupt the data sequence and lead to the loss of temporal information; meanwhile, gated recurrent units have limited effectiveness in capturing information among multiple variables, making it challenging to significantly improve prediction accuracy. To address these shortcomings, this study innovatively introduces a cross-attention mechanism into the core prediction algorithm and proposes a new forecasting module, the CA module, which can capture information among variables while maintaining the orderliness of the time series. Furthermore, to address the time lag between weather information and power data, this study designs a novel data alignment algorithm based on increments and combines it with variational mode decomposition to achieve multi-source feature alignment, further improving prediction accuracy. Additionally, a reinforcement learning algorithm is used as a hyperparameter optimization framework during the forecasting process to determine the optimal hyperparameters for different tasks, thus enhancing the model’s accuracy and practicality. The experiments are based on weather and corresponding power data from four wind power stations in eastern China during spring, including ablation, comparison, and generalization experiments. The results show that the proposed model improves wind power output forecasting accuracy by more than 15% across different time scales compared to current mainstream algorithms, with short- to medium-term prediction accuracy improvements of up to 58.1%. Note to Practitioners—The motivation of this paper is to improve the accuracy of wind power generation forecasting, thereby facilitating better regulation and dispatch of wind farms and enhancing resource utilization. Most existing forecasting models are based on methods such as Transformer and LSTM. However, the disorderly nature of these models and their limited ability to capture information among multiple variables constrain their performance, making it difficult to achieve breakthrough improvements. This paper proposes a novel forecasting module based on cross-attention. Initially, the raw data is decomposed into different frequency modes, and then the similarity between power and weather factors at different frequencies is calculated to assign different weights, enabling the model to fully capture the correlation information among multiple variables. Additionally, this paper introduces an incremental data alignment algorithm, which combines the multi-step average relative growth rate with an exponential weighted average to construct a new curve, which is used for dynamic time warping. This significantly improves alignment and eliminates the time lag effect of weather factors on power. The experimental results show that the proposed solution significantly improves prediction accuracy compared to current mainstream models. However, due to the potential influence and similarity between different variables and frequency bands, future research will explore using cross-band forecasting to address this issue. Moreover, this model can also be applied to other multi-dimensional time series forecasting tasks, such as weather forecasting, traffic congestion, and generator oil temperature prediction, demonstrating high practical value.}, archive = {J_TASE}, author = {Li Jiang and Yifan Wang and Zhaowei Lu and Yuebing Gao}, doi = {10.1109/TASE.2025.3548789}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12857--12870}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptive wind power prediction model based on cross-attention mechanism with incremental alignment}, volume = {22}, year = {2025}, }
• (2025). Wind farm control via offline reinforcement learning with adversarial training. TASE, 22, 12845--12856. (www)

  @article{ , abstract = {In a wind farm, wakes produced by upstream wind turbines significantly diminish the wind capture of downstream ones, resulting in reduced power generation for the entire farm. Reinforcement learning (RL) control can alleviate the wake effect by enhancing coordination among turbines in arrays. However, training such a collaborative control policy through online RL necessitates millions of interactions with a computational fluid dynamics-based simulator, making it computationally expensive. Wind farm operators possess extensive datasets from past operations, which can tackle the sample generation difficulty. Nevertheless, online RL struggles with the direct use of these offline datasets due to covariate shift and biased value estimation. In this paper, we introduce an offline RL method named Multi-Agent Offline Behavior Imitation (MAOBI) to cope with this challenge. First, by employing f-GANs (Generative Adversarial Networks), MAOBI estimates the divergence between the learning policy and the behavior policy based on samples generated by them. By minimizing this divergence, MAOBI can mimic the behavior policy hidden in the offline datasets. The method then identifies state-action pairs that yield high returns, further improving the control policy. Results demonstrate that MAOBI-trained control policies achieve performance comparable to state-of-the-art online RL methods when deployed in a high-fidelity wind farm simulator. Note to Practitioners—This study addresses a pressing demand in the wind industry: developing high-performance control systems for wind farms to maximize power generation. Although online RL is a popular approach for training collaborative control policies, its reliance on constructing wind farm environments consumes substantial computational resources, such as a supercomputer, to generate training samples. To overcome this limitation, we propose an offline RL algorithm. This algorithm leverages historical operational data from wind farms to search for optimal control policies, eliminating the need for a virtual environment. Our method offers advantages such as being model-free, communication-free, and capable of real-time response.}, archive = {J_TASE}, author = {Yubo Huang and Xiaowei Zhao}, doi = {10.1109/TASE.2025.3548565}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12845--12856}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Wind farm control via offline reinforcement learning with adversarial training}, volume = {22}, year = {2025}, }
• (2025). Robust second-order LiDAR bundle adjustment algorithm using mean squared group metric. TASE, 22, 12831--12844. (www)

  @article{ , abstract = {The bundle adjustment (BA) algorithm is a widely used nonlinear optimization technique in simultaneous localization and mapping (SLAM) systems. By leveraging the co-view relationships of landmarks from multiple perspectives, the BA method constructs a joint estimation model for both poses and landmarks, enabling the system to generate refined maps and reduce front-end localization errors. However, exploring a robust LiDAR BA estimator and achieving accurate solutions is a challenge. In this work, firstly we propose a novel mean square group metric (MSGM) to build the optimization objective of the LiDAR BA algorithm. This metric applies a mean square transformation to uniformly process the measurements of plane landmarks during one sampling period. The transformed metric ensures scale interpretability and does not require a time-consuming point-by-point calculation. Secondly, by integrating a robust kernel function, the metrics involved in the BA algorithm are reweighted, thus enhancing the robustness of the solution process. Thirdly, based on the proposed robust LiDAR BA model, we derived an explicit second-order estimator (RSO-BA). This estimator employs analytical formulas for Hessian and gradient calculations, ensuring the precision of the BA solution. Finally, we verify the merits of the proposed RSO-BA estimator against existing implicit second-order and explicit approximate second-order estimators using publicly available datasets and physical experiments. The experimental results demonstrate that the RSO-BA estimator outperforms its counterparts in terms of registration accuracy and robustness, particularly in dynamic or complex unstructured environments. Note to Practitioners—The motivation of this paper is to develop a novel LiDAR bundle adjustment (BA) algorithm that ensures accurate and consistent 3D scene modeling. Currently, most LiDAR BA algorithms use “group” processing to construct cost metrics, aiming to reduce the computational complexity of point-by-point operations. However, the cost metrics of these approaches lack scale interpretability, which makes it difficult to incorporate robust kernel functions into the model design, ultimately weakening the system’s robustness and reducing estimation accuracy. To address these issues, we propose a mean square group metric (MSGM) that considers the number of measurement points, to construct the optimization objective for the LiDAR BA (RSO-BA) problem. In each optimization iteration, a robust kernel function reweights each metric to ensure robustness in the solution. Additionally, we derive the analytical Hessian matrix and gradient vector required for solving the RSO-BA, with the inclusion of second-order terms enhancing estimation accuracy. The proposed method can be directly applied to the mobile robot system for automatic driving, home service and unmanned security applications. Furthermore, the proposed algorithm can be extended to sensors such as depth cameras, which provide direct depth information, for broader practical applications.}, archive = {J_TASE}, author = {Tingchen Ma and Bingyi Xia and Yongsheng Ou and Jiankun Wang and Sheng Xu}, doi = {10.1109/TASE.2025.3541199}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12831--12844}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust second-order LiDAR bundle adjustment algorithm using mean squared group metric}, volume = {22}, year = {2025}, }
• (2025). An insulin-sensitivity-aware meal-bolus decision method based on event-triggered adaptive dynamic programming. TASE, 22, 12816--12830. (www)

  @article{ , abstract = {Postprandial glucose management is crucial for patients with diabetes. However, the currently adopted meal-bolus decision algorithms normally rely on subject-specific parameters and are impacted by time-varying insulin requirements caused by behavioral or metabolic variations. In this work, a data-driven meal-bolus decision method with an insulin-sensitivity-aware event-triggered learning mechanism is proposed. Specifically, the algorithm is developed within the framework of adaptive dynamic programming, which utilizes a Gaussian process regression approach to construct the model network and applies two shallow neural networks to construct the critic and action networks. To allow the algorithm to quickly perceive and respond to the variation of glycemic dynamics, different utility functions characterized by insulin-sensitivity-dependent penalty terms are constructed and switched based on the estimated insulin sensitivity conditions. The effectiveness of the proposed method is evaluated by in silico experiments utilizing the 10-adult cohort from the FDA-accepted UVa/Padova T1DM simulator. For the scenario of decreased insulin sensitivity, the proposed method leads to statistically significant improvements in terms of the percent time in euglycemia 70-180 mg/dL (86.2% vs. 72.4%, $p=0.025$ ) and the mean glucose (142.3 mg/dL vs. 155.3 mg/dL, $p=0.020$ ) without increasing the risk of hypoglycemia compared with the standard bolus calculator. Besides, an advisor-mode analysis using clinical data for subjects who initially exhibited significant hyperglycemia upon starting insulin therapy is conducted, which proves its effectiveness and safety in practical applications. Note to Practitioners—This work is motivated by the issue of postprandial blood glucose management, which is indispensable to diabetes patients. An effective meal-bolus decision method can help individuals resist the surge in blood glucose levels caused by carbohydrate ingestions and reduce the risk of hyperglycemia. However, for the state-of-the-art bolus decision algorithms, the inter/intra-subject variations of physiological parameters and the fluctuating insulin requirements caused by physical activity, stress, or emotional change are the two main causes of performance degeneration. To this end, we propose an insulin-sensitivity-aware event-triggered learning approach to meal bolus optimization. Specifically, without explicitly adjusting the subject-specific parameters, an adaptive dynamic programming decision framework is designed to perform the bolus decision based on the data-driven network learning and the iteration of control law and cost function. Besides, the estimated insulin sensitivity information is utilized in the learning phase of the decision framework to optimize the ultimate bolus, bolus dosage so as to make the algorithm quickly respond to the variation of insulin requirements. The in silico experiments and clinical data advisor-mode analysis demonstrate the effectiveness of the proposed algorithm and its feasibility in clinical application for postprandial blood glucose management.}, archive = {J_TASE}, author = {Xiang Lu and Deheng Cai and Wei Liu and Linong Ji and Dawei Shi}, doi = {10.1109/TASE.2025.3548101}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12816--12830}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An insulin-sensitivity-aware meal-bolus decision method based on event-triggered adaptive dynamic programming}, volume = {22}, year = {2025}, }
• (2025). Fully asynchronous neuromorphic perception for mobile robot dodging with loihi chips. TASE, 22, 12802--12815. (www)

  @article{ , abstract = {Sparse and asynchronous sensing and processing in natural organisms lead to ultra low-latency and energy-efficient perception. Inspired by these characteristics, event cameras have attracted extensive attention from academia and industry. However, the mainstream event stream processing paradigms (e.g., event frames, 3D voxels) encounter issues such as feature loss, event stacking, and high computational burden. These problems deviate from the intended purpose of event cameras. To address these issues, we propose a fully asynchronous neuromorphic paradigm that integrates event cameras, spiking networks, and neuromorphic processors (Intel Loihi). This paradigm can faithfully process each event asynchronously as it arrives, mimicking the spike-driven signal processing in biological brains. We first propose a Key-Event-Point module to extract the key event stream from the raw event stream, effectively addressing the issue of limited transmission bandwidth when processing events on neuromorphic processors. Then, we propose the complete pipeline for the dodging network based on spiking neural networks to achieve offline training and online inference. Finally, we compare the proposed paradigm with event frames and 3D voxels processing paradigms in detail on the real mobile robot dodging task. Experimental results show that our scheme exhibits better robustness than image-like methods with different time windows and light conditions. Additionally, the energy consumption per inference of our scheme on the embedded Loihi processor is only 4.30% of that of the event spike tensor method on NVIDIA Jetson Orin NX with energy-saving mode, and 1.64% of that of the event frame method on the same neuromorphic processor. To the best of our knowledge, this is the first time that a fully asynchronous neuromorphic paradigm has been implemented for solving sequential tasks on a real mobile robot.Note to Practitioners—As a neuromorphic visual sensor, the event camera offers a novel approach to achieving robust, low-power perception in robotics, owing to its high temporal resolution, wide dynamic range, and low information redundancy. However, the sparse and asynchronous nature of event streams presents a significant challenge for data processing. The mainstream approach preprocesses event streams into various representations (e.g., event frames, 3D voxels) before performing subsequent operations, leading to issues such as feature loss, event stacking, and high computational burden. In this paper, we propose a fully neuromorphic system that leverages the asynchronous characteristics of event streams and spiking neural networks to enable asynchronous processing of events. Our asynchronous processing of event streams enhances robustness across different time windows and lighting conditions while significantly reducing power consumption during inference. Our fully asynchronous neuromorphic system has been validated on a real mobile robot and is expected to advance the robot perception system towards biological perception.}, archive = {J_TASE}, author = {Junjie Jiang and Delei Kong and Chenming Hu and Zheng Fang}, doi = {10.1109/TASE.2025.3547338}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12802--12815}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully asynchronous neuromorphic perception for mobile robot dodging with loihi chips}, volume = {22}, year = {2025}, }
• (2025). Predefined-time secure distributed energy management for microgrids against DoS attack based on dynamic event-triggered approach. TASE, 22, 12791--12801. (www)

  @article{ , abstract = {This article studies how to solve the energy management (EM) problem while balancing economy and security. Firstly, a defense scheme is developed to relieve the negative effects of denial-of-service (DoS) attacks. Secondly, a secure predefined-time distributed EM algorithm via the time-base generator (TBG) is devised which can be suitable for directed graphs, and sensitive information is decomposed into two dynamic variables for key information privacy preservation. Thirdly, a dynamic event-triggered (DET) mechanism is proposed to economize communication resources. By constructing the Lyapunov energy function, it is proven that the convergence time of the devised distributed algorithm can be predefined by users independent of initial conditions. Zeno behavior is proven to be avoided, which exhibits the effectiveness of the DET approach. At last, the simulations are given to show the validity and advancement of the designed secure distributed algorithm.Note to Practitioners—With the deep integration of the cyber layer and physical layer in microgrids, the microgrid system is facing the threat of attacks. This article considers a typical cyber attack, a DoS attack, and the privacy-preserving problem. For solving the EM problem considering cyber security, a DoS attack defense scheme and a novel state decomposition mechanism have been proposed. After a DoS attack occurs, the microgrid system has an urgent need for EM fast recovery, while frequent information exchange can lead to unnecessary communication bandwidth occupation. Hence, a predefined-time distributed optimization approach based on the DET method is devised to achieve fast EM after DoS attacks and relieve the communication burden. The article adopts the TBG function to design dynamic gain which allows practitioners to flexibly adjust the convergence speed of the EM algorithm in practical applications.}, archive = {J_TASE}, author = {Feisheng Yang and Jiaming Liu and Yuhua Du and Chao-Bo Yan}, doi = {10.1109/TASE.2025.3544892}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12791--12801}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined-time secure distributed energy management for microgrids against DoS attack based on dynamic event-triggered approach}, volume = {22}, year = {2025}, }
• (2025). Volatile discrete-time observation stabilization of time-varying hybrid stochastic large-scale networks. TASE, 22, 12780--12790. (www)

  @article{ , abstract = {This paper focuses on the exponential stability of time-varying hybrid stochastic large-scale networks, in which a novel discrete-time observation control strategy with volatile control gain is imposed. The control gain exhibits sign-indefinite characteristics that follow certain volatile patterns. Then, the strategy can preferably respond to extrinsic perturbances and defend actuators and facilities by comparison with the position in constant gain. To quantify the volatile control gain, a concept of average volatile discrete-time observation control gain is proposed. In light of stochastic analysis theory and Lyapunov method, we establish improved stability criteria. In contrast to earlier results, the piecewise continuous, time-varying, and indefinite functions replace the constant coefficient of the upper bound estimation for the operator of the Lyapunov function, i.e., it can be positive or negative along time evolution, which shows wider applications of our results. Furthermore, considering the phenomenon of controller failure and packet losses, obtained results are applicable for analyzing the stabilization via intermittent discrete-time observation control and nonuniform discrete-time observation control, respectively. Finally, two applications to oscillator systems and single-link robot arms are presented and numerical simulations are exploited to illustrate the feasibility. Note to Practitioners—This paper is motivated by existing results on discrete-time observation control for the dynamics of hybrid stochastic large-scale networks. The existing results mainly require continuous updates to the control signals while maintaining a constant control gain. This is challenging to achieve on digital computers and often does not effectively respond to external disturbances. This paper designs a new hybrid control called volatile discrete-time observation control, and a concept of the control gain is proposed to quantify the volatile control gain. Specifically, the obtained results also apply to scenarios involving control failures and packet loss. The results are applied to oscillator systems and single-link robot arms, demonstrating their effectiveness and it is expected that the proposed approach can be extended to more practical physical engineering systems.}, archive = {J_TASE}, author = {Dongsheng Xu and Yuanning Zhang and Sen Li and Huan Su}, doi = {10.1109/TASE.2025.3544678}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12780--12790}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Volatile discrete-time observation stabilization of time-varying hybrid stochastic large-scale networks}, volume = {22}, year = {2025}, }
• (2025). Image-driven imitation learning: Acquiring expert scanning skills in robotics ultrasound. TASE, 22, 12768--12779. (www)

  @article{ , abstract = {A promising ultrasound (US) image acquisition requires experienced sonographers holding the probe with proper force and pose to ensure an excellent acoustic coupling. To enable a robotic ultrasound system (RUSS) to acquire the sonographers’ skills from ultrasound image demonstrations, this paper proposes a cutting-edge framework that integrates an expert technique discrimination network and a robotic strategy generation network to learn expert scanning skills. In this framework, the expert technique discrimination network focuses on learning expert scanning techniques from the pre- and post-frame ultrasound images. Furthermore, to acquire expert scanning skills and obtain a standard image view, we design a knowledge-based algorithm grounded on inverse reinforcement learning (IRL) to generate a series of scanning policies concerning the expert technique discrimination network. Both simulations and experiments are conducted to validate the effectiveness of the proposed framework by comparing it with MI-GPSR and PTR. The scanning success rate and trajectory tracking error of the algorithm in the simulation environment are 68% and 12.0782, respectively, while in the phantom environment are 94% and 11.8367. The results demonstrate good performance in the task of imitating expert techniques for autonomous scanning. Note to Practitioners—The motivation for this work originates from the need for ultrasound scanning tasks, such as carotid plaque and thyroid scans, to follow specific procedures. In clinical practice, sonographers require extensive learning and training to acquire these scanning skills. Traditional autonomous robotic ultrasound research often focuses on achieving the final standard view, neglecting the logical flow of the scanning process. Moreover, current studies on imitation learning for robotic ultrasound typically require not only ultrasound images from expert demonstrations but also additional data like probe position, adding complexity and potential errors to the data collection process. The proposed method addresses this by designing a framework that learns and comprehends expert techniques solely from image demonstrations. This endows RUSS with the ability to perform a human-like ultrasound scanning task. This work can be applied in the field of autonomous ultrasound robotics to assist sonographers in achieving more precise scans.}, archive = {J_TASE}, author = {Jiaming Li and Haohui Huang and Cong Guo and Qingguang Lin and Jing Guo and Chenguang Yang}, doi = {10.1109/TASE.2025.3545043}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12768--12779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Image-driven imitation learning: Acquiring expert scanning skills in robotics ultrasound}, volume = {22}, year = {2025}, }
• (2025). Triple-gated bidirectional variational pyramid network for multirate industrial soft sensing. TASE, 22, 12756--12767. (www)

  @article{ , abstract = {In various industrial processes, soft sensors have become important tools for predicting key quality variables. However, traditional soft sensor models only use data from the same sampling moments as the key quality variables, thus wasting information from other sampling moments. In light of this, a novel soft sensor model named triple-gated bidirectional variational pyramid network (G3-BiVPN) is proposed. Within G3-BiVPN, the multirate dataset is first segmented into multiple datasets each with a single sampling rate. For each dataset, a corresponding bidirectional variational autoencoder (BiVAE) is utilized for feature extraction. BiVAEs are used as backbone models to form a bidirectional pyramid structure. A triple gating mechanism consisting of attention gate (AG), temporal gate (TG), and spatial gate (SG) is integrated into BiVAEs to regulate information flow. Information can be flowed bidirectionally through different levels, with each level establishing a regression relationship with the key quality variables and selecting the optimal level as the final output. The core advantage of G3-BiVPN lies in its utilization of the multirate nature of the data. Finally, the efficiency of G3-BiVPN has been validated through two sets of real-world industrial process data with multiple sampling rates.Note to Practitioners—Due to sensor specification differences or practical needs, industrial process variables often exhibit a multiplicity in sampling rates. The core concept of G3-BiVPN is the bidirectional transmission and fusion of features at different rates. Initially, multirate features are extracted using BiVAEs and filtered through AGs to retain the relevant feature information for the output. Subsequently, these filtered features are stacked, with higher sampling rate features forming the bottom layers and lower sampling rate features occupying the top layers. In this architecture, feature information from different sampling rates can freely flow bidirectionally between the layers. During the downward transmission of information, upsampling is performed. To mitigate issues of information redundancy, further filtering is applied through TG after the upsampling. And, during the upward transmission, downsampling is employed. To prevent loss of detailed information during downsampling, SG is utilized before downsampling. This bidirectional propagation enhances the effective fusion of information across different layers, ensuring comprehensive understanding of the data at each level. Through this intricate feature interaction, G3-BiVPN can adapt to the complexity of multirate data, effectively utilizing available information from different sampling rates to accurately predict key quality variables.}, archive = {J_TASE}, author = {Yan-Lin He and Wei Zhang and Lei Chen and Yuan Xu and Qun-Xiong Zhu and Huihui Gao}, doi = {10.1109/TASE.2025.3548053}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12756--12767}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Triple-gated bidirectional variational pyramid network for multirate industrial soft sensing}, volume = {22}, year = {2025}, }
• (2025). NL-WCS: A novel data-driven algorithm for extracting the dynamics of serial robots considering non-linear friction. TASE, 22, 12741--12755. (www)

  @article{ , abstract = {Recently, developed data-driven SINDy-based techniques can identify the dynamic model of serial robots without simplifying assumptions nor pre-knowledge of all kinematics and geometric details. However, these techniques cannot handle non-linear friction models, which significantly affects the precision of the dynamic model identification. This study proposes a novel data-driven approach for dynamic model identification considering the non-linear friction model along with SINDy concept. This approach is termed as non-linear-weighted-constrained SINDy (NL-WCS). The SINDy concept is extended to accommodate any non-linear friction model by efficiently incorporating the Levenberg-Marquardt (LM) algorithm. Weighted L1 regularization is combined with the physics and the data constraints, to promote sparsity in the recovery of the dynamic equations. Moreover, this combination makes the approach robust against the regression matrix’s ill-conditionality and noise. LM is integrated with the robot’s SIMULINK model to get initial values for the non-linear friction empirical parameters. NL-WCS is experimentally evaluated by utilizing three distinct trajectories to validate the extracted dynamic model of 6-DOF UR10 and 7-DOF KUKA robots. In addition, five non-linear friction models are compared. NL-WCS outperforms all the previous SINDy-data-driven methods since it reduced the RMSE significantly up to 60.65%. NL-WCS also demonstrated robustness when tested against different levels of noise. Note to Practitioners—High-performance serial robot model-based controllers need precise dynamic model identification. The classical analytical methods for modeling serial robots rely on deriving the dynamic equations with simplified assumptions and then identifying the inertial parameters. Specifically, they assume that all the robot’s geometric details are known. Furthermore, there are uncertainties in geometric parameter values due to manufacturing/assembly errors. Data-driven-based SINDy methods can derive the dynamic model without pre-knowledge of the geometric parameters of the robot. However, it can’t incorporate non-linear friction models. Thus, this paper proposed a novel data-driven technique to extract the dynamic model of any serial manipulator by coupling SINDy approach with the nonlinear friction models which is the realistic case. The practitioners can benefit from a technique like that, in such a way of applying NL-WCS to different types of industrial robots, particularly, those that have missing manufacturer data sets or sheets. Simple and reliable dynamic models can be derived only by running the robot with various trajectories. Also, this helps eliminate any uncertainties in the dynamic model discovery/building process. Furthermore, Incorporating the non-linear friction model in the derivation process enhances the dynamic modeling accuracy as it effectively takes into account all friction characteristics. The proposed approach can be converted into a software package that the practitioners can use to derive the dynamics without getting their heads around the overwhelming robot’s kinematic details.}, archive = {J_TASE}, author = {Mohamed Omar and Ke Wang and Ruifeng Li and Dai Kun and Ossama B. Abouelatta and Tao Xie and Mohamed Alkalla}, doi = {10.1109/TASE.2025.3546920}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12741--12755}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {NL-WCS: A novel data-driven algorithm for extracting the dynamics of serial robots considering non-linear friction}, volume = {22}, year = {2025}, }
• (2025). Distributed output feedback prescribed performance control for high-order nonlinear multi-agent systems. TASE, 22, 12730--12740. (www)

  @article{ , abstract = {This paper presents an output feedback prescribed performance control method for a class of high-order nonlinear uncertain multiagent systems. General prescribed performance consensus control for multiagent systems requires that the initial consensus error is constraining within a boundary value. Unlike existing works, the conservative condition of prescribed performance consensus control for multiagent systems is relaxed as the consensus error is independent of the initial state. In this condition, we employ a control scheme with improved prescribed performance to constraint the transient behavior of the system. By designing reduce order dynamic gain K-filter, the state variables of the systems are reconstructed. Based on the baskstepping method and dynamic surface control technology, we introduce an innovative prescribed performance event-triggered approach aimed at ensuring prescribed performance levels for both transient and steady-state aspects of consensus control and reducing the communication bandwidth resource. Furthermore, we rigorously demonstrated through the Lyapunov function that all agents can achieve consensus with the leader driven by the controller. The simulation results have demonstrated the effective tracking performance of the developed control approach. Finally, the effectiveness and reliability of the proposed control strategy are verified through the successful execution of multi-QUAVs formation encirclement control.Note to Practitioners—This paper considers output feedback prescribed performance consensus control problem of multiagent systems, which can be applied to some practical systems, e.g., low-altitude formation flight of autonomous aerial vehicle with specified transient performance, multi underwater robot for oil field line-cruising, etc. Furthermore, in harsh working environments, the states of agent devices are difficult to measure or cannot be measured. Hence the control objective in these applications can be transformed into output feedback prescribed performance consensus control problem of multiagent systems. Besides, in military applications, high-speed aircraft need to fly at a certain height above the ground in order to avoid radar detection, which poses strong limitations on the position of the aircraft. This challenging issue can be well addressed through prescribed performance control. Compared with the existing results, this paper proposed a novel prescribed performance control strategy for multiagent systems with unrestricted initial state. It should be noted that in practical applications, communication bandwidth resources are extremely limited, so we propose event triggered control to effectively alleviate the communication pressure of multiagent system cooperative control. The nonlinear high-order system model studied in this article can be converted into a Lagrangian system, multi-complexity manipulator system and an autonomous aerial vehicle system, which has strong engineering practical significance.}, archive = {J_TASE}, author = {Zhijie Li and Xiaofei Wang and Hao Guo and Lele Xi and Guopin Liu and Yafeng Li and Changchun Hua}, doi = {10.1109/TASE.2025.3544306}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12730--12740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed output feedback prescribed performance control for high-order nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Dynamic observer-based H-infinity consensus control of fractional-order multi-agent systems. TASE, 22, 12720--12729. (www)

  @article{ , abstract = {This paper considers the H-infinity consensus control problem of fractional-order multi-agent systems (FOMASs) with external disturbances (order $0\lt \alpha \lt 1$ ) based on dynamic observer method. Using matrix spectral decomposition and the algebra properties of Kronecker product, the distributed H-infinity consensus problem in FOMASs is converted to an H-infinity problem for a group of independent fractional-order systems (FOSs) having the same dimension as that of a single agent. Based on the lemma of H-infinity stability for FOS in term of linear matrix inequalities (LMIs) and by introducing a new linearizing change of variables, the design conditions for the H-infinity consensus controller are derived. A multi-step procedure for constructing controller and dynamic observer is further presented. In the simulation part of the paper, an example of the PD $^{\alpha }$ -controlled Newcastle robot model is given to illustrate the effectiveness of the theoretical results. Note to Practitioners—Due to the unique memorability of fractional-order calculus, its application to H-infinity consensus control of multi-agent systems brings system performance metrics closer to reality. This paper investigates the H-infinity consensus problem of FOMASs based on dynamic observers under disturbance. In practical applications, it is hard to directly measure the states of some FOMASs. Therefore, to solve the H-infinity consensus control problem of such systems, we introduce concepts including multi-agents, dynamic observers, and fractional-order. The primary objective is to achieve equilibrium in such FOMASs as closely to reality as possible. Optimizing system performance indicators brings the system to its desired state. Finally, an example of the PD $^{\alpha }$ -controlled Newcastle robot model is provided to demonstrate the validity of the theorems.}, archive = {J_TASE}, author = {He Li and Shuo Liu and Guanglei Meng and Quanyong Fan}, doi = {10.1109/TASE.2025.3547980}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12720--12729}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic observer-based H-infinity consensus control of fractional-order multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Voltage regulation in microgrids via modulation communication: A distributed output consensus approach. TASE, 22, 12710--12719. (www)

  @article{ , abstract = {Lately, voltage regulations for distributed generation (DG) in direct current (DC) microgrid (MG) have been extensively investigated based on various communication methods. However, the communication is not always reliable or even nonexistent in actual remote area. In this paper, a distributed coordinated voltage control strategy based on power/information hybrid modulation without additional communication lines is proposed. First, the whole state-space model in DC MG under double carriers differential phase shift keying (DC-DPSK) modulation method is constructed. Then, distributed dynamic output feedback control strategy is designed according to established model. The designed distributed control can address leader-following consensus issue in linear heterogeneous time delay multi-agent system which is equal to solve output voltage consensus issue in DC MG. Finally, the effectiveness and potential of the proposed design strategy is verified by simulation results.Note to Practitioners—Output voltage is the vital index for evaluating the quality of power system. Although various accomplishments have been achieved in the area of voltage regulation, the study of voltage regulation control without extra communication lines is rarely considered in DC MG. To fill in the gap, the paper concentrates on designing output feedback consensus control method under DC-DPSK modulation communication to ensure the stability and output voltage consensus in power system. It is worth noting that proposed control strategy based on multi-agent algorithm can be implemented distributed performance without additional communication lines, which is practical in real application. The results aim to provide a helpful reference for secondary control design in the remote area power system, such that the corresponding application research can be promoted.}, archive = {J_TASE}, author = {Menglin Liu and Peng Shi and Imre J. Rudas}, doi = {10.1109/TASE.2025.3546899}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12710--12719}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Voltage regulation in microgrids via modulation communication: A distributed output consensus approach}, volume = {22}, year = {2025}, }
• (2025). Dynamic splitting and merging control strategy for vehicle platoon based on trust evaluation in a zero-trust environment. TASE, 22, 12693--12709. (www)

  @article{ , abstract = {Most of the past research on platooning control has not considered the impact of changes in the level of trust between vehicles in the platoon on the platoon control, and the communication process between connected vehicles is often subjected to malicious attacks such as time delays or interruptions, as well as tampering with status information, and the level of trust between vehicles changes, which affects the cooperative behaviors such as driving styles and inter-vehicle interval strategies. Changes in the level of trust between vehicles can affect the control process strategy of the vehicle, such as where the spacing strategy changes, thereby affecting the inputs to the controller, which in turn affects the outputs of the controller, and ultimately affects changes in the driving style of the vehicle, so considering the level of trust between vehicles in vehicle platoon control is critical to the safe operation of the vehicle platoon. To address this challenge, this paper proposes a dynamic splitting and merging control strategy for vehicle platoons based on the trust evaluation of vehicle nodes. Firstly, trust is evaluated using the Certainty Factor (C-F) uncertain reasoning model, which is a process that starts from initial evidence of uncertainty and derives reasonable conclusions with a certain degree of uncertainty by utilizing the uncertainty of evidence. An autoregressive model predicts short-term vehicle trajectories, and multi-source information from communication and perception is compared to determine vehicle node trust. Utilizing Bayesian reasoning methods, the trust level of vehicle nodes is updated. Based on traditional platoon control and trust evaluation, a software-level dynamic splitting and merging strategy is proposed to enhance resilience against unknown disturbances in a zero-trust environment. Finally, the system’s internal and string stability are analyzed, and the scheme’s effectiveness is validated through simulations. Note to Practitioners—The motivation of this paper is to solve the problem of vehicle platoon security control for connected vehicles in default distrust scenarios. Previous approaches were designed under the premise of considering mutual trust between vehicle nodes, ignoring the problems of inaccurate information interaction between connected vehicles and attacked interaction processes in real scenarios. To resist the risks of various types of attacks faced by intelligent networked vehicles in real traffic scenarios, this paper designs a vehicle platoon control strategy under zero-trust scenarios. Firstly, using the C-F uncertainty reasoning method, we propose a data-based node trust evaluation algorithm, and utilize its node trust evaluation results to re-establish the vehicle state equation under the zero-trust scenario. To make the vehicle platoon control under the zero-trust scenario scalable and resilient, this paper further proposes a vehicle platoon dynamic splitting and merging strategy based on the vehicle node trust evaluation scheme without changing the original communication topology of the vehicle platoon. The preliminary experimental results show that compared with the previous vehicle platoon control method under default vehicle node trust, the security of vehicle platoon operation is effectively improved. In the future, to better match real traffic scenarios, we will study how to dynamically adjust the trust threshold of vehicle nodes based on the vehicle operation environment, spacing strategy, sensor parameters, and other factors.}, archive = {J_TASE}, author = {Darong Huang and Ling Zhu and Zhenyuan Zhang and Yuhong Na and Zhongmei Li and Xin Fang}, doi = {10.1109/TASE.2025.3545555}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12693--12709}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic splitting and merging control strategy for vehicle platoon based on trust evaluation in a zero-trust environment}, volume = {22}, year = {2025}, }
• (2025). Enhancing RGB-D mirror segmentation with a neighborhood-matching and demand-modal adaptive network using knowledge distillation. TASE, 22, 12679--12692. (www)

  @article{ , abstract = {Recent breakthroughs in computer vision have led to remarkable progress in the areas of autonomous vehicles and robotics. However, ordinary objects such as mirrors pose unique challenges to computer vision systems owing to occlusion, reflection, and distortion. Moreover, existing deep learning models suffer from issues such as excessive parameters and high computational complexity, making it challenging to implement numerous studies offline. To address these issues, we propose an innovative solution: a neighborhood-matching and demand-modal adaptive network using knowledge distillation (KD), called NDANet-S $^{\ast }$ , specifically designed for red-green-blue depth mirror segmentation. NDANet-S $^{\ast }$ operates by iteratively matching detailed and semantic difference between neighborhood features during the encoding phase. It then complements information across different modalities through demand-modal adaptation, enhancing heteromodal cross-complementation during the KD stage. In the decoding phase, semantic enhancement features and iterative encoding features are deeply integrated, forming a strong foundation for multistage progressive knowledge transfer in the KD process. Furthermore, we introduce a multistage teacher-assisted KD scheme, guided by sample complexity, to work synergistically with the mirror segmentation model. This innovative scheme includes a sample complexity rater, heterogeneous cross-complementarity, and hierarchical progressive knowledge transfer. Experimental evaluations on publicly available datasets indicate that NDANet-S $^{\ast }$ significantly enhances segmentation accuracy while preserving a consistent number of parameters. Additionally, it achieves state-of-the-art performance in mirror segmentation. The source code for our model is publicly available and can be accessed at: https://github.com/2021nihao/NMDANet. Note to Practitioners—This study presents a neighborhood-matching and demand-modal adaptive network for RGB-D mirror segmentation, incorporating knowledge distillation (KD). The combination of the KD framework and segmentation network enables sample complexity discrimination, cross-modal distillation, and multilevel distillation (guided by the former) to achieve targeted distillation. The newly proposed NDANet-S $^{\ast }$ surpasses current state-of-the-art methods, with a reduction of 93.2% in parameters and 87.65% in floating-point operations (FLOPs) compared to NDANet-T.}, archive = {J_TASE}, author = {Wujie Zhou and Han Zhang and Yuanyuan Liu and Ting Luo}, doi = {10.1109/TASE.2025.3547613}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12679--12692}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing RGB-D mirror segmentation with a neighborhood-matching and demand-modal adaptive network using knowledge distillation}, volume = {22}, year = {2025}, }
• (2025). RL-USRegi: Autonomous ultrasound registration for radiation-free spinal surgical navigation using reinforcement learning. TASE, 22, 12668--12678. (www)

  @article{ , abstract = {Registration of intraoperative ultrasound (iUS) with preoperative CT represents a significant yet challenging task in the context of radiation-free spinal surgical navigation. The presence of thickness response artifacts in US images poses a considerable obstacle to the accurate extraction of bone boundaries. Furthermore, US-CT registration typically necessitates the detection and correspondence of high-quality landmarks at the initial stage. This can be accomplished by surgeons who have undergone extensive training in the localization of standard spinal US views, enabling them to identify key vertebral landmarks for subsequent precise registration. In this paper, we propose a fully automated iUS registration method that employs a limited number of spinal US views as observation objects. Specifically, three-dimensional vertebral meshes segmented from the preoperative CT images are superimposed on the US images and then fed to the reinforcement learning (RL) agent for sequential decision-making. The proposed method achieves fully automatic US-CT registration without relying on prespecified initialization. This is achieved by training the agent to approach bone surfaces on several randomly selected 2D US views. The instability of RL-based iUS registration is primarily attributable to the difficulty of correlating long-range information within the neural network. To address this issue, we propose a Field of View Separation (FoVS) module. The proposed approach employs separate encoders for US and mesh images, followed by cross-attention aggregation, which facilitates information flow between non-adjacent pixels. This approach enables pretraining of feature extraction on distinct encoders and the application of supplementary loss for enhanced feature matching precision, thereby significantly improving the learning capability and stability of the network. Furthermore, a refinement module is introduced to correct the results of the RL registration, which improves the stability of the registration process. To ascertain the efficacy of each module, action, and auxiliary task, comprehensive experiments are conducted. The results demonstrate that the performance of the RL agent is enhanced by the associated modules and auxiliary tasks. The registration exhibited an angular error of $8.83 \; \pm \; 4.69$ degrees and a translational error of $3.34 \; \pm \; 1.42$ mm, achieving the state-of-the-art (SOTA) results. It is noteworthy that fine-tuning the model prior to the surgical phase can significantly reduce the registration error, which is a promising outcome for its clinical translation.Note to Practitioners—The objective of this study is to address the issue of image registration using iUS in conjunction with preoperative CT scans in the context of spine surgery. The current 2D/3D image registration methods are constrained by several limitations. Firstly, they often exhibit reduced accuracy, and require high-quality images in substantial quantities. Secondly, there is a lack of effective mechanisms to rectify errors identified after the registration process. This paper proposes a fully automated registration framework based on RL, which incorporates image rendering and mesh clipping to enable continuous adjustment of the pose of 3D data, thereby facilitating 2D/3D registration. The framework employs the distinctive attributes of iUS images and incorporates a refinement module to evaluate registration accuracy, thereby facilitating the rectification of any registration issues. The proposed framework was tested on both sheep lumbar subjects and human lumbar phantoms, demonstrating the highest level of performance to date and indicating its potential for integration into surgical navigation systems.}, archive = {J_TASE}, author = {Ang Li and Jiayi Han and Yongjian Zhao and Max Q.-H. Meng and Li Liu}, doi = {10.1109/TASE.2025.3544413}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12668--12678}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RL-USRegi: Autonomous ultrasound registration for radiation-free spinal surgical navigation using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Tensor-based model reduction and identification for generalized memory polynomial. TASE, 22, 12655--12667. (www)

  @article{ , abstract = {Power amplifiers (PAs) are essential components in wireless communication systems, and the design of their behavioral models has been an important research topic for many years. The widely used generalized memory polynomial (GMP) model suffers from rapid growth in the number of parameters with increasing memory depths and nonlinearity order, which leads to a significant increase in model complexity and the risk of overfitting. In this study, we introduce tensor networks to compress the unknown coefficient tensor of the GMP model, resulting in three novel tensor-based GMP models. These models can achieve comparable performance to the GMP model, but with far fewer parameters and lower complexity. For the identification of these models, we derive the alternating least-squares (ALS) method to ensure the rapid updates and convergence of model parameters in an iterative manner. In addition, we notice that the horizontal slices of the third-order data tensor constructed from the input signals are Vandermonde matrices, which have a numerically low-rank structure. Hence, we further propose the RP-ALS algorithm, which first performs a truncated higher-order singular value decomposition on the data tensor to generate random projections, then conducts the ALS algorithm for the identification of projected models with downscaled dimensions, thus reducing the computational effort of the iterative process. The experimental results show that the proposed models outperform the full GMP model and sparse GMP model via LASSO regression in terms of the reduction in the number of parameters and running complexity. Note to Practitioners—This paper is motivated by the demand for low-complexity and accurate PA behavioral models, offering a viable option for modeling wideband PAs in modern wireless communication systems. A good model should provide satisfactory performance with as low complexity as possible and allow for quick identification of model parameters. Based on the advantages of tensor networks, this paper proposes three tensor-based GMP models that can utilize the potential low-rank tensor structure of the GMP model, and derives the alternating least-squares method to rapidly identify the model parameters. These models can efficiently reduce the number of parameters and running complexity while maintaining comparable accuracy to the GMP model.}, archive = {J_TASE}, author = {Yuchao Wang and Yimin Wei}, doi = {10.1109/TASE.2025.3546774}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12655--12667}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tensor-based model reduction and identification for generalized memory polynomial}, volume = {22}, year = {2025}, }
• (2025). Adaptive fixed-time prescribed performance control of non-smooth systems subject to Injection/Deception attacks. TASE, 22, 12645--12654. (www)

  @article{ , abstract = {This paper studies a fixed-time output tracking control problem of non-smooth cyber-physical systems (CPSs) subject to injection/deception attacks and performance constraints. Cellina approximate selection theorem is employed to convert the non-smooth system into an equivalent smooth one, based on which we introduce a mapping transformation of the output tracking error to solve the performance constraints. Then an adaptive prescribed performance controller that involves a fuzzy logic system (FLS) to restrain the effect of unknown attacks is derived in the context of fixed-time control theory. By Lyapunov stability analysis, we prove that the proposed method is able to assure, in addition to the fixed-time stabilization of the closed-loop system, the evolvement of output tracking error along within the preset boundary regardless of injection/deception attacks. The effectiveness of the result, together with the superiority over the existing methods, is demonstrated by two simulation examples. Note to Practitioners—Numerous plants like connected vehicle and manipulator system actually belong to a class of CPSs with non-smooth dynamics, and thus are vulnerable to cyber attacks due to the use of network communication. Under the circumstance, the existing methods cannot enforce the control performance as required. Therefore, by introducing the mapping transformation based on measured error, we propose a practical prescribed performance control method, and this method is then improved by involving FLS to address the nonlinearity arising from attacks such that it is capable of achieving the pre-designated transient and steady-state performances despite the presence of either injection or deception attacks. Such merits make the proposed scheme widely suitable for the output tracking tasks that desire both control safety and performance guarantees, e.g., positioning-grabbing of the linked manipulator, collision-free-path tracking of the connected vehicle.}, archive = {J_TASE}, author = {Ge Guo and Qian Xu and Chen-Liang Zhang}, doi = {10.1109/TASE.2025.3544459}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12645--12654}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fixed-time prescribed performance control of non-smooth systems subject to Injection/Deception attacks}, volume = {22}, year = {2025}, }
• (2025). Integrated dual torque sensors and DOB of robust torque control for flexible joint. TASE, 22, 12633--12644. (www)

  @article{ , abstract = {Flexible joints are the integral drive and control components for manipulators in interaction applications. Ensuring optimal performance while implementing torque control is crucial to maintaining stability during interactions. However, the system is susceptible to numerous disturbances, such as motor inaccuracies, nonlinearity caused by friction, and hysteresis. These disturbances are distributed across the motor-side, reduction gear, and load-side, leading to limitations in system performance, including response speed and steady-state accuracy. This paper proposes a disturbance compensation method combined with dual torque sensors and an improved disturbance observer (DOB). Two torque sensors are installed on the fixed side and load-side, respectively. These sensors are utilized to measure the load-side disturbances without considering accuracy models. Furthermore, an improved DOB is designed based on one of the torque sensors to estimate the motor-side disturbances. The disturbance compensator and feedback controller are introduced as the control algorithms for the flexible joint to achieve high-precision torque control. The robust stability of the proposed method is analyzed. Finally, several comparative experiments are conducted under various conditions. The results demonstrate that the proposed method enhances torque control performances and backdrivability compared with the traditional methods. Note to Practitioners—Generally, lots of nonlinear disturbances in the flexible joint system are distributed across low-speed/high-torque and high-speed/low-torque ports. These disturbances will lead to limitations in torque control performance. To address the problem above, this paper focuses on integrating dual torque sensors with an improved DOB. Specifically, two torque sensors are installed on the fixed side of the reduction gear and load-side, respectively. This enables the real-time measurement and calculation of load-side disturbances. On the other hand, considering the torque sensor mounted at the fixed side and motor-side model, an improved DOB is designed to estimate the motor-side disturbances. Subsequently, the closed-loop control architecture is designed based on the torque sensor attached to the load. The complex modeling process can be reduced by installing and applying torque sensors. The model and parameter errors are not considered, and the torque control performance is improved. The research outcome of this paper provides an effective approach that can be used to achieve stable and rapid torque tracking, as well as effortless human-robot interaction. The proposed method still has good torque tracking performance in the case of a large load.}, archive = {J_TASE}, author = {Junjie Dai and Chin-Yin Chen and Guilin Yang and Chi Zhang and Yanbiao Li}, doi = {10.1109/TASE.2025.3546220}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12633--12644}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated dual torque sensors and DOB of robust torque control for flexible joint}, volume = {22}, year = {2025}, }
• (2025). Cascade disturbance observer-based control design for cascaded systems with considerable inner-loop dynamics. TASE, 22, 12621--12632. (www)

  @article{ , abstract = {Cascade control permeates a wide range of industrial processes, due to its significant improvement in disturbance rejection. However, conventional cascade control turns laborious to control the sluggish outer loop and suffers from complex inner-outer-loop coupling dynamics. Combining the merits of conventional cascade control and powerful capability of disturbance observer (DOB), this paper presents a cascade disturbance observer (CDOB) based control approach for cascaded systems. The resulting cascade control system features equivalent decoupling of the inner- and outer-loop disturbance rejection (DR), which allows the relevant filters to be designed separately. For the inner-loop filter, a robust design approach, directly based on the performance specifications and robustness constraints, is proposed, which enables an intuitive trade-off between performance and robustness. A novel filter is devised to improve the sluggish outer-loop DR performance. With the feedforward controller designed based on the nominal model, a composite control strategy is finally constructed. Some simulations and experiments are presented to validate the efficacy of the proposed method, and the significant performance improvements depict application prospects in process control.Note to Practitioners—This paper focuses on cascade control design for cascaded systems and presents a CDOB based control scheme as well as its robust design approach. The DR response of the control system and detailed control design, including inner- & outer-loop filters design, compensated plant analysis and robust stability analysis, are demonstrated, which can provide a design reference for practitioners. Specifically, the inner- and outer-loop filters are constructed subject to the performance and robustness trade-offs, which can help practitioners grasp the performance and robustness of the control system and facilitate control practice. The proposed control method features equivalent decoupling of the inner- and outer-loop DR and the CDOB can also serve as a patch for the baseline controller just like the conventional DOB. Further, the CDOB structure is also promising to be extended to more general cascade systems, such as non-minimum phase processes. The simulation and experiment results demonstrate the superiority of the proposed method and show promise for practical application in other systems.}, archive = {J_TASE}, author = {Zhuo Chen and Yong-Sheng Hao and Zhi-Gang Su and Li Sun}, doi = {10.1109/TASE.2025.3544563}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12621--12632}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cascade disturbance observer-based control design for cascaded systems with considerable inner-loop dynamics}, volume = {22}, year = {2025}, }
• (2025). Online monitoring of high-dimensional data streams with deep Q-network. TASE, 22, 12606--12620. (www)

  @article{ , abstract = {With the fast advancements in Internet of Things (IoT) technology and sensing infrastructure, a wide range of systems continues to generate a massive amount of data. Meanwhile, practical resource constraints such as limited bandwidth or processing capability restrict the full observability of data streams in real time. As a result, the practitioners often need to dynamically decide which data streams to observe given the resource constraints in order to quickly detect any system anomaly as soon as possible. In this article, we propose a reinforcement learning framework based on deep Q-learning and combine it with statistical process control (SPC) techniques to effectively monitor high-dimensional data streams when only partial observations are available at each acquisition time due to resource constraints. To the best of our knowledge, this is the first work that integrates deep reinforcement learning, which considers long-term rewards associated with the dynamic data sampling strategy, into SPC charts; this integration effectively addresses the challenge of resource constraints in monitoring high-dimensional data streams. Specifically, we construct a nonparametric monitoring statistic for each data stream and develop a reinforcement learning framework to automatically identify the most informative data streams for observation at each time epoch. The state space, action space, and rewards in the reinforcement learning framework are carefully designed and a Double Dueling Q-network is trained accordingly. Unlike existing methods, which rely on heuristic approaches to determine the sampling strategy, the proposed framework maximizes the long-term reward, thus leading to superior performance compared to existing benchmarks. Numerical simulations and a case study are thoroughly conducted, showing that the proposed method outperforms the state-of-the-art algorithms by significantly reducing detection delay. Note to Practitioners—This paper is motivated by the practical issue of online process monitoring and anomaly detection with resource constraints. In particular, due to resource constraints, practitioners can only select a subset of data streams to monitor at each time epoch. Thus, the central challenges are to dynamically choose which data streams to observe and to decide when to raise an alarm. Unlike the existing methodologies which are heuristic and only consider short-term rewards from the dynamic sampling, this paper proposes a novel reinforcement learning framework that allows practitioners to monitor high-dimensional heterogeneous data streams more efficiently by considering the long-term rewards associated with the dynamic data sampling strategy. Four main steps are involved in the proposed method: (i) construct a nonparametric local monitoring statistic for each data stream; (ii) train the Double Dueling Q-network offline according to the proposed deep reinforcement learning framework; (iii) at each time epoch of the online monitoring, determine the most informative data streams to observe based on the deep Q-network; and (iv) determine whether to raise an alarm in the system.}, archive = {J_TASE}, author = {Haoqian Li and Ziqian Zheng and Kaibo Liu}, doi = {10.1109/TASE.2025.3547250}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12606--12620}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online monitoring of high-dimensional data streams with deep Q-network}, volume = {22}, year = {2025}, }
• (2025). Resilient tracking control of cyber-physical systems against false data injection attacks and obstacle avoidance. TASE, 22, 12596--12605. (www)

  @article{ , abstract = {In this paper, the reliable tracking control and collision avoidance problems for cyber-physical systems (CPSs) with false data injection (FDI) attacks are investigated. FDI attacks can significantly compromise the safety and performance of CPSs by corrupting control and navigation data. Safety is an important aspect of CPSs. Unmanned ground vehicles and aerial vehicles are important applications of CPSs. The dual challenge of maintaining system safety and stability under deliberate cyberattacks, while ensuring reliable obstacle avoidance in dynamic environments, remains unresolved in many current methodologies. These challenges are amplified in CPSs owing to their reliance on real-time data and their susceptibility to adversarial manipulation. The main objective of this study is to develop a resilient tracking control strategy that can effectively mitigate the impact of FDI attacks and achieve obstacle avoidance. We propose a novel framework based on the exponential control barrier function (ECBF) and a novel observer-based auxiliary signal approach that can ensure the resilience of CPSs against FDI attacks and obstacle avoidance. The effectiveness of our methods is shown through extensive simulations and physical experiments, which depict improved tracking accuracy and system stability in the presence of FDI attacks as compared with those of traditional control methods.Note to Practitioners—This research tackles the problem of maintaining dependable tracking control and avoiding collisions in CPSs vulnerable to FDI attacks, with applications in fields such as smart transportation and smart city. To mitigate the impact of FDI attacks, we propose a resilient control strategy that uses exponential ECBF and an observer-based auxiliary signal approach. This framework mitigates the impact of corrupted data, maintains system stability, and avoids collisions, even in the presence of cyberattacks. Practitioners in CPS design and deployment can benefit from this approach by integrating it into existing systems to improve safety and security. The proposed method is validated through both simulations and physical experiments, demonstrating its practicality for real-world applications.}, archive = {J_TASE}, author = {Daotong Zhang and Peng Shi and Chee Peng Lim and Imre Rudas}, doi = {10.1109/TASE.2025.3546583}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12596--12605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient tracking control of cyber-physical systems against false data injection attacks and obstacle avoidance}, volume = {22}, year = {2025}, }
• (2025). Backstepping control with tunable transformation parameters for a coupled PDE rijke tube with boundary input delay. TASE, 22, 12583--12595. (www)

  @article{ , abstract = {In this paper, the active control problem for thermoacoustic instability in the Rijke tube is investigated. First of all, the first-order hyperbolic partial differential equations (PDEs) for one-dimensional compressible gas dynamics, combined with the Heckl heat release model, are proposed to describe the unstable thermoacoustic system in the Rijke tube. For the dynamic system, the decoupling of the nonlinear heat release with delayed velocity fluctuation and the acoustic field, along with the boundary input delay, is addressed. Secondly, the transport PDE is used to represent the boundary input delay. This transforms the unstable thermoacoustic system into a cascaded form of the transport equation and linear hyperbolic PDEs, resulting in a system of $5\times 5$ first-order linear hyperbolic PDEs. Then, due to the limitations of the existing backstepping control method for this system, a novel backstepping control approach with tunable parameters is proposed to stabilize the system. Next, a boundary feedback controller with optimized transformation parameters is designed by solving a programming problem to exponentially stabilize the unstable thermoacoustic system with boundary input delay in the Rijke tube. Finally, the effectiveness of the proposed approach is verified through a numerical simulation using a Rijke tube. Note to Practitioners—This paper was motivated by the problem of thermoacoustic instability taking place frequently in combustion systems, from aero-engines to land-based gas turbines. The Rijke tube, as a classical and time-honored experimental platform in simulating thermoacoustic processes, is studied in this paper. Based on a first-order hyperbolic partial differential equation (PDE) model for one-dimensional compressible gas dynamics, combined with the Heckl heat release model, we present a novel stabilization approach to the thermoacoustic instability in the Rijke tube. The objective of this paper is to furnish a foundational framework that provides the control strategies for thermoacoustic instability in aero-engines combustion environments. A critical gap in the existing literature concerns the stabilization of thermoacoustic instability in the presence of boundary input delays, an inherent feature of practical applications that significantly affects stability. This paper processes the coupled PDE model through Riemann transformation and folding transformation. Then a new backstepping boundary feedback control with tunable parameters is proposed to stabilize the coupled unstable thermoacoustic system, which allows for the fastest convergence rate of the thermoacoustic system by adjusting the tunable parameters. We also prove that the existing method is ineffective to solve the coupled PDE thermoacoustic system with input delays. The method presented in this paper can be also applicable to a class of linear hyperbolic systems with opposite convection. In future work, we will address the stabilization of thermoacoustic systems with dual time delays, i.e., input delay and thermal inertia.}, archive = {J_TASE}, author = {Dan Ma and Yuqian Xing and Yanyan Liu and Hongjun Lin}, doi = {10.1109/TASE.2025.3546344}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12583--12595}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Backstepping control with tunable transformation parameters for a coupled PDE rijke tube with boundary input delay}, volume = {22}, year = {2025}, }
• (2025). Control synthesis of singular PDT switched systems with state constraints and application to HyTAQs. TASE, 22, 12572--12582. (www)

  @article{ , abstract = {This paper investigates the problem of control synthesis for a class of continuous-time singular switched systems, based on which a modeling and control framework is developed for Hybrid Terrestrial and Aerial Quadrotors (HyTAQs). To accurately describe the frequent switching phenomenon for HyTAQs during the landing process, persistent dwell time (PDT) switching signals are taken into consideration. Besides, mode-dependent state constraints are introduced to singular switched systems, which are more practical and general than the ones considered in earlier studies. Numerically testable stability criteria are obtained, and thereby the conditions on the existence of controllers capable of satisfying the mode-dependent state constraints are derived. Finally, two illustrative examples including HyTAQs are presented to demonstrate the necessity and validity of the designed controllers. Note to Practitioners—This paper is motivated by the modeling and stabilization of Hybrid Terrestrial and Aerial Quadrotors (HyTAQs), a class of vehicles with both terrestrial and aerial mobility, which have received extensive attention due to the low energy requirement and high maneuverability at the same takeoff weight. Note that the dynamic differences between aerial and terrestrial modes, exacerbate the challenges for performance analysis and controller design, let alone the differences in state dimensions. Therefore, the paper proposes a unified framework for modeling HyTAQs based on singular switched systems, and the mode-dependent state constraints are considered to reduce conservatism. In addition, the frequent switching phenomenon during the landing process has never been considered in the field. To this end, the PDT switching signals are introduced to describe the bounce and shake behaviors during the mode switching process. Numerically testable stability criteria are obtained, based on which the stabilization conditions for such systems are provided. Finally, a numerical example and an application for HyTAQs are presented to demonstrate the validity and applicability of the theoretical results.}, archive = {J_TASE}, author = {Lixian Zhang and Yuting Ma and Jianan Yang and Tong Wu and Yimin Zhu}, doi = {10.1109/TASE.2025.3545069}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12572--12582}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Control synthesis of singular PDT switched systems with state constraints and application to HyTAQs}, volume = {22}, year = {2025}, }
• (2025). Boosting global-local feature matching via anomaly synthesis for multi-class point cloud anomaly detection. TASE, 22, 12560--12571. (www)

  @article{ , abstract = {Point cloud anomaly detection is essential for various industrial applications. The huge computation and storage costs caused by the increasing product classes limit the application of single-class unsupervised methods, necessitating the development of multi-class unsupervised methods. However, the feature similarity between normal and anomalous points from different class data leads to the feature confusion problem, which greatly hinders the performance of multi-class methods. Therefore, we introduce a multi-class point cloud anomaly detection method, named GLFM, leveraging global-local feature matching to progressively separate data that are prone to confusion across multiple classes. Specifically, GLFM is structured into three stages: Stage-I proposes an anomaly synthesis pipeline that stretches point clouds to create abundant anomaly data that are utilized to adapt the point cloud feature extractor for better feature representation. Stage-II establishes the global and local memory banks according to the global and local feature distributions of all the training data, weakening the impact of feature confusion on the establishment of the memory bank. Stage-III implements anomaly detection of test data leveraging its feature distance from global and local memory banks. Extensive experiments on the MVTec 3D-AD, Real3D-AD and actual industry parts dataset showcase our proposed GLFM’s superior point cloud anomaly detection performance.Note to Practitioners—The proposed GLFM is employed for point cloud anomaly detection in industrial inspection, capable of simultaneously processing data across multiple classes. GLFM requires the collection of a set of normal product samples for model training, where the features of these samples are stored. If the feature distribution of a test sample deviates substantially from that of the normal samples, it is flagged as anomalous. GLFM not only exhibits outstanding performance on public datasets but has also been validated on a real-world industrial parts point cloud dataset.}, archive = {J_TASE}, author = {Yuqi Cheng and Yunkang Cao and Dongfang Wang and Weiming Shen and Wenlong Li}, doi = {10.1109/TASE.2025.3544462}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12560--12571}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Boosting global-local feature matching via anomaly synthesis for multi-class point cloud anomaly detection}, volume = {22}, year = {2025}, }
• (2025). Distributed estimation and motion control in multi-agent systems under multiple attacks. TASE, 22, 12548--12559. (www)

  @article{ , abstract = {This paper addresses the problem of distributed estimation and motion control (DEMC) in multi-agent systems (MASs) with both linear and Lipschitz nonlinear dynamics. Unlike conventional DEMC methods designed for MASs under ideal conditions, this work investigates scenarios where all agents are vulnerable to various forms of attacks. The considered attacks comprise false-data injection (FDI) attacks and denial of service (DoS) attacks that affect the communication channels among agents to destabilize the MAS. Also, the unbounded actuator attacks which exist in practical environments to intentionally degrade the MAS performance is considered. To cope with these kinds of attacks, two novel resilient approaches are established aimed at estimating and following a mobile target under attacks. The proposed distributed attack-resilient control strategies are designed based on a dual-layer structure, guaranteeing effective DEMC with an ultimately bounded error. The results from two simulation examples are provided to validate the presented algorithms. Note to Practitioners—The motivation of this work is to deal with the DEMC problem for MASs under multiple attacks. In most of the existing DEMC schemes for MASs, having a healthy network and dynamics is a requirement. However, in practical environments, MASs as an important subclass of cyber-physical systems are subject to different types of attacks that affect the network and dynamics of MASs and may seriously jeopardize the performance of the DEMC algorithm, or even worse, lead to instability. Therefore, a resilient hierarchical DEMC algorithm is proposed for MASs which allows agents to estimate and follow a mobile target under multiple attacks. The proposed scheme is resilient to most existing cyber-attacks and is designed for MASs with both linear and nonlinear dynamics. It can be applied to various practical engineering systems such as autonomous vehicles, mobile robots, and intelligent transportation systems. The stability and convergence of the proposed algorithms are analyzed mathematically, and it is shown that the agents not only track the estimated target but also can cope with multiple attacks through simulation experiments.}, archive = {J_TASE}, author = {Ahmadreza Jenabzadeh and Zhan Shu and Tingwen Huang and Quanmin Zhu and Yilun Shang and Yukang Cui}, doi = {10.1109/TASE.2025.3545462}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12548--12559}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed estimation and motion control in multi-agent systems under multiple attacks}, volume = {22}, year = {2025}, }
• (2025). Automatic point cloud clustering for surface defect diagnosis. TASE, 22, 12538--12547. (www)

  @article{ , abstract = {Point cloud clustering is a promising method for 3D surface defect diagnosis in manufacturing but requires manual clustering parameter selection, reducing usability. This paper proposes an automatic point cloud clustering method to address this issue. It employs a strategy that progresses from coarse to fine. In the coarse searching stage, a K-Nearest Neighbor (KNN) graph analysis technique is developed to recognize potential defective regions in parallel. Moving on to the fine stage of extracting detailed defects, a modified DBSCAN algorithm is proposed, in which the clustering parameters are calculated automatically from the KNN graph analysis results. Experimental results showed that the proposed method achieved cloud clustering with automatically calculated clustering parameters for surface defect diagnosis. The proposed method outperformed the traditional region growing algorithm in accuracy (0.942 vs. 0.680) and processing speed (21500 points/sec vs. 8740 points/sec) without requiring manual intervention.Note to Practitioners—This paper presents a method for diagnosing defects on automobile and flat steel surfaces. Current 3D point cloud techniques for surface defect diagnosis require manual parameter adjustments, reducing usability. This paper proposes an automatic method without manual intervention. The proposed method uses a coarse-to-fine strategy. The 3D point cloud is divided into sub-blocks to locate potential defects, and a clustering algorithm then extracts detailed defects with automatically determined parameters. We mathematically characterize changes in point density caused by surface defects and show how these features can be used for clustering parameter calculation. Experimental results demonstrate the method’s efficiency on flat as well as some curved surfaces, but it has yet to be evaluated on complex structures. Future work will aim to broaden its application to include a more extensive variety of surfaces and integrate it with robotic vision systems.}, archive = {J_TASE}, author = {Jidong Ye and Xingjian Liu and Harikrishnan Madhusudanan and Yue Wang and Junhui Zhu and Yong Wang and Changhai Ru and Xinyu Liu and Yu Sun}, doi = {10.1109/TASE.2025.3542076}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12538--12547}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic point cloud clustering for surface defect diagnosis}, volume = {22}, year = {2025}, }
• (2025). Design and validation of a vision-integrated multi-hinge exosuit for dorsiflexion assistance: A feasibility study with healthy individuals. TASE, 22, 12526--12537. (www)

  @article{ , abstract = {Exosuits play an essential role in facilitating rehabilitation training for patients with functional impairments. However, the safety of human-exosuit interactions and their adaptability to varied environments are crucial challenges that hinder their transition from laboratory prototypes to practical real-world applications. This study presents an innovative active exosuit designed to correct foot drop, characterized by its safe interactive performance and terrain daptability. Regarding safe human-exosuit interaction, meticulous design of the dimensions of hinge module ensures that, upon contact between the hinges, the driving rope reaches its maximum contraction length, thereby mechanically safeguarding against excessive dorsiflexion of the ankle joint by preventing further stretch. In terms of adaptability to various terrains, a visual system capable of terrain recognition has been integrated, enabling a seamless transition between control modes based on the detected terrain, achieving a recognition accuracy of 99.1% through the Transformer in Transformer (TNT) algorithm. To quantify the effect of the exosuit in correcting drop-foot, five healthy participants with artificially induced impairment were recruited to participate in the experiment. The experimental results revealed that the peak of the plantarflexion angle was decreased by 21.3%, 26.0%, and 46.5%, respectively, with the assistance of an exosuit when walking at the terrains of level ground (LG), upramps (UR), and upstairs (US). The root mean square (RMS) of the electromyography (EMG) signals of the tibialis anterior (TA) muscle was reduced by 26.9%, 32.5%, and 20.2%, respectively. These initial results demonstrate that the exosuit can effectively assist dorsiflexion at different terrains. Note to Practitioners—This paper has presented the design and evaluation of a soft exosuit equipped with multiple hinges and a vision system, which offers dorsiflexion assistance to improve ground clearance across various terrains. The multi-hinge structure facilitates effective assistance force transmission and acts as a mechanical limiter, preventing damage to the ankle joint from overstretching the Bowden cable due to control errors. The vision subsystem enables precise terrain recognition, facilitating tailored assistance based on the current terrain conditions. Moreover, a position-based control strategy is adopted based on the reference trajectory measured from the healthy side, enabling the exosuit to adapt to three terrains: LG, UR, and US. This work has the potential to advance the development of exosuits for practical applications significantly.}, archive = {J_TASE}, author = {Zhuo Wang and Chunjie Chen and Fangliang Yang and Hui Chen and Sida Du and Yao Liu and Xiangyang Wang and Xinyu Wu}, doi = {10.1109/TASE.2025.3545899}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12526--12537}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and validation of a vision-integrated multi-hinge exosuit for dorsiflexion assistance: A feasibility study with healthy individuals}, volume = {22}, year = {2025}, }
• (2025). Multi-vehicle collaborative trajectory planning based on kaldor-hicks improvement. TASE, 22, 12512--12525. (www)

  @article{ , abstract = {This paper employs lateral and longitudinal trajectory planning to generate candidate trajectories and discards those that do not satisfy the constraints imposed by single-vehicle conditions. Next, a collaborative trajectory combination set for multiple vehicles is derived from the candidate trajectories, with multi-vehicle constraints applied to eliminate combinations that fail to meet the required conditions. The objective function for each candidate trajectory set is first calculated using a single-vehicle objective function, after which a multi-vehicle objective function based on the Kaldor-Hicks improvement principle is constructed. Finally, the paper introduces an improved particle swarm optimization method for multi-vehicle collaborative trajectory planning. The results demonstrate that the dynamic spatiotemporal occupancy growth rate, under varying planning times and frequencies, is at least 19%. Furthermore, the proposed algorithm ensures efficient allocation of travel resources, preventing competition among vehicles that could compromise system feasibility. When verified with HighD trajectory data, the algorithm not only delivers superior optimization results but also exhibits lower standard deviations in dynamic spatiotemporal occupancy and speed compared to real-world data. Finally, the algorithm’s superiority in real-time decision-making and stability is confirmed. Note to Practitioners—In the context of mixed traffic comprising both autonomous and human-driven vehicles, this paper tackles the challenge of coordinating autonomous vehicles to improve traffic efficiency and safety in real-time environments. It presents a promising approach to enhancing cooperative behavior in complex scenarios by integrating real-time data streams to optimize adaptability and ensure equitable driving efficiency across different vehicle types.}, archive = {J_TASE}, author = {Donglei Rong and Wenbin Yao and Chengcheng Yang and Congcong Bai and Sheng Jin}, doi = {10.1109/TASE.2025.3542840}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12512--12525}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-vehicle collaborative trajectory planning based on kaldor-hicks improvement}, volume = {22}, year = {2025}, }
• (2025). Rectifying systematical measurement error through majority pattern mining and semi-supervised prediction for industrial instrument. TASE, 22, 12500--12511. (www)

  @article{ , abstract = {Systematical measurement errors in industrial sensors should be avoided or rectified, as severe discrepancies in data can impede control, operation, and evaluation. However, some systematical errors are hard to estimate or experiment with. These errors often arise due to the change in external conditions. This means the measurement errors can be ignored when the external conditions remain within the designed scope. Generally, the relationship between the external condition and the systematical error is of complex nonlinearity, which may not be analytically tractable so rectifying the error is challenging. Nonetheless, a general assumption can be made in typical industrial scenarios: most external conditions fall within the design scope so that the corresponding systematical errors are 0. This assumption holds because the sensor is generally installed and calibrated under the most common operating mode. Based on these rationales, we first propose an intermediate sample-enhanced clustering strategy to identify the majority pattern, aiding in figuring out the zero systematical measurement error points. Then, leveraging the zero systematical measurement error information and partly known labels, a semi-supervised learning method is employed for estimating the complex nonlinear mapping from the external condition and the measurement, thereby rectifying the errors. The effectiveness of our approach is demonstrated through the rectification of a density meter in a real industrial aluminum oxide process, validated by the comparison with the laboratory assay outcomes. Note to Practitioners—Measurement instruments in industrial systems are designed to meet the precision requirement under specific conditions. Once the condition x is out of scope, the measurement y will be accompanied by the systematical measurement errors which is the function of x, denoted by $f(x)$ . Thus the measurement model is $ y=y_{t}+f(x)+e$ , where $y_{t}$ is the true value and e is the random measurement error. However, for many scenarios, the errors are complex. $f(x)$ is nonlinear and even intractable. From the perspective of engineering practice, the condition x in most of the running period should remain within the design scope to ensure normal use. However, it is also common for the running status to drift from the original designed working points over time. Then, an intolerant systematical measurement error occurs, causing the instrument which can be expensive lose their function. To estimate the systematical measurement error and implement the rectification of the deviated measurement, this paper proposes a data-driven strategy when the condition x is measurable and part of the label for y is available. We demonstrated the effectiveness of the strategy using a real industrial application example.}, archive = {J_TASE}, author = {Saige Cheng and Yonggang Li and Kai Wang and Chunhua Yang}, doi = {10.1109/TASE.2025.3544863}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12500--12511}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rectifying systematical measurement error through majority pattern mining and semi-supervised prediction for industrial instrument}, volume = {22}, year = {2025}, }
• (2025). A maximally permissive robustness analysis for automated manufacturing systems allowing multiple server failures. TASE, 22, 12485--12499. (www)

  @article{ , abstract = {As one of the essential issues when designing optimal control policies for automated manufacturing systems (AMSs) with unreliable resources, maximally permissive behavior should be ensured for the controlled system. This paper is devoted to the maximally permissive robustness analysis for AMSs allowing multiple server failures. First, all subnet systems are categorized by their usage of unreliable resources so as to provide more general definitions of the robust, strongly robust, and weakly robust markings. Second, we provide algorithms to compute reduced reachability graphs (R2Gs) and strongly connected components (SC2s), respectively. Then, the necessary and sufficient conditions to verify the robustness of markings are obtained by analyzing the transitions involved in each component in an SC2 and the relation between markings in the disconnected marking subset and SC2. Finally, the effectiveness and efficiency of the theoretical results are validated via examples. Note to Practitioners—Maximally permissive robustness analysis is an essential criterion for evaluating the control performance of robustness for AMSs with failure-prone resources. Ineffective handling of resource failures can lead to unexpected disruptions, decreased throughput, and diminished product quality. While much research has focused on developing robust control policies to ensure system resilience, there remain certain states that do not comply with these policies, even when they are designed to be robust. We present significant findings that assess the robustness of all system states when unreliable resources experience various failure scenarios. These results allow the researchers to design optimal robust controllers, thereby significantly enhancing the flexibility and efficiency of manufacturing processes. By applying the results of our robustness analysis, manufacturers can better prepare for and respond to resource failures, ultimately improving overall production performance.}, archive = {J_TASE}, author = {Junyan Li and Hesuan Hu}, doi = {10.1109/TASE.2025.3544585}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12485--12499}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A maximally permissive robustness analysis for automated manufacturing systems allowing multiple server failures}, volume = {22}, year = {2025}, }
• (2025). VIOMA: Video-based intelligent ocular misalignment assessment. TASE, 22, 12470--12484. (www)

  @article{ , abstract = {The measurement of ocular alignment is critical for the diagnosis of strabismus. Current clinical methods for assessing ocular misalignment are subjective and frequently rely on the expertise of practitioners and the extent of patient cooperation. Computer-aided diagnosis methods in recent years have improved automation and precision of measurement, but still, fall short of the requirement of clinical practice. In this study, a video-based intelligent ocular misalignment assessment (VIOMA) system, was proposed to provide an objective, repeatable, user-friendly and highly-automated alternative modality for clinical ocular misalignment measurement, in which the automatic cover tests were performed under a control and motor unit, simultaneously the eye movements were tracked using a motion-capture module and assessed through video analysis techniques, determining the presence, type, and magnitude of eye deviation. For system evaluation, an automatic cover tests video dataset for strabismus (StrabismusACT-76) was established, which consists of data from 76 participants. The Bland-Altman plot, used to compare the results of the VIOMA system and human expert, showed a mean value of 1.26 prism diopter (PD) and a half-width of the 95% limit of agreement of ±7.17 PD. VIOMA system presented a mean absolute error of 3.04 PD in measuring the deviation magnitude, within a 5 PD error tolerance. Additionally, the system’s measurements were strongly correlated with that of video labeling with the mean value of -0.26 PD, a half-width of the 95% limit of agreement of ±3.56, and the average error of 1.31 PD. The experiment results indicated that the proposed method has the capability to offer accurate and efficient assessment of ocular misalignment. Note to Practitioners—The motivation behind this work stems from the need to develop an accurate and efficient system for automated ocular misalignment assessment. The subjectivity of manual cover test performed by examiners has led to variability in outcomes, and certain existing computerized methods have limitations in terms of automation, measurement accuracy, and applicability in clinical practice. Faced with these challenges, we proposed VIOMA, by establishing an apparatus for automatic implementation of cover tests and developing assessment algorithms based on strabismic video analysis. This system can objectively and precisely measure ocular misalignment, offering a promising practical solution for clinical intelligent diagnosis of strabismus. The VIOMA system’s potential applications are not limited to strabismus but may extend to other eye-related conditions and beyond.}, archive = {J_TASE}, author = {Yang Zheng and Hong Fu and Ruimin Li and Carly Siu Yin Lam and Jimin Liang and Kaitai Guo and Wai-Lun Lo}, doi = {10.1109/TASE.2025.3545870}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12470--12484}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {VIOMA: Video-based intelligent ocular misalignment assessment}, volume = {22}, year = {2025}, }
• (2025). A combination feature-based reinforcement learning approach via mathematical optimization. TASE, 22, 12455--12469. (www)

  @article{ , abstract = {Reinforcement learning is a promising method for solving decision problems, and its potential has been increasingly recognized for large-scale combinatorial optimization problems in recent years. However, the existing studies on reinforcement learning for cutting stock problems mostly rely on sequence-to-sequence or graph neural network approaches that use the learned experience to make decisions while neglecting the combination features of cutting stock problems. In this paper, we propose a novel reinforcement learning framework for cutting stock problems that integrate integer programming and monotone comparative statics to construct a Markov decision process with a high-quality action space. We start by constructing a new Markov decision process that considers the diagonal structure of the integer programming model for combinatorial optimization problems, and then use column generation to obtain each action by combining multiple decision variables. Furthermore, we design a bipartite graph and related bipartite graph convolutional network to find the solutions. The results show that the proposed reinforcement learning framework provides a high-quality action space, and the designed bipartite graph convolutional network can effectively select the best actions from the action set.Note to Practitioners—This article was motivated by the cutting stock problems that exist in various industrial scenarios such as the wood, steel, paper, and glass industries. We improve the reinforcement learning for the cutting stock problem that can be adopted in industrial scenarios, which can increase the profile and reduce the production cost of industrial enterprises. Our improvement can also be referred to when solving other combinatorial optimization problems that can promote making decisions in industrial production.}, archive = {J_TASE}, author = {Fengyuan Shi and Ying Meng and Jiyin Liu and Lixin Tang}, doi = {10.1109/TASE.2025.3544431}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12455--12469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A combination feature-based reinforcement learning approach via mathematical optimization}, volume = {22}, year = {2025}, }
• (2025). Enhancing the resilience of IEC 61131–3 software with online reconfigurations for fault handling. TASE, 22, 12436--12454. (www)

  @article{ , abstract = {In automated production, resilience describes a system’s capacity to absorb disturbances by reconfiguring itself, thus retaining its Overall Equipment Effectiveness at least partially. This includes online behavior reconfiguration to automatically recover from or prevent faults, collectively called fault handling. Promising research exists for fault handling in automated Production Systems. In process engineering, fault diagnosis and automatic parameter adaptions are already industrially available. However, handling faults in discrete manufacturing requires a series of distinct operations, which cannot be achieved by parameter changes alone. Further, core requirements must remain fulfilled by automatic fault handling approaches, including real-time control and extra-functional requirements like changing operation modes, monitoring interlocks, and an alarming and communication system. This article proposes a concept for reconfigurable IEC 61131–3 software for automatic fault handling, validated by a public reference implementation for a demonstrator, an industrial production system, and a modified industrial test rig. Eight experiments were successfully conducted, showcasing four use cases of the concept: The prevention of faults by avoiding anomalous components, the recovery from a fault state to automatic operation, the definition of previously undefined state variables, and the monitoring of global interlocks to trigger a controlled stop. All mentioned extra-functional requirements are fulfilled. Note to Practitioners—Identification, reporting, diagnosis, and recovery of faults in automated production incur substantial effort. Project-specific code is required for diagnosis, and the recovery and re-initialization are often performed manually. To our knowledge, automatic recovery approaches from scientific literature are not widely used in discrete manufacturing. Reasons may include a frequent disregard of extra-functional requirements mentioned above. Further, some approaches are incompatible with IEC 61131–3 or industry-typical software modularization. This article proposes a PLC software concept that aims to be compatible with real-world challenges and solutions. The functional software is vertically modularized from organizational hardware-level code. The horizontal modularization separates devices or equipment groups. Support for multiple changing operation modes including two types of controlled stop (run to completion or abort), alarming, data exchange, and global interlocks are incorporated. A prototypical IEC 61131–3 implementation is publicly available that separates a reusable generic part from hardware-specific and project-specific code. The resulting control code is highly reusable, such that all modes (derived from PackML), including dynamic reconfigurations, are composed from the same software modules. Note that we do not expect the concept to be well-adoptable in continuous processes, as elaborated in the Preliminaries section.}, archive = {J_TASE}, author = {Jan Wilch and Birgit Vogel-Heuser and Florian Sax and Simon Rüth and Ulrich Oeckl and Bernhard Wohlschläger and Yu-Ming Hsieh and Fan-Tien Cheng}, doi = {10.1109/TASE.2025.3543626}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12436--12454}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing the resilience of IEC 61131–3 software with online reconfigurations for fault handling}, volume = {22}, year = {2025}, }
• (2025). FoundationGrasp: Generalizable task-oriented grasping with foundation models. TASE, 22, 12418--12435. (www)

  @article{ , abstract = {Task-oriented grasping (TOG), which refers to synthesizing grasps on an object that are configurationally compatible with the downstream manipulation task, is the first milestone towards tool manipulation. Analogous to the activation of two brain regions responsible for semantic and geometric reasoning during cognitive processes, modeling the intricate relationship between objects, tasks, and grasps necessitates rich semantic and geometric prior knowledge about these elements. Existing methods typically restrict the prior knowledge to a closed-set scope, limiting their generalization to novel objects and tasks out of the training set. To address such a limitation, we propose FoundationGrasp, a foundation model-based TOG framework that leverages the open-ended knowledge from foundation models to learn generalizable TOG skills. Extensive experiments are conducted on the contributed Language and Vision Augmented TaskGrasp (LaViA-TaskGrasp) dataset, demonstrating the superiority of FoundationGrasp over existing methods when generalizing to novel object instances, object classes, and tasks out of the training set. Furthermore, the effectiveness of FoundationGrasp is validated in real-robot grasping and manipulation experiments on a 7-DoF robotic arm. Our code, data, appendix, and video are publicly available at https://sites.google.com/view/foundationgrasp. Note to Practitioners—This research is motivated by the challenge of generalizable task-oriented grasping skill learning. Solving such a challenge could significantly improve the robot’s level of automation and intelligence in tool manipulation for household and industrial tasks. Existing methods struggle with handling unseen objects and tasks in dynamic, open-world environments. To overcome this limitation, we propose to leverage the open-ended knowledge from foundation models to improve the generalization capabilities of existing TOG methods. This way, the robot can perform TOG w.r.t. unseen objects and tasks, facilitating downstream tool manipulation. Overall, this research has broad applicability to various scenarios involving tool manipulation, such as cleaning kitchenware and assembling parts in industrial contexts.}, archive = {J_TASE}, author = {Chao Tang and Dehao Huang and Wenlong Dong and Ruinian Xu and Hong Zhang}, doi = {10.1109/TASE.2025.3542418}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12418--12435}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FoundationGrasp: Generalizable task-oriented grasping with foundation models}, volume = {22}, year = {2025}, }
• (2025). Time-optimal trajectory planning with clearly defined initial guess for aerial suspended payload throwing. TASE, 22, 12406--12417. (www)

  @article{ , abstract = {Autonomous Aerial Vehicles (AAVs), particularly quadrotors, have gained substantial attention in recent years due to their high agility, substantial convenience, and significant potential in hazardous missions such as military surveillance and disaster relief. This paper focuses on the aerial throwing problem, aiming to develop a time-optimal method for air-dropping cable-suspended payloads. The contributions of the paper are: 1) a fast approach is presented to streamline the quadrotor’s state management by directly mapping and planning at the quadrotor state space (position, velocity, acceleration); 2) a clearly defined initial guess is provided for aerial suspended throwing tasks, which speeds up the planning process. This methodology not only enhances the convenience of quadrotor navigation, but also fosters a more direct and efficient control scheme. The efficacy and feasibility of the proposed method are validated through both numerical simulations and practical experiments, demonstrating the potential for rapid and accurate payload throwing with cable-suspended systems. Note to Practitioners—This study is driven by the need to enhance aerial payload throwing task in hazardous scenarios, such as disaster relief and military surveillance where precision and speed are crucial. While quadrotors serve as agile platforms for such operations, existing methods lack a rapid planning approach that can directly plan at the quadrotor state space (position, velocity, acceleration). Our work introduces a warm start strategy, which significantly hastens the planning process, enabling faster throwing of cable-suspended payloads. Future extensions of this work could focus on integrating adaptive elements that respond to environmental feedback in real-time, thus broadening the practical applicability of the method in real-world conditions.}, archive = {J_TASE}, author = {Rui Cao and Yongchun Fang and Xiao Liang}, doi = {10.1109/TASE.2025.3544272}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12406--12417}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time-optimal trajectory planning with clearly defined initial guess for aerial suspended payload throwing}, volume = {22}, year = {2025}, }
• (2025). Distributed cooperative guidance model-free control for a cluster of disk-type autonomous underwater gliders. TASE, 22, 12395--12405. (www)

  @article{ , abstract = {This paper focuses on a distributed cooperative guidance model-free control method for a cluster of under-actuated disk-type autonomous underwater gliders (AUGs) in the presence of unknown kinetic model parameters and ocean disturbances. Firstly, a distributed cooperative motion generator is designed to generate reference path points, and then a cooperative control method is proposed based on the update path parameters. Secondly, a three-dimensional (3D) guidance law is constructed by employing closed 3D vector fields. Finally, data-driven filtered adaptive extended state observers (DFAO) are proposed to deal with the unknown input gains, internal uncertainties and external disturbances of the disk-type AUGs, and an adaptive kinetic control law is designed by using the knowledge learned from the observers. Simulation results demonstrate the effectiveness of the proposed 3D distributed cooperative guidance model-free control method for disk-type AUGs subject to fully unknown kinetics. Note to Practitioners—The disk-type AUG has the characteristic of long operational endurance, capable of functioning continuously for several months when fully loaded. Consequently, this type of glider offers an advantage in establishing oceanic sensor networks. To achieve this goal, two technical challenges arise: the coordination among multiple gliders and the anti-disturbance control of individual glider. Our research focuses on distributed cooperative guidance and model-free control issues for multiple underwater gliders. First, we propose a distributed cooperative guidance scheme to maintain a specific formation among the gliders. Additionally, while ensuring control effectiveness, we employ data-driven methods to estimate uncertain kinetic model parameters. Our approach is not only theoretically viable but also ready for industrial application, thus filling a gap in underwater glider technology.}, archive = {J_TASE}, author = {Liyu Lu and Haoliang Wang and Nan Gu and Zhouhua Peng and Weidong Zhang}, doi = {10.1109/TASE.2025.3543530}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12395--12405}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed cooperative guidance model-free control for a cluster of disk-type autonomous underwater gliders}, volume = {22}, year = {2025}, }
• (2025). Bandwidth awareness related event triggered robust control on power system with VSG-LFC scheme under unreliable network. TASE, 22, 12383--12394. (www)

  @article{ , abstract = {This paper investigates the frequency regulation problems with high level renewable energy sources (RESs) penetration under unreliable network. A virtual synchronous generator (VSG) incorporated with load frequency control (LFC) method (VSG-LFC) is proposed by introducing an adaptive event-triggered mechanism with the awareness of the bandwidth occupation rate (AETM-ABOR) to reduce the system sensitivity to the penetration level of RESs and achieve a balance between control performance and network resource usage under limited communication environments. Then, a Lyapunov-Krasovskii functional (LKF) with the two-sided looped functional terms is proposed to derive criteria that satisfies the $H_{\infty }$ performance index and controller design. Finally, simulation evaluations on numerical models and a two-area four-machine power system are conducted to verify the availability and effectiveness of this development. Note to Practitioners—This paper was motivated by the problem of frequency regulation. As the penetration of RESs into the power system is gradually increasing, the lack of inertial characteristics of power electronics that connect RESs to the power system poses a threat to the frequency stability of the power system. A VSG-LFC strategy is applied to effectively compensate the equivalent system inertia. Meanwhile, an adaptive event-triggered mechanism with the awareness of the bandwidth occupation rate is designed to compromise the data transmission and system control performance. Qualitative and quantitative analyses are conducted to illustrate that this design is capable of providing expected performance.}, archive = {J_TASE}, author = {Yingying Shao and Zhenzhen Zhang and Hao Chen and Shouming Zhong}, doi = {10.1109/TASE.2025.3545901}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12383--12394}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bandwidth awareness related event triggered robust control on power system with VSG-LFC scheme under unreliable network}, volume = {22}, year = {2025}, }
• (2025). Dynamic modeling and control for an offshore semisubmersible floating wind turbine. TASE, 22, 12371--12382. (www)

  @article{ , abstract = {Floating wind turbines (FWTs) hold significant potential for the exploitation of offshore renewable energy resources. Nevertheless, prior to the construction of FWTs, it is imperative to tackle several critical challenges, especially the issue of performance degradation under combined wind and wave loads. This study initiates with the development of a simplified nonlinear dynamical model for a semi-submersible FWT. In particular, both the rotor dynamics and the finite rotations of the platform are considered in presented modeling approach, thereby effectively capturing the complex interplay between the platform, tower, nacelle, and rotor under combined wind and wave loads. Subsequently, based on the developed FWT model, a novel adaptive nonlinear pitch controller is formulated with the goal of striking a trade-off between regulating power generation and reducing platform motion. Notably, the proposed control strategy adopts a continuous control approach, strategically beneficial in circumventing the chattering phenomenon commonly associated with sliding mode control. Furthermore, the controller integrates an online approximator and a robust integral of the sign of the tracking error, facilitating real-time learning of system unknown dynamics while compensating for bounded disturbances. Finally, both the accuracy of the established nonlinear FWT model in predicting key dynamics and the superiority of the presented pitch controller are validated through comprehensive comparative studies. Note to Practitioners—This paper addresses the conflicting goals between power regulation and load mitigation for floating wind turbines (FWTs) to ensure the reliable operation of wind turbine systems. This remains an ongoing challenge due to the inherent complexity of existing FWT models, frequently resulting in controllers crafted using linearized representations that fail to accommodate real-world uncertainties effectively. Through the utilization of a simplified physical-based nonlinear FWT model, a novel adaptive nonlinear pitch controller emerges as a promising solution. Notably, the developed nonlinear FWT model elucidates the coupling between rotor and platform degrees of freedom clearly and succinctly, facilitating the design of intelligent controllers. Our approach demonstrates the capability to concurrently regulate power production and stabilize the platform. Additionally, an online approximator is integrated into the controller to capture system dynamics, thus augmenting adaptability and diminishing reliance on high-gain feedback compensation. Importantly, this control strategy holds promise for extension and implementation in various other renewable energy systems.}, archive = {J_TASE}, author = {Yingjie Gong and Qinmin Yang and Hua Geng and Wenchao Meng and Lin Wang}, doi = {10.1109/TASE.2025.3541730}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12371--12382}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic modeling and control for an offshore semisubmersible floating wind turbine}, volume = {22}, year = {2025}, }
• (2025). Independent observers-based fault diagnosis for multiple sensor faults of full-vehicle active suspension systems with inaccurate models. TASE, 22, 12358--12370. (www)

  @article{ , abstract = {In this paper, an independent observers-based fault diagnosis method is proposed for multiple sensor faults of the full-vehicle active suspension system with an inaccurate model. To address the problem of model uncertainties brought by the linearized full-vehicle suspension model, unmodelled dynamics, parametric uncertainties, and external disturbances are first combined into an integrated uncertain term. Disturbance observers are designed to online track the integrated uncertain terms, whose estimates will be employed to decrease the influence of model uncertainties on fault diagnosis. An independent fault diagnosis observer is designed for each sensor separately, where only the measurement of the matched sensor is taken as the observer input. In this way, each fault diagnosis observer works independently and interactions between measurements of multiple faulty sensors can be decoupled to locate the sensor faults. Anomalies of each sensor are monitored by the fault diagnosis observer in a one-to-one relationship such that fault detection and isolation can be realized at the same time. In particular, the sensitivity and robustness of the fault diagnosis method are improved with the estimates of the integrated uncertain terms injected into the fault diagnosis observers to emulate model uncertainties. Effectiveness of the proposed scheme has been validated via simulation results considering multiple sensor faults occurring at different time or simultaneously. Note to Practitioners—Good reliability of vehicle subsystems is demanded to provide stable rideability and high maneuverability for vehicles, especially for electric vehicles. To guarantee the reliability of the essential components in practice motivates us to focus on the fault diagnosis problems for sensors of vehicle suspension systems. In this paper, considering the possible scenario of multiple sensor faults, an independent observers-based fault diagnosis scheme is designed for full-vehicle suspension systems. By using the proposed method, we can detect the sensor failure and isolate the faulty sensors simultaneously. An independent fault diagnosis observer is built for each sensor with its single sensor measurement taken as the input of the observer. In this way, interactions of the incorrect measurements from the multiple faulty sensors can be removed to avoid interference from these measurement couplings in locating the faulty sensors. Anomalies of each sensor are monitored by its matched fault diagnosis observer in a one-to-one relationship. Different from the existing methods, model uncertainties brought by the linearized full-vehicle suspension model are observed online and their estimates are employed to build the fault diagnosis observers to narrow the gap between the reality and model, and promote the accuracy and effectiveness of the model-based fault diagnosis method. However, incipient sensor faults cannot be effectively diagnosed with our method, which is one drawback of this proposed fault diagnosis scheme. For future work, fault accommodation strategy will be investigated to compensate for the sensor failure and retain satisfactory system performance. Real vehicle experiments will be conducted to test and improve the algorithm, and thus promote the designed method into practical use. The proposed fault diagnosis method for multiple sensor faults can be extended to other mechanical control systems with similar structure of system equations, in addition to vehicle suspension systems.}, archive = {J_TASE}, author = {Shuai Yan and Yuanqing Xia and Di-Hua Zhai}, doi = {10.1109/TASE.2025.3543338}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12358--12370}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Independent observers-based fault diagnosis for multiple sensor faults of full-vehicle active suspension systems with inaccurate models}, volume = {22}, year = {2025}, }
• (2025). G²-mapping: General gaussian mapping for monocular, RGB-d, and LiDAR-inertial-visual systems. TASE, 22, 12347--12357. (www)

  @article{ , abstract = {In this paper, we introduce G2-Mapping, a novel method to comprehensively support online monocular, RGB-D, and LiDAR-Inertial-Visual systems, employing 3D gaussian points as scene representation. There are several issues when applying 3d gaussian splatting (3DGS) techniques to simultaneous localization and mapping (SLAM) 1) for monocular, the lack of depth information makes scene initialization difficult and large baseline positioning challenging; 2) differentiable rendering with respect to depth and pose has not been implemented in 3DGS, making it difficult to directly apply to the SLAM system; 3) strategy for updating the scene with incoming online frames is not present, which may lead to memory overflow. In order to overcome problems mentioned above, we formulate a mathematical derivation and propose a differentiable rendering approach that leverages both depth and color to optimize the scene and pose. We introduce a simplified odometry that provides a metric depth estimation for monocular and enhance the low-overlap scene availability. A scale consistency and uncertainty weighted optimization is further proposed to eliminates the impact of inaccurate depth prediction. Our proposed scene updating strategy effectively prevents rapid memory growth. Tracking and mapping are performed alternatively to achieve precise localization and synchronous high-fidelity map reconstruction. Extensive experiments demonstrate that our G2-Mapping surpasses feature-based SLAM in localization precision and exceeds state-of-the-art neural SLAM methods in the fidelity of view synthesis. Note to Practitioners—This paper is dedicated to tackling the efficiency challenges in multi-source SLAM and map reconstruction, aiming to generate pose and high-fidelity maps synchronously. We introduce G2-Mapping, a novel framework that leverages the power of 3D Gaussian points for scene representation, offering a universal solution for monocular, RGB-D, and LiDAR-Inertial-Visual systems. By developing a comprehensive differentiable renderer and presenting a strategy for dynamic scene updating, G2-Mapping significantly advances the state-of-the-art in localization precision and view synthesis fidelity. Although the approach is highly promising, it currently relies on the accuracy of depth prediction networks and requires further optimization for handling sparse LiDAR data. Future research will focus on enhancing these aspects, aiming to seamlessly integrate G2-Mapping into practical applications within robotics, autonomous vehicles, and augmented reality, where robust and efficient SLAM solutions are paramount.}, archive = {J_TASE}, author = {Lin Chen and Boni Hu and Jvboxi Wang and Shuhui Bu and Guangming Wang and Pengcheng Han and Jian Chen}, doi = {10.1109/TASE.2025.3541255}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12347--12357}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {G²-mapping: General gaussian mapping for monocular, RGB-d, and LiDAR-inertial-visual systems}, volume = {22}, year = {2025}, }
• (2025). Fully distributed model-free adaptive sliding mode control for MASs with hybrid-attacked topology. TASE, 22, 12336--12346. (www)

  @article{ , abstract = {A fully distributed model-free adaptive sliding mode control (MFASMC) strategy is proposed in this paper for unknown nonlinear multi-agent systems (MASs), in which topology networks are exposed to hybrid attacks consisting of denial-of-service (DoS) and false data injection attacks. Hybrid attacks in topology networks can result in neighbor information dropouts and inaccuracies among agents. First, the MASs with unknown dynamics are translated to equivalent linear data equations by the dynamic linearization technique. Second, the impact of neighbor information dropouts caused by DoS attacks is mitigated by a designed attack compensation mechanism, in which the compensation error is guaranteed to be bounded in the sense of mathematical expectation. Then, a fully distributed MFASMC algorithm, which does not depend on knowledge of the Laplacian matrix, is designed to improve the robustness of MASs with topology networks exposed to hybrid attacks, thus indirectly mitigating the impact of neighbor information inaccuracies. Finally, the consensus error is rigorously proved to be bounded in the sense of mathematical expectation, and the validity of the proposed strategy is confirmed by simulations. Note to Practitioners—This paper aims to develop a fully distributed MFASMC method to address the consensus problem for unknown MASs with hybrid attacks in network topologies. A hybrid attack compensation mechanism is proposed to mitigate the effects of neighbor information dropouts caused by hybrid attacks. By combining the sliding mode control theory with the model-free adaptive control strategy, the system’s robustness is improved to relieve the impact of neighbor information inaccuracies attributed to hybrid attacks. Since the proposed algorithm does not depend on the mathematical model and Laplace matrix of systems, it can be applied to large-scale MASs with unknown models to solve network topology security problems, such as multiple subway trains, wireless communication systems, and microgrid systems. Furthermore, the proposed method is simple in design, widely adaptable, and robust, making it easier to apply to practical systems and more friendly to control engineers.}, archive = {J_TASE}, author = {Shitao Duan and Guangdeng Chen and Qi Zhou and Hongyi Li and Tingwen Huang}, doi = {10.1109/TASE.2025.3543537}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12336--12346}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully distributed model-free adaptive sliding mode control for MASs with hybrid-attacked topology}, volume = {22}, year = {2025}, }
• (2025). An online learning-based mACO approach for hot rolling scheduling problems involving dynamic order arrivals. TASE, 22, 12322--12335. (www)

  @article{ , abstract = {The hot rolling scheduling problem involving dynamic order arrivals (HRSP-DOA) is pivotal in promoting Industry 4.0 initiatives in steel companies. This paper introduces a novel multi-objective ant colony optimization (mACO) algorithm enhanced with online learning strategies to address the HRSP-DOA. The problem is formulated as a variant of the prize-collecting dynamic vehicle routing problem (PC-DVRP). Its primary mission is to select high-reward orders to form a set of rolling rounds and sequence them within each rolling round to satisfy various technical constraints, thereby minimizing penalty costs and maximizing collected prizes. As new orders arrive, the problem’ s decision and objective spaces undergo unexpected changes, requiring the algorithm to make a quick response. To address these challenges, we propose an mACO with multiple heuristic matrices to search for the Pareto optimal set (POS) for the HRSP-DOA. Furthermore, we introduce the following online learning strategies to enhance mACO’s performance: 1) the upper confidence bound (UCB1) algorithm for selecting the most promising heuristic matrix, which is analogous to a multi-armed bandit problem; and 2) the K-means clustering method for transferring pheromone knowledge from the current event time to the following one. The well-designed comparative tests demonstrate that the proposed approach outperforms state-of-the-art algorithms on all synthetic instances. Component analysis further validates the effectiveness of the proposed online learning mechanisms. Finally, we present a real-world case study from an iron and steel company to verify its applicability. Note to Practitioners—In the context of Industry 4.0, the hot rolling scheduling problem involving dynamic order arrivals (HRSP-DOA) is pivotal for smart steel manufacturing, and it is formulated as a variant of the prize-collecting dynamic vehicle routing problem (PC-DVRP). By leveraging problem-specific properties, this paper develops a multi-objective ant colony optimization (mACO) algorithm with multiple heuristic matrices to address the challenges posed by changes in both decision-making and objective spaces caused by dynamic order arrivals. We also introduce two online learning strategies to enhance search efficiency: 1) the upper confidence bound (UCB1) algorithm for selecting the most promising heuristic matrix; and 2) the K-means clustering method for transferring pheromone knowledge in dynamic environments. The effectiveness of the proposed approach is validated through computational experiments. Furthermore, our dynamic multi-objective optimization algorithm can be implemented in real-world decision support systems and applied to other variants of the DVRP.}, archive = {J_TASE}, author = {Sheng-Long Jiang and Qie Liu and Ling-Ling Cao and Liangliang Sun}, doi = {10.1109/TASE.2025.3540042}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12322--12335}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An online learning-based mACO approach for hot rolling scheduling problems involving dynamic order arrivals}, volume = {22}, year = {2025}, }
• (2025). NeSLAM: Neural implicit mapping and self-supervised feature tracking with depth completion and denoising. TASE, 22, 12309--12321. (www)

  @article{ , abstract = {In recent years, there have been significant advancements in 3D reconstruction and dense RGB-D SLAM systems. One notable development is the application of Neural Radiance Fields (NeRF) in these systems, which utilizes implicit neural representation to encode 3D scenes. However, the depth images obtained from consumer-grade RGB-D sensors are often sparse and noisy, which poses significant challenges for 3D reconstruction and affects the accuracy of the representation of the scene geometry. Furthermore, existing methods select random pixels for camera tracking, leading to inaccurate localization in real-world indoor environments. To this end, we present NeSLAM, an advanced framework that achieves accurate and dense depth estimation, robust camera tracking, and realistic synthesis of novel views. First, a depth completion and denoising network is designed to provide dense geometry prior and guide the neural implicit representation optimization. Second, we propose a NeRF-based self-supervised feature tracking algorithm for robust real-time tracking. Experiments on various indoor datasets demonstrate the effectiveness and accuracy of the system in reconstruction, tracking quality, and novel view synthesis. Note to Practitioners—Traditional SLAM methods usually use the sparse point cloud to represent the scene, resulting in poor scene representation capability. Our method proposes a neural implicit representation method with depth completion and denoising network and feature tracking method, achieves accurate scene reconstruction and accurate pose estimation in various indoor scenes. The depth completion and denoising network provide accurate depth information associated with depth uncertainty, which is used to improve the geometry consistency. The NeRF-based self-supervised feature tracking method improve the accuracy and robustness for camera tracking. The experimental results demonstrate the accuracy and effectiveness of this method in different scenes.}, archive = {J_TASE}, author = {Tianchen Deng and Yanbo Wang and Hongle Xie and Hesheng Wang and Rui Guo and Jingchuan Wang and Danwei Wang and Weidong Chen}, doi = {10.1109/TASE.2025.3541064}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12309--12321}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {NeSLAM: Neural implicit mapping and self-supervised feature tracking with depth completion and denoising}, volume = {22}, year = {2025}, }
• (2025). Design and implementation of fuzzy-mode-based fault isolation and fault-tolerant control for aircraft electric braking systems. TASE, 22, 12297--12308. (www)

  @article{ , abstract = {This paper addresses the fault isolation, estimation, and fault-tolerant control scheme for the aircraft electric anti-skid braking system (EABS) in the presence of actuator and sensor faults. First, the inherently nonlinear dynamics of EABSs are represented by a Takagi-Sugeno (T-S) fuzzy model, incorporating immeasurable antecedent variables to capture the time-varying characteristics. Second, based on the output equivalence principle, a fuzzy observer with unmatched antecedent variables is proposed to achieve isolation and estimation of actuator and sensor faults. The designed observer can guarantee the sensitivity to specific faults while enhancing the robustness to disturbances. The estimated fault information is then utilized to develop a fault-tolerant control strategy, ensuring effective fault compensation and tracking performance. Subsequently, the design of separate and integrated frameworks for the estimation and control units is considered, taking their interaction into account to achieve state and fault isolation, estimation, fault compensation, and tracking control. Finally, hardware-in-the-loop experimental results verify the effectiveness and real-time performance of the proposed fault isolation and fault-tolerant control method, demonstrating the practical applicability of the proposed framework. Note to Practitioners—The aircraft anti-skid braking system (ABS) is crucial for ensuring the safety during landing, taxiing, and other ground movements. This paper focuses on developing reliable fault isolation and fault-tolerant control strategies to maintain ABS performance and efficiency in the presence of faults. The proposed approach employs a fuzzy model to analyze the effects of various faults on system outputs, enabling precise fault isolation and estimation for simultaneous multiple faults. The reconstructed fault information is then integrated to enhance the fault-tolerant control mechanism. This ensures that braking performance can be maintained, even in the presence of multiple simultaneous faults, thereby enhancing system robustness and safety. Moreover, the proposed strategy holds potential applications in other safety-critical domains, such as rail transportation and aerospace vehicles. Future research will explore the integration of historical data to further enhance the accuracy of the fault diagnostic and accommodation units.}, archive = {J_TASE}, author = {Yiyun Zhao and Zheng Wu and Fanbiao Li and Tao Yang and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2025.3543647}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12297--12308}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and implementation of fuzzy-mode-based fault isolation and fault-tolerant control for aircraft electric braking systems}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance-based distributed event-triggered data-driven frequency control for hybrid power systems under jamming attacks. TASE, 22, 12283--12296. (www)

  @article{ , abstract = {The reliance on open communication networks poses significant challenges to power systems. This paper studies a prescribed performance-based scalable distributed event-triggered model-free adaptive control (DETMFAC) scheme for multi-area load frequency control (LFC) systems that are integrated with wind-energy, analyzing both scenarios without and with jamming attacks. First, a transformed-error-based dynamic linearization model (DLM) for distributed outputs is developed to handle the unknown system dynamics under performance constraints. By directly incorporating the DLM and the prescribed performance function into the criterion function, we derive a novel prescribed performance-based DETMFAC scheme. This algorithm improves the control performance while overcoming the problems caused by global topology information and model dependence as in the traditional distributed LFC design approaches by introducing a parameter adaptation mechanism. Additionally, a novel jamming attack model based on the signal-to-interference-plus-noise ratio (SINR) is introduced to illustrate the interaction between the power system and the malicious attacker in a Stackelberg game framework. Then a secure scheme with attack compensation is proposed to mitigate the negative attack impact by reconstructing the LFC commands via linear estimation. Experimental results for a four-area power system demonstrate the designed scheme. Note to Practitioners—LFC is a fundamental strategy for frequency regulation in networked power systems. However, traditional LFC systems are vulnerable to cyber-attacks, tracking performance constraints, and bandwidth limitations due to the interconnection between different subareas. This paper proposes a prescribed performance-based DETMFAC scheme and a secure control scheme with a resilience mechanism to effectively address these challenges in a coordinated manner, integrating an event-trigger, parameter identifier, attack compensator, and input update law. The control input is only updated when the triggering condition is met, based on the distributed output and identified adaptive parameters. In the case of an attack, the event-trigger ceases operations and the attack compensator generates an estimated input signal. Importantly, the proposed LFC scheme relies solely on local input and output data (i.e., model-free) and can be implemented in a fully distributed architecture (i.e., scalable). Thus, practitioners can seamlessly integrate the designed strategies with existing LFC systems to enhance performance, security, and efficiency. The results provide a valuable reference for scalable, cost-effective and secure distributed data-driven control design of multi-area LFC systems and promote related application research.}, archive = {J_TASE}, author = {Rui Hou and Li Jia and Xuhui Bu and Jianfang Li}, doi = {10.1109/TASE.2025.3543578}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12283--12296}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance-based distributed event-triggered data-driven frequency control for hybrid power systems under jamming attacks}, volume = {22}, year = {2025}, }
• (2025). Adaptive tracking control of uncertain nonlinear multiagent systems: A dynamic event-based method via prescribed-time characteristics. TASE, 22, 12271--12282. (www)

  @article{ , abstract = {The tracking control issue for a class of nonlinear multiagent systems under adaptive fuzzy prescribed-time dynamic event-triggered control is explored in this paper. A DETM relying on the characteristics of prescribed-time functions is designed and characterized by utilizing a prescribed-time mapping function as the triggering threshold. This approach not only mitigates the oscillation issues associated with traditional threshold-based event control inputs, which are often caused by unreasonable thresholds and the demand for enhanced performance, but also significantly reduces the communication burden. Additionally, a conversion strategy for time-varying constraints is considered, allowing users to define the tracking accuracy and settling time regardless of system parameters and initial conditions. It has been proven that the proposed strategy enables the tracking error to converge within a prescribed error range within a specified time frame. The effectiveness and practicality of this scheme are validated through simulation results. Note to Practitioners—The core focus of this work lies in enhancing the effectiveness of engineering applications, with a particular emphasis on optimizing the performance of MASs, efficiently utilizing communication resources, and adapting to uncertainties and external disturbances. These requirements are crucial across various domains, such as circuit systems, unmanned aerial vehicle formations, and sensor network systems. Within the theoretical framework of prescribed-time control, this study integrates prescribed-time characteristics to design a novel DETM. This mechanism not only significantly reduces communication resources but also effectively mitigates potential oscillations in control inputs during critical moments, thereby providing a more robust and efficient solution for practical application scenarios.}, archive = {J_TASE}, author = {Renhua Huang and Wencheng Wang and Jiuxiang Dong}, doi = {10.1109/TASE.2025.3540851}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12271--12282}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tracking control of uncertain nonlinear multiagent systems: A dynamic event-based method via prescribed-time characteristics}, volume = {22}, year = {2025}, }
• (2025). Self-healing control for multivariable processes based on simple and canonical correlation analyses. TASE, 22, 12257--12270. (www)

  @article{ , abstract = {Abnormal working conditions often occur in modern industrial processes. Subsequently, if maintenance is not carried out in a timely manner, such abnormal working conditions can cause changes in system output variables, resulting in a decrease in product quality and even production accidents. This study proposes a self-healing control framework that extracts system input and output variable historical data under normal conditions. Then, based on simple correlation analyses and canonical correlation analyses, the correlation between historical input and output data is constructed, and the self-healing control law is calculated using the current batch of system output. The setpoint of the underlying controller and process variables are adjusted according to the control law to restore the system output variables to normal or near normal, thus ensuring the stability of product quality. This study demonstrates the convergence of system output variables and analyzes the computational complexity of the two algorithms. Finally, applied to water storage and continuous stirring reactor systems, the proposed self-healing control framework is validated to have a significant effect in restoring system output variables to normal or near normal, and presents significant safety and economic value in industrial production. Note to Practitioners—Abnormal conditions and fault conditions often occur in traditional complex industrial production processes, such as temperature, pressure fluctuations and pipeline blockages. When abnormal working conditions occur, certain key variables will fluctuate, which will have a huge impact on system safety and stable production. Therefore, this paper proposes a self-healing control system. The system can timely detect the occurrence of abnormal working conditions and make corresponding self-healing control decisions for abnormal working conditions. This system is compatible with several types of algorithms, and two different algorithms are designed in this article. Both algorithms can maintain the stability of key system output variables, and each has its own characteristics of adjustment and operation. In order to explore the application scope of the system, this paper carried out tests in various application scenarios and abnormal working conditions. Experimental results show that the proposed self-healing control system has significant self-healing effects in various industrial production processes.}, archive = {J_TASE}, author = {Jingfeng Zhao and Yukun Shi and Youqing Wang}, doi = {10.1109/TASE.2025.3543328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12257--12270}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-healing control for multivariable processes based on simple and canonical correlation analyses}, volume = {22}, year = {2025}, }
• (2025). Data-driven reinforcement learning-based forgetting factor iterative learning control. TASE, 22, 12245--12256. (www)

  @article{ , abstract = {In this paper, a data-driven control scheme is proposed by integrating both reinforcement learning (RL) and iterative learning control (ILC) methodologies. To address the limitations of conventional ILC and extend its applicability, a new approach called Forgetting Factor ILC (FFILC) has been introduced. The FFILC exhibits improved performance in the presence of non-repetitive noise, disturbances, and system uncertainties compared to traditional ILC. This improvement is attributed to its ability to incorporate past data, enhancing adaptability to changing conditions. RL is utilized to determine the optimal forgetting factor. During initial iterations, larger errors are common, and a dynamic forgetting factor expedites convergence. In addition, RL equips the system with intelligence, facilitating superior adaptive performance and resilience as the environment changes throughout iterations. The paper provides a mathematical proof to analyze the convergence of RL-FFILC. Simulations involving a quadrotor system and a drum-type boiler turbine plant demonstrate the superior performance of RL-FFILC compared to some advanced control methods. Note to Practitioners—Industrial automation frequently involves repetitive processes, for which ILC is a suitable method to enhance performance. However, conventional ILC methods can be limited by special systems, non-repetitive disturbances and noise. For practitioners, RL-FFILC, as an intelligent robust control, offers a practical solution to enhance control performance in dynamic settings. This method is data-driven, meaning it does not require accurate model identification prior to implementation, making it suitable for applications where obtaining precise system models is difficult or impractical. The effectiveness of RL-FFILC is demonstrated through simulations on two distinct systems, a quadrotor and a drum-type boiler turbine plant. The quadrotor, an autonomous system with six degrees of freedom, showcases the method’s applicability to other autonomous systems with similar or fewer degrees of freedom. The drum-type boiler turbine plant, crucial in many industrial processes, highlights the relevance of the method in maintaining system stability and efficiency despite variable operational conditions. These applications fall within Cyber-Physical Systems (CPS), where intelligent data-driven control enhances adaptability and resilience.}, archive = {J_TASE}, author = {Ehsan Soleimani and Ali Khaki Sedigh and Amirhossein Nikoofard}, doi = {10.1109/TASE.2025.3540699}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12245--12256}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven reinforcement learning-based forgetting factor iterative learning control}, volume = {22}, year = {2025}, }
• (2025). ELabrador: A wearable navigation system for visually impaired individuals. TASE, 22, 12228--12244. (www)

  @article{ , abstract = {Visually impaired individuals encounter significant challenges when walking and acting in unfamiliar environments, particularly in outdoor scenarios. The complexity of outdoor environments, characterized by diverse obstacles, traffic signals, and societal norms, poses substantial barriers to mobility of visually impaired individuals and makes long-distance walking especially arduous. Although GPS-based navigation systems can facilitate long-distance travel, they often suffer from location inaccuracies in urban areas and even completely fail indoors. Moreover, these systems lack the capability to provide detailed information about walkways and immediate surroundings, which are crucial for safe and efficient walking. To address these limitations, we introduce a proof-of-concept wearable navigation system named eLabrador, designed to assist visually impaired individuals in long-distance walking in unfamiliar outdoor environments. The eLabrador integrates public maps (e.g. Amap or Google Maps) and GPS for global route planning, while leveraging computational visual perception to provide precise and safe local guidance. This hybrid approach enables accurate and safe navigation for visually impaired individuals in outdoor scenarios. Specifically, the eLabrador utilizes a head-mounted RGB-D camera to capture environmental geometric terrain and objects in outdoor urban environments. These inputs are processed into a 3D semantic map, offering a detailed representation of the surrounding environment. The planning module then integrates this 3D semantic map with route information from the global map (i.e. Amap) to generate an optimized walking path. Finally, the interaction module utilizes the audio-haptic dual-channel to relay navigation instructions to visually impaired user. Together, these three modules work seamlessly to facilitate long-distance navigation for visually impaired individuals in outdoor environments. The eLabrador is evaluated with two real-world outdoor scenarios, involving 10 visually impaired and visually masked participants. The experiments show that eLabrador successfully guides visually impaired participants to their destinations in outdoor environments. Additionally, the eLabrador provides descriptive information about landmarks and other navigation cues, helping visually impaired users better understand their surroundings. Subjective evaluations further indicate that most participants felt a sense of safety and reported an acceptable cognitive load during navigation, indicating its usability and effectiveness. Note to Practitioners—Visually impaired individuals almost cannot walk long distance in unfamiliar outdoor environments. Without proper assistance, their mobility and quality of life can be severely impacted. To address this issue, this article presents a wearable navigation system eLabrador to assist visually impaired individuals in walking outdoors, such as traveling from a residential entrance to a nearby park. Experimental results from real-world scenarios involving 10 participants demonstrate that eLabrador safely guides visually impaired users to their destination, significantly enhancing their mobility and independence.}, archive = {J_TASE}, author = {Meina Kan and Lixuan Zhang and Hao Liang and Boyuan Zhang and Minxue Fang and Dongyang Liu and Shiguang Shan and Xilin Chen}, doi = {10.1109/TASE.2025.3541055}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12228--12244}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ELabrador: A wearable navigation system for visually impaired individuals}, volume = {22}, year = {2025}, }
• (2025). Value set-based method for solving stability boundary of systems with uncertain time delays. TASE, 22, 12213--12227. (www)

  @article{ , abstract = {Existing methods for computing stability boundary of systems with multiple uncertain time delays are either conservative or suffer from low computational efficiency. Addressing this challenge, a value set based method is proposed to solve stability boundary for time delays with high efficiency. Firstly, the characteristic quasi-polynomial of the system is constructed in the periodic multiplication between frequency and time delays (FDM) space. Subsequently, the characteristic quasi-polynomial is transformed by the value set approximations, which is then refined through parameter space subdivision to achieve the required accuracy. Finally, the stability boundary is solved by traversing the parameter space with a bisecting step size. This method is characterized by its efficiency without introducing conservativeness, which is achieved through the decreasing step size space traversal. Numerical examples and a case study based on a microgrid with three distributed generators (DGs) that introduce communication delays are served to validate the accuracy and efficiency.Note to Practitioners—This paper proposes a value set based method to solve stability boundary of the system with multiple uncertain time delays. The method is developed based on the necessary and sufficient criteria. The method leverages the periodicity of FDM and employs a variable step traversal algorithm to decrease the computational burden and achieve high efficiency. Notably, this method can be used for systems with multiple time delays, regardless of the range or quantity of the delays. This method is effective for analyzing systems with multiple time delays, offering support for controller design and enhancing system stability.}, archive = {J_TASE}, author = {Hua Ye and Xiubao Shi and Xiaodong Chu and Yutian Liu and Shumin Sun}, doi = {10.1109/TASE.2025.3540425}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12213--12227}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Value set-based method for solving stability boundary of systems with uncertain time delays}, volume = {22}, year = {2025}, }
• (2025). Motion planning and control of active robot in orthopedic surgery by CDMP-based imitation learning and constrained optimization. TASE, 22, 12197--12212. (www)

  @article{ , abstract = {Current orthopedic surgical robots are widely used in pedicle screw implantation tasks due to their precise positioning capabilities. However, the surgical operation processes, including surgical pose alignment and the drilling of the pedicle screw placement path, remain heavily dependent on the surgeon, indicating that the level of automation still needs improvement. This paper aims to enhance automation in pedicle screw implantation tasks through a combination of imitation learning and constraint optimization, thereby ensuring both reliability and safety. Firstly, a high-level motion planning method, leveraging Cartesian space dynamic movement primitives (CDMP) based imitation learning and an image-guided optical navigation system (I-GONS), is proposed to generate the task space path of the rough surgical pose alignment and fine surgical pose alignment, as well as for the drilling of pedicle screw placement path. Secondly, end-effector velocity control based on position and orientation errors (POE-EVC) is employed to follow the high-level planned path. This is achieved by constructing a quadratic programming (QP) problem with the robot kinematic constraints and manipulability optimization. Concurrently, the low-level motion control is addressed online using a modified linear variational inequality based primal dual neural network (mLVI-PDNN). Experimental results demonstrate the distance errors in multiple pedicle screw implantation tasks are $0.738\pm 0.080$ mm and $0.154\pm 0.031$ mm at the entry and target points, respectively. And the angular errors between the actual drilling path compared to the planned screw placement path are $0.005\pm 0.002$ degrees. These results show that the proposed methodology offers a reliable and innovative solution for the higher level of automation in the pedicle screw implantation procedure.Note to Practitioners—The study is driven by the need for higher levels of automation in orthopedic surgery for pedicle screw implantation tasks. Orthopedic surgical robots are usually deployed in complicated surgeries due to their excellent positioning capabilities. However, they primarily rely on basic kinematic constraints and cooperative control modes, thus lacking a reliable mechanism for active surgical pose alignment and screw path drilling, especially in complex surgical scenarios that involve only optical tracking systems (OTS) without additional vision sensors. For example, current popular OTS systems have a limited field of view and cannot fully acquire information about the surgical scene, which impacts collision-free path planning. To address these issues, a novel methodology combining CDMP based imitation learning with QP based motion control for the task of pedicle screw implantation is introduced. Incorporating surgeons’ prior knowledge of the surgical environment and their participation in the robot’s motion planning using the CDMP model enhances the robot’s autonomy and increases reliability in surgeries, without requiring additional vision sensors. This approach allows for accurate reconstruction of multiple surgical targets of automatic surgical pose alignment and pedicle screw drilling, while respecting manipulability optimization and the robot’s kinematic constraints, rather than merely imitating and ignoring the robot’s kinematic structure. This methodology is also applicable to various clinical surgical operations and industrial applications.}, archive = {J_TASE}, author = {Xingqiang Jian and Yibin Song and Dongdong Liu and Yu Wang and Xueqian Guo and Bo Wu and Nan Zhang}, doi = {10.1109/TASE.2025.3541594}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12197--12212}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Motion planning and control of active robot in orthopedic surgery by CDMP-based imitation learning and constrained optimization}, volume = {22}, year = {2025}, }
• (2025). Traffic agents trajectory prediction based on enhanced bidirectional recurrent network and adaptive social interaction model. TASE, 22, 12182--12196. (www)

  @article{ , abstract = {Accurate prediction of the future trajectory of traffic agents is imperative to the effective motion planning of autonomous vehicles and mobile robots. Despite enormous progress that has been made toward trajectory prediction, dynamic and crowded traffic scenarios pose major challenges to the understanding and forecasting of traffic agents’ motion behavior. In this paper, we propose a novel trajectory prediction method from the perspective of temporal modeling and social interaction. Specifically, we first put forward a recurrent modeling approach to learn temporal features in favor of capturing long-range and short-range temporal dependencies of individual agents. Then, we construct a social feature learning module to capture the sparse and directional interactions among agents while suppressing the spurious connections. Finally, to reduce the accumulated error during prediction, a coordinated bidirectional decoding module is developed where temporal and social features can be properly integrated into the forward and backward prediction processes. Extensive experiments are performed on four real-world trajectory prediction benchmarks, and the results demonstrate the superiority of our method compared with other competing approaches. Detailed ablation studies are also performed to evaluate the effectiveness of each model component. Note to Practitioners—Motion planning is one of the crucial components of autonomous systems, such as intelligent vehicles and mobile robots. For example, the safety and efficiency of motion planning can be drastically improved if the future trajectories of surrounding agents, e.g., pedestrians, bicyclists, cars, etc., can be accurately forecasted. Motivated by the above requirements, this article develops an advanced deep learning model that can learn temporal and social features from trajectory data and perform accuracy prediction. This work aims to enhance the bidirectional recurrent network for dealing with trajectory data with evident temporal characteristics. In addition, this work introduces a novel adaptive social interaction modeling approach that overcomes the inherent defect of the fixed threshold method. This work also addresses the error accumulation problem in the prediction. The proposed model is evaluated on several datasets, and the results demonstrate its effectiveness. Our approach has broad application prospects in autonomous driving and mobile robots.}, archive = {J_TASE}, author = {Xiaobo Chen and Yuwen Liang and Chuan Hu and Hai Wang and Qiaolin Ye}, doi = {10.1109/TASE.2025.3540929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12182--12196}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Traffic agents trajectory prediction based on enhanced bidirectional recurrent network and adaptive social interaction model}, volume = {22}, year = {2025}, }
• (2025). Discrete-time embedded model control scheme for disturbed nonlinear systems with application to quadrotor UAVs. TASE, 22, 12173--12181. (www)

  @article{ , abstract = {In this paper, a discrete-time embedded model control (EMC) scheme is proposed for a class of disturbed nonlinear systems under model uncertainty. First, by invoking the current state, a linear discrete-time embedded model (EM) is real-timely generated by linearizing and discretizing nominal model of nonlinear system. Then, as for the reference dynamics for our control scheme, a reference generator is designed to generate a group of reference control input and reference state. Meanwhile, as for control dynamics for proposed scheme, an auxiliary system and a predictor are combined to compensate for the error of control input. Based on above, a synthetic controller is derived to stabilize the controlled system by combining the reference dynamics and control dynamics. Thus, the discrete-time Lyapunov stability theory is utilized to analyze the overall closed-loop system, and a sufficient stability condition is proposed to guarantee that all the closed-loop states under the proposed EMC scheme are semi-globally ultimately uniformly bounded (UUB), ensuring the ultimate error bounds to be adjusted into tolerable regions. Finally, as for the nonlinear system of quadrotor UAV, some simulations are conducted to illustrate the effectiveness of the proposed control scheme. Note to Practitioners—The motivation of this paper aims to investigate the control stability issue of a disturbed nonlinear system. Currently, these relevant results proposed in existent paper mainly relied on the complicated continuous-time controllers, which are hard to be implemented in practice. Yet, this paper proposes a control scheme based on a discrete-time embedded model, which can ensure the desired control performance. First, a reference signal generator is designed by generating a linear discrete-time embedded model. Based on this, the control error is compensated by combining an auxiliary system and a predictor. Then, a synthetic controller is derived to stabilize the controlled system. The effectiveness of the overall control scheme is validated using a quadrotor UAV nonlinear system, demonstrating that all signals in the closed-loop system under our control scheme are semi-globally uniformly bounded, and the ultimate error bounds can be adjusted to a tolerable range. In future work, we will investigate the problem on input delay compensation based on the control scheme of this paper.}, archive = {J_TASE}, author = {Bing Zhu and Mou Chen and Tao Li and Shuyi Shao}, doi = {10.1109/TASE.2025.3541953}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12173--12181}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discrete-time embedded model control scheme for disturbed nonlinear systems with application to quadrotor UAVs}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time robust synchronization of networked heterogeneous euler-lagrange systems. TASE, 22, 12160--12172. (www)

  @article{ , abstract = {In this paper, we propose a prescribed-time synchronization (PTS) algorithm for networked Euler-Lagrange systems subjected to external disturbances. Notably, the system matrix and the state of the leader agent are not accessible to all agents. The algorithm consists of distributed prescribed-time observers and local prescribed-time tracking controllers, dividing the PTS problem into prescribed-time convergence of distributed estimation errors and local tracking errors. Unlike most existing prescribed-time control methods, which achieve prescribed-time convergence by introducing specific time-varying gains and adjusting feedback values, we establish a class of ${\mathcal {K}}_{T}$ functions and incorporate them into comparison functions to represent time-varying gains. By analyzing the properties of class ${\mathcal {K}}_{T}$ and comparison functions, we ensure the prescribed-time convergence of distributed estimation errors and local tracking errors, as well as the uniform boundedness of internal signals in the closed-loop systems. External disturbances are handled and dominated by the time-varying gains that tend to infinity as time approaches the prescribed time, while the control signal is still guaranteed to be bounded. Finally, a numerical example and a practical experiment demonstrate the effectiveness and innovation of the algorithm. Note to Practitioners—This paper aims to address the issue of prescribed-time synchronization for networked Euler-Lagrange systems. Existing research on asymptotic and finite-time convergence reveals that the settling time for synchronization is significantly influenced by the system’s initial values and controller parameters, making it challenging to be freely pre-designed. In contrast, our proposed prescribed-time synchronization algorithm ensures that all Euler-Lagrange systems achieve synchronization within a prescribed time. The effectiveness of our algorithm has been validated through numerical simulations and physical experiments. In practical applications, our algorithm can be utilized for cooperative control in robotic manipulators and drones. Compared to traditional PD controllers, our proposed algorithm not only offers the advantage of arbitrary settling time configuration in cooperation but also ensures faster response speeds and higher control accuracy, owing to the incorporation of time-varying gains.}, archive = {J_TASE}, author = {Gewei Zuo and Yaohang Xu and Mengmou Li and Lijun Zhu and Han Ding}, doi = {10.1109/TASE.2025.3541052}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12160--12172}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time robust synchronization of networked heterogeneous euler-lagrange systems}, volume = {22}, year = {2025}, }
• (2025). Improved finite-time sliding mode control for multi-agent systems under fuzzy topologies. TASE, 22, 12147--12159. (www)

  @article{ , abstract = {In this paper, a novel sliding mode control (SMC) method is designed to investigate the finite-time consensus tracking (FTCT) problem of the second-order leader-following multi-agent systems (MASs) with imprecise communication topology of each agent. First, a T-S fuzzy model is introduced to characterize the inexact communication topology of the leader-following MASs. Moreover, a fuzzy SMC law is designed to ensure the reachability of the constructed sliding mode surface (SMS). Meanwhile, in light of the partitioning strategy, sufficient conditions for FTCT of the leader-following MASs are established. It is worth mentioning that the distributed SMC method proposed in this paper is based on the unknown sliding gain, which can greatly reduce the conservatism. Finally, the simulation study on a group of single-link robots and a numerical simulation demonstrate the feasibility of proposed control method. Note to Practitioners—This paper aims to ensure the transient performance and good robustness of second-order multi-agent systems. Generally, the classic nonlinear control method, sliding mode control (SMC), is accompanied by a certain level of conservativeness due to its fixed sliding gain matrix. Moreover, excessive conservatism may prevent the proposed control strategy from being implemented in practical engineering. To address this issue, we formulate an improved finite-time SMC framework to reduce the conservatiness. We also designed an improved algorithm based on genetic algorithm (GA) and linear matrix inequalities (LMIs) to solve the difficulties brought by the new control algorithm, thereby facilitating its practical implementation and enhancing the overall performance of the second-order multi-agent systems.}, archive = {J_TASE}, author = {Rong Nie and Wenli Du and Zhongmei Li and Shuping He}, doi = {10.1109/TASE.2025.3539341}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12147--12159}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Improved finite-time sliding mode control for multi-agent systems under fuzzy topologies}, volume = {22}, year = {2025}, }
• (2025). Safe and interpretable human-like planning with transformer-based deep inverse reinforcement learning for autonomous driving. TASE, 22, 12134--12146. (www)

  @article{ , abstract = {Human-like decision-making and planning are crucial for advancing the decision-making level of autonomous driving and increasing acceptance in the autonomous vehicle market, as well as for achieving data closed loop for autonomous driving. However, human-like decision-making and planning methods still face challenges in safety and interpretability, particularly in multi-vehicle interaction scenarios. In light of this, this paper proposes an interpretable human-like decision-making and planning method with Transformer-based deep inverse reinforcement learning. The proposed method employs a Transformer encoder to extract features from the scenario and determine the attention assigned by the ego vehicle to each traffic vehicle, thereby improving the interpretability of planning outcomes. Furthermore, for improved safety in planning, the model is trained on both positive and negative expert demonstrations. The experimental results show that the proposed method enhances model safety while maintaining imitation levels compared to conventional methods. Additionally, the attention allocation results closely align with those of human drivers, indicating the model’s ability to elucidate the importance of each traffic vehicle for decision-making and planning, thereby improving interpretability. Therefore, the proposed method not only ensures high levels of imitation and safety but also enhances interpretability by providing accurate attention allocation results for decision-making and planning. Note to Practitioners—This paper presents a method for enhancing the planning of autonomous vehicles by making it more interpretable and safer. Using Transformer-based deep reinforcement learning, the approach improves clarity by showing how the vehicle prioritizes other traffic participants and learning from both positive and negative examples. This not only enhances safety and decision accuracy but also provides insights into the vehicle’s reasoning process, which is crucial for debugging and increasing user trust. Future work could focus on adapting this method for even more complex driving scenarios.}, archive = {J_TASE}, author = {Jiangfeng Nan and Ruzheng Zhang and Guodong Yin and Weichao Zhuang and Yilong Zhang and Weiwen Deng}, doi = {10.1109/TASE.2025.3539340}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12134--12146}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe and interpretable human-like planning with transformer-based deep inverse reinforcement learning for autonomous driving}, volume = {22}, year = {2025}, }
• (2025). A lyapunov-based framework for trajectory planning of wheeled vehicle using imitation learning. TASE, 22, 12118--12133. (www)

  @article{ , abstract = {Trajectory planning with a learning-based approach has emerged as a crucial element in autonomous unmanned systems and has attracted substantial interest from both academia and industry. However, unresolved issues persist concerning data efficiency, safety, convergence, and generalization within the control pipeline. To address this gap, this work presents a trajectory planning method that combines the differential flatness of wheeled vehicle with global convergence property. Our proposed framework transforms the trajectory planning problem, integrating kinematic constraints into a motion planning paradigm. This transformation significantly reduces the state space associated with trajectory planning. Initially, Gaussian mixture regression (GMR) is employed to learn the nonlinear mapping from flat input, leveraging a limited number of demonstrations solved by the optimal control method. Subsequently, we design an asymmetric quadratic Lyapunov function that incorporates both random barrier information and the potential convergence property of the demonstration trajectories. Based on the optimized parameterized Lyapunov function incorporating the convergence and safety criteria, the analytical supplementary control is subsequently obtained by solving a quadratic programming problem to compensate for the prediction errors of GMR, which make the framework complete. Both numerical and real-world experiments are performed to validate the effectiveness of our framework.Note to Practitioners—This work is motivated by the concept of robot motion primitives. While direct segmented planning based on dynamical system movement is feasible for robots with multiple degrees of freedom, it cannot be directly applied to wheeled vehicles with non-holonomic constraints. In this paper, we propose a method that utilizes the differential flatness properties of wheeled vehicles to bridge the gap between these two application objects. By converting the trajectory planning problem of the mobile vehicles into an optimal control problem, we can obtain a satisfactory optimal trajectory. Solving the optimal control problem involves an iterative process due to its nature as an initial value problem. While some learning-based method offer end-to-end learning capabilities and can perform real-time optimal control, they requires a substantial number of training dataset (different optimal trajectories) and lack convergence validation. The foundation of this work is the imitation learning paradigm. Although learning-based approaches are also capable of directly learning from demonstrations, our approach realizes a combination of Lyapunov theory, which allows us to extract the planning task’s potential properties from a limited number of demonstrations, and the vehicle’s differential flat properties to construct a framework with analytical solution. The framework can effectively re-plan obstacle-free trajectories, even in scenarios with obstacles not given in learning phase. This effectively compensates for the shortcomings of learning-based methods in terms of data reliance, convergence, and interpretability.}, archive = {J_TASE}, author = {Jialun Lai and Zongze Wu and Zhigang Ren and Ci Chen and Qi Tan and Shengli Xie}, doi = {10.1109/TASE.2025.3541409}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12118--12133}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A lyapunov-based framework for trajectory planning of wheeled vehicle using imitation learning}, volume = {22}, year = {2025}, }
• (2025). Co-optimization of tool orientations, kinematic redundancy, and waypoint timing for robot-assisted manufacturing. TASE, 22, 12102--12117. (www)

  @article{ , abstract = {In this paper, we present a concurrent and scalable trajectory optimization method to improve the quality of robot-assisted manufacturing. Our method simultaneously optimizes tool orientations, kinematic redundancy, and waypoint timing on input toolpaths with large numbers of waypoints to improve kinematic smoothness while incorporating manufacturing constraints. Differently, existing methods always determine them in a decoupled manner. To deal with the large number of waypoints on a toolpath, we propose a decomposition-based numerical scheme to optimize the trajectory in an out-of-core manner, which can also run in parallel to improve the efficiency. Simulations and physical experiments have been conducted to demonstrate the performance of our method in examples of robot-assisted additive manufacturing. Note to Practitioners—In robot-assisted manufacturing, how to determine the motion commands according to a sequence of waypoints is a typical problem to be solved where the waypoints represent the positions of a tool-tip. Factors in three aspects need to be planned at each waypoint, including the tool orientation, the tool speed and the redundant degrees-of-freedom on the robotic system. In the trajectory planning step, the objective is always defined as improving the kinematic performance of joint motion in terms of velocity, acceleration, and jerk. Taking the strategy of existing methods that consider these aspects separately will generate less optimal results. This paper presents a new formulation that optimizes all these together while assigning certain manufacturing constraints. Considering that a toolpath can consist of a large number of waypoints in practice, how to improve planning efficiency with limited computer memory is an important issue to be solved. A decomposition based numerical scheme is developed to tackle this problem. The aforementioned issues can be effectively solved by the method proposed in this paper, the performance of which has been demonstrated on a dual robotic system with 6+2 DoFs. The proposed method is general and can also be applied to other types of systems with single or multiple robots as well as other manufacturing methods (e.g. milling).}, archive = {J_TASE}, author = {Yongxue Chen and Tianyu Zhang and Yuming Huang and Tao Liu and Charlie C. L. Wang}, doi = {10.1109/TASE.2025.3542218}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12102--12117}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Co-optimization of tool orientations, kinematic redundancy, and waypoint timing for robot-assisted manufacturing}, volume = {22}, year = {2025}, }
• (2025). State-dependent switching control for nonlinear impulsive switched systems with dwell time. TASE, 22, 12093--12101. (www)

  @article{ , abstract = {This paper investigates the problem of globally exponential stability (GES) for impulsive switched systems via a hybrid switching strategy. First, a switching law combining state-dependent and time-dependent (dwell time) switching strategies is presented in the framework of impulse action. On the basis of the switching law, some GES criteria are proposed by using multiple Lyapunov function technique. A potential relationship among system parameters, impulse action including impulse amplitude and impulse frequency, and the combined switching law is established, which is of vital importance to achieving GES. In addition, when the impulse action involved in switched system is given in terms of stabilizing impulse, the impulsive controller including impulsive gains and impulsive frequency can be designed with the help of the proposed switching law to guarantee the stability of the addressed system. Finally, a simulation example is demonstrated to verify the validity of the proposed switching law and corresponding results. Note to Practitioners—This paper is motivated by the control problem of hybrid systems under a discontinuous environment in practical terms, such as optimal control models in economics, motor controllers in electric vehicles, and traffic flow control in intelligent transportation systems. Due to the hybrid nature of switched systems, they are challenging to compare independently with continuous or discrete systems. Therefore, this paper thoroughly investigates the stability control of nonlinear switched systems in the framework of impulse action. State-dependent switching strategy is an effective control method for on-demand switching. However, existing results often require continuous monitoring of system states, leading to significant computational and communication resource consumption. To address this issue, a hybrid switching strategy including state-dependent and time-dependent switching is firstly proposed for impulsive switched systems in this paper. A relationship among system parameters, impulse action, and the combined switching law is established. Then easy-checked exponential stability criteria are presented based on the proposed hybrid switching strategy. The results demonstrate that the use of hybrid switching strategy significantly reduces switching frequencies and saves control costs while ensuring system stability. In future work, we will consider to explore the application of this novel and efficient switching strategy to study the finite-time stability of hybrid systems, particularly in the presence of inevitable external uncertainties in practical environments.}, archive = {J_TASE}, author = {Dan Yang and Xiaodi Li and Guofeng Zhang and Heung Wing Joseph Lee}, doi = {10.1109/TASE.2025.3541186}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12093--12101}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {State-dependent switching control for nonlinear impulsive switched systems with dwell time}, volume = {22}, year = {2025}, }
• (2025). Adaptive event-triggered fault detection for interval type-2 fuzzy markov jump systems under fading channels. TASE, 22, 12082--12092. (www)

  @article{ , abstract = {This paper focuses on the asynchronous fault detection (FD) problem for interval type-2 fuzzy Markov jump systems under bandwidth constraints and fading channels. A novel adaptive event-triggered scheme (AETS) is introduced, which not only saves communication resources but also guarantees a minimum data transfer rate. Based on the released date under the trigger protocol, a fuzzy asynchronous FD filter is designed, where the network-induced transmission delays and fading measurements are both considered. Then, using the residual signals generated by the filter, a generalized detection strategy for actuator faults is proposed. By applying the information about membership functions and the Lyapunov functional approach, the design conditions of the filter are obtained that ensure the stochastic stability and strict dissipativity of the dynamic error system. Compared with the existing methods, the proposed AETS is more capable of reducing the data release rate and avoiding some maltriggering events aroused by disturbance and noise. Finally, a simulation example is provided to validate the effectiveness of the proposed approach. Note to Practitioners—Fault detection has a crucial role in the modern production environment, which provides solutions for equipment health monitoring and fault prevention and has a positive impact on the efficiency and safety of industrial production. Considering the bandwidth constraints and fading channels in practical engineering, an adaptive event-triggered-based fault detection method is proposed in this paper. The proposed event-triggered strategy can effectively save communication resources on the premise of ensuring system performance. Besides, the delay information and fading measurements are fully utilized in the design of the filter, which is more suitable for practical scenarios. The proposed fault detection method can provide operators with early warning of faults for abnormal conditions to avoid more serious damages.}, archive = {J_TASE}, author = {Li-Xiang Feng and Guang-Hong Yang}, doi = {10.1109/TASE.2025.3542397}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12082--12092}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive event-triggered fault detection for interval type-2 fuzzy markov jump systems under fading channels}, volume = {22}, year = {2025}, }
• (2025). Event-triggered impulsive synchronization for neural networks subject to deception attacks: Saturation control and optimization. TASE, 22, 12071--12081. (www)

  @article{ , abstract = {This paper is concerned with the mean-square synchronization issue of the coupled neural networks (CNNs) subject to deception attacks, where the saturation constraint on impulsive control signal is adequately taken into account. To compensate the effect of deception attacks, a hybrid event-triggered mechanism is developed and a minimum inter-event interval is introduced simultaneously to avoid the Zeno phenomena. Under deception attacks, sufficient conditions to guarantee the achievement of the mean-square exponentially synchronization for the CNNs are presented with a novel method combined with mathematical induction, polyhedral representation of saturation nonlinearity, proof by contradiction, which address the actuator saturation in discrete-time control signals effectively. Meanwhile, in consideration of different impulsive effects, three optimization problems are constructed for the sake of acquiring the maximum estimation of the domain of attraction, the feasible maximal impulsive interval and the admissible feasible minimum impulsive interval, respectively. Finally, two numerical simulations are carried out to illustrate the validity of the proposed theoretical analysis. Note to Practitioners—In the engineering and industry, the synchronization control of CNNs exists in many different fields, including image processing, fluid dynamics, and secure communication. The control signals and plant information are transmitted over communication network which is vulnerable to deception attacks. To save communication resources and compensate the effect of deception attacks, a hybrid event-triggered scheme is proposed. Simultaneously, the synthesis of impulse effects and actuator saturation are considered to make results more practical. In addition, three optimization problems are proposed to derive the maximum estimation of the domain of attraction, some related parameters selection of event-triggered scheme.}, archive = {J_TASE}, author = {Wenjing Sun and Ze Tang and Jianwen Feng and Ju H. Park}, doi = {10.1109/TASE.2025.3540772}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12071--12081}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered impulsive synchronization for neural networks subject to deception attacks: Saturation control and optimization}, volume = {22}, year = {2025}, }
• (2025). Application of robust fuzzy cooperative strategy in global consensus of stochastic multi-agent systems. TASE, 22, 12058--12070. (www)

  @article{ , abstract = {This investigation introduces a sophisticated robust fuzzy distributed protocol, which synergistically merges the strengths of robust control and fuzzy control to confront the global consensus conundrum in unknown multi-agent systems. The technological ingenuity of this protocol lies in its integration of a seamless switching function, which ensures the robust and effective functionality of the fuzzy protocol across a broad global spectrum. Furthermore, the study delves into the global consensus dilemma in both first-order and second-order stochastic unknown multi-agent systems, outlining the specific design framework for robust fuzzy controllers. To uphold the stability of the closed-loop systems, the investigation innovatively formulates a novel type of Lyapunov function, inspired by the tenets of Lyapunov quadratic form design. Conclusively, through a series of simulation experiments, the investigation substantiates the practical effectiveness of the proposed algorithms. Note to Practitioners—Practitioners in automation, robotics, and distributed decision-making face a significant challenge in achieving global consensus in multi-agent systems amidst uncertainties and disturbances. This research introduces a sophisticated robust fuzzy distributed protocol that integrates robust control and fuzzy control, leveraging a switching function to ensure effectiveness across various scenarios. The study provides a detailed design framework for robust fuzzy controllers in first- and second-order stochastic unknown multi-agent systems, crucial for developing resilient strategies to maintain system stability and performance. Innovatively, a novel Lyapunov function, inspired by Lyapunov quadratic form design, upholds closed-loop system stability, offering a theoretical foundation for control strategies. Simulation experiments confirm the protocol’s practical effectiveness, achieving high-efficiency and reliable global consensus in unknown MASs. Preliminary results suggest promising practical implementation, benefiting practitioners across various fields.}, archive = {J_TASE}, author = {Jiaxi Chen and Jitao Shen and Weisheng Chen and Junmin Li and Shuai Zhang}, doi = {10.1109/TASE.2025.3539624}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12058--12070}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Application of robust fuzzy cooperative strategy in global consensus of stochastic multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Performance-constraint-based adaptive fuzzy prescribed-time containment control for nonlinear multi-agent systems. TASE, 22, 12047--12057. (www)

  @article{ , abstract = {In this article, a fuzzy adaptive prescribed-time performance constrain containment control issue is researched for nonlinear multi-agent systems (MASs). The original system’s prescribed time control issue will be converted into the corresponding controlled system’s asymptotic tracking control issue by using the time-domain mapping technique. Fuzzy logic systems (FLSs) are ultilized to identify the unknown dynamics. Combining prescribed-performance control and time-domain mapping techniques, a fuzzy adaptive prescribed-time performance constraint containment control algorithm is proposed, which can demonstrate that all signals in controlled system are bound within prescribed-time, and followers can converge to the convex hull created by the leaders by using Lyapunov stability theory. Furthermore, the virtual errors and local containment error asymptotically converge to a preset region. Finally, the simulations are provided to illustrate the effectiveness and feasibility of the developed control method. Note to Practitioners—This article investigates a fuzzy adaptive prescribed-time containment control issue for nonlinear multi-agent systems (MASs). The time-domain mapping technique can be used to convert the prescribed time control issue of the original system in finite horizon into the asymptotic tracking control issue of the corresponding controlled system in infinite horizon. In contrast to previous similar works, this article proposes a prescribed-time control algorithm, where the convergence time can be set offline and it does not rely on any design parameters. Moreover, this study can guarantee containment and virtual errors converge to the prescribed region by using the prescribed performance function and the barrier Lyapunov function.}, archive = {J_TASE}, author = {Kewen Li and Xiao Liu and Yongming Li}, doi = {10.1109/TASE.2025.3541253}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12047--12057}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Performance-constraint-based adaptive fuzzy prescribed-time containment control for nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Energy saving design of gear hobbing machine based on analytical target cascading: Modeling, decomposition, and independent optimization. TASE, 22, 12033--12046. (www)

  @article{ , abstract = {Design parameters optimization is regarded as an effective way to reduce energy consumption of a machine tool, which are contributed by several subsystems with coupling relationships affecting the total energy consumption. Therefore, it is of importance to study the coupling relationships between subsystems at the design stage for energy saving. However, there are multiple key energy consuming subsystems associated with various design parameters, leading to a complex optimization model of this problem. It is rather time-consuming to optimize such a complex model and hard to obtain global optimal results. To address this challenging problem, this article proposes an energy saving design approach of a gear hobbing machine based on Analytical Target Cascading (ATC) method. Firstly, the energy consumption models of each subsystem are established and the coupling relationships are analyzed. Then, the decomposition of the energy coupling model is conducted and the comprehensive optimization model is constructed by ATC. One novel adaptive sequential sampling method is proposed with the surrogate model to achieve the collaborative optimization of both structural parameters and control parameters. Finally, three optimization algorithms are developed to comparatively study the energy-optimal schemes. Testing results indicate that the energy consumption can be reduced by 9.72%, with lighter moving component and reduced deformation.Note to Practitioners—This article provides one method for the energy saving of the gear hobbing machine at the design stage. Previous studies only focus on the influence of one design factor on energy consumption, and the optimization is under a mixed optimization model, which is hard to obtain global optimal results. This article develops an energy saving design method considering the comprehensive influence of design factors for energy saving of multiple subsystems. Based on ATC, an energy coupling model can be divided into several energy consuming subsystems and the independent optimization of each subsystem can be achieved. To ensure the lightweight and control abilities, a surrogate model and a novel sequential sampling method are applied. Results show the effectiveness of the proposed method and the energy reduction can be achieved with a good structural and control performance.}, archive = {J_TASE}, author = {Shaoqing Wu and Congbo Li and Yan Jin and Xikun Zhao and Jinwen Zhang}, doi = {10.1109/TASE.2025.3541175}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12033--12046}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy saving design of gear hobbing machine based on analytical target cascading: Modeling, decomposition, and independent optimization}, volume = {22}, year = {2025}, }
• (2025). State estimation of a spatial 2-D linear diffusion process with mobile sensors. TASE, 22, 12024--12032. (www)

  @article{ , abstract = {This paper studies the state observer design of a spatial two-dimensional (2-D) linear diffusion process described by a linear parabolic partial differential equation (PDE) under mobile sensors. Firstly, we analyze the well-posedness of the PDE system and give the structure form of the state observer with mobile sensors. Subsequently, according to the number of mobile sensors, the 2-D space domain is divided into multiple sub-domains, and the mobile sensors are guided by the projection operator method, which can guarantee that the mobile sensors can only move in their respective 2-D sub-domains. Then, in the light of Lyapunov theory, Poincaré-Wirtinger inequality and Barbalat lemma, we propose an observation-plus-guidance design method to ensure the asymptotic stability of the state estimation error system. In the designed mobile strategy, the actual guidance of mobile sensors is essentially a physical synthesis of two direction guidance laws, where two dimensional guidance laws are designed separately. Moreover, the existence condition of the observer is given by linear matrix inequalities. At last, a numerical example is provided to demonstrate the efficacy of the proposed design scheme. Note to Practitioners—For the actual physical temporal-space process, the spatial 2-D case makes more sense. With the increase of space dimension, the difficulty of system design increases sharply. The observer design methodology and mobile sensor guidance approach developed for spatial 1-D systems are not directly extendable to 2-D systems. However, there have also been few reports on the observer design of 2-D physical temporal-space process using mobile sensors, and it is still a challenging problem. In this article, this study addresses the problem of designing state observer for 2-D linear parabolic PDE systems utilizing mobile sensors. To prevent collisions among mobile sensors, the 2-D spatial domain is partitioned into several subregions, and the projection operator approach is employed to develop the motion strategy. Then, a Lyapunov-based observation-plus-guidance design method is provided to achieve the desired design goals. At last, the finite difference method is used to verify the design method.}, archive = {J_TASE}, author = {Xiao-Wei Zhang and Kui Chen and Huai-Ning Wu and Jin-Liang Wang and Zi-Peng Wang}, doi = {10.1109/TASE.2025.3540855}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12024--12032}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {State estimation of a spatial 2-D linear diffusion process with mobile sensors}, volume = {22}, year = {2025}, }
• (2025). GelStereo tip: A spherical fingertip visuotactile sensor for multi-finger screwing manipulation. TASE, 22, 12012--12023. (www)

  @article{ , abstract = {Dexterous hands are the key element for robots to achieve human-like manipulation capabilities. An outstanding challenge is to provide fingertips of dexterous hands with precise tactile deformation sensing capabilities. In this paper, we present the GelStereo Tip, a spherical and easy-to-integrate GelStereo-type visuotactile sensor capable of sensing high-resolution 3D elastomer deformation. Previous calibration method does not take into account the impact of imaging errors caused by the sensor’s compact and high-curvature structural characteristics on the accuracy of tactile sensing. Therefore, we propose a novel self-calibration method based on the Refractive Stereo Ray Tracing model, named GTSC, and demonstrate the accuracy of less than 0.3 mm for deformation sensing. Furthermore, we also propose a Contact Retention Tactile Controller to address the issue of fingertips being unable to overcome obstructive torque during the multi-finger bottle cap screwing. After integrating GelStereo Tip into fingertips of Allegro Hand, the controller adjusts the joint positions of the given trajectory using proportional control based on the difference between the sensor’s actual deformation and the reference state for contact retention. We believe that the GelStereo Tip sensor combined with robotic dexterous hands has great application potential in the field of multi-finger fingertip manipulation. Note to Practitioners—The motivation of this paper is to design a fingertip visuotactile sensor with high-precision 3D tactile deformation sensing capabilities for multi-finger robotic hands and to validate its sensing performance. Additionally, it aims to address the issue of overcoming resistance in multi-finger screwing manipulations. Currently, most sensors do not consider the refraction effect or ignore the impact of planar imaging errors in refractive calibration. This paper proposes a visuotactile sensor along with a corresponding self-calibration method to ensure its sensing accuracy. Experiments show that our sensor possesses high-precision and robust 3D deformation sensing capabilities. On the other hand, multi-finger hands often struggle to complete screwing tasks along the given trajectory due to disturbances from torque resistance. This paper proposes a tactile controller that evaluates the contact state through aforementioned tactile sensing to improve subsequent trajectory and achieve continuous screwing. Comparative experiments highlight the necessity of this controller and the reliability of tactile sensing. We hope that the design of our sensor, the self-calibration method, and the tactile controller applied to multi-finger screwing can provide new insights for other practitioners.}, archive = {J_TASE}, author = {Boyue Zhang and Shaowei Cui and Chaofan Zhang and Jingyi Hu and Rui Wang and Shuo Wang}, doi = {10.1109/TASE.2025.3540913}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {12012--12023}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GelStereo tip: A spherical fingertip visuotactile sensor for multi-finger screwing manipulation}, volume = {22}, year = {2025}, }
• (2025). Self-prior guided spatial and fourier transformer for nighttime flare removal. TASE, 22, 11996--12011. (www)

  @article{ , abstract = {When capturing scenes with intense light sources, extensive flare artifacts often obscure the background and degrade image quality. Most flare removal methods directly process the flare-corrupted image as the optimization target, limiting the model’s understanding and generalization in complex real-world scenarios. In this paper, we propose a novel Self-prior Guided Spatial and Fourier Transformer (SGSFT) for nighttime flare removal. Specifically, we first establish a Self-prior Extraction Network to capture inherent priors in different scenes. Subsequently, we introduce a Semantic Contrast Enhancement Strategy to reinforce the semantic irrelevance between flare and light source, enabling the flare removal network to learn pattern differences between them and thus preserve light source. Finally, we build a Spatial and Frequency Flare Removal Network with Spatial Contextual Attention Block (SCAB) and Frequency Global Information Adjustment Block (FGIAB) to generate flare-free image. SCAB can perceive rich contextual information from self-prior guided regions and infer reasonable content. FGIAB captures global luminance representation in the frequency domain to maintain luminance consistency between the inferred regions and the flare-free areas. Extensive experiments demonstrate that the proposed approach achieves optimal performance in real nighttime scenes and exhibits robust generalization across various flare scenarios captured by different electronic devices. Note to Practitioners—The motivation of this paper is to remove flare artifacts in imaging. Flares degrade image quality and impact the performance of advanced vision tasks such as semantic segmentation and depth estimation in autonomous driving. Existing methods that indiscriminately extract contextual information from the entire image limit the model’s understanding of flares. This study proposes a self-prior guided flare removal network. The network first extracts self-prior information from flare-damaged images, then aggregates non-local information from the context indicated by the self-prior information to remove flares and infer semantically plausible fill content. Additionally, we model the global luminance information of the image in the frequency domain to enhance the global luminance consistency of the flare-free image. Experimental results show that our method has strong flare removal capabilities, but it also has a limitation. The training phase of this method requires paired flare-damaged images and flare images, which are difficult to obtain in real-world scenarios. Therefore, we will explore unsupervised flare removal methods in the future.}, archive = {J_TASE}, author = {Tianlei Ma and Zhiqiang Kai and Xikui Miao and Jing Liang and Jinzhu Peng and Yaonan Wang and Hao Wang and Xinhao Liu}, doi = {10.1109/TASE.2025.3539632}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11996--12011}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-prior guided spatial and fourier transformer for nighttime flare removal}, volume = {22}, year = {2025}, }
• (2025). Omniveyor: An assembled logistics sorting system powered by reinforcement learning. TASE, 22, 11982--11995. (www)

  @article{ , abstract = {To improve logistics efficiency, smart logistics sorting is an inevitable trend in logistics development. Existing smart logistics sorting systems suffer from high construction costs or limited scalability. To solve these problems, we design a brand-new two-dimensional conveyor system called Omniveyor, which transports and sorts high-density packages within a limited space. It is assembled from multiple repetitive square conveyor modules, achieving the goal of cost-effectiveness and easy maintenance. To realize automatic sorting, we model the planning problem on Omniveyor and propose a scheduling strategy named MMPPO by reinforcement learning. Unlike traditional path planning, MMPPO assigns actions to modules rather than packages, which reduces scheduling overhead in high-throughput scenarios. Furthermore, we develop a simulation environment to inspect the effectiveness of our method, which solves an intractable package-following problem that has plagued simulation implementation in this field. Experimental results show that MMPPO outperforms baselines in terms of throughput and overall consumption at high densities. Besides, we implement a physical prototype of Omniveyor to validate its feasibility. Note to Practitioners—The motivation of the paper is to solve the sorting problems in the logistics system. Existing logistics sorting systems often have problems such as low sorting efficiency and high costs, making them unsuitable for small-sized warehouses. In this paper, we present a modular 2D desktop logistics system that is both scalable and efficient for sorting. We mathematically describe the platform package transportation process and propose an algorithm that addresses scheduling and planning problems, capable of continuous planning under pipeline input. Preliminary simulation tests indicate that our approach is feasible, and we have also built a small-scale prototype. In future research, we will further expand the scale of the platform and conduct research.}, archive = {J_TASE}, author = {Mingrui Yin and Hao Zhang and Chenxin Cai and Meiyan Liang and Jie Liu}, doi = {10.1109/TASE.2025.3540511}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11982--11995}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Omniveyor: An assembled logistics sorting system powered by reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Lightweight approach for nonlinear state-space system identification subjected to skewed measurement noise. TASE, 22, 11968--11981. (www)

  @article{ , abstract = {In this paper, the skewed output noise is considered and we propose a lightweight robust algorithm for nonlinear state-space system identification based on the generalized hyperbolic variance gamma (GHVG) distribution, which enhances the robustness of the proposed algorithm. To facilitate the realization of the proposed algorithm, the hidden variables are introduced to decompose the GHVG distribution into the Gaussian gamma mixture (GGM) distribution, which improves the computational efficiency of the proposed algorithm. The expectation maximization (EM) and the particle smoothing (PS) approaches are combined to solve the hidden variables and unknown states problems, which contributes to derive the estimation formulas of the model parameters and noise parameters simultaneously. To further reduce the computational burden of PS method for estimating the nonlinear states, a novel nearest neighbor idea is used in the identification process which ensures the performance of the proposed algorithm while reducing the number of particles involved in the calculation of the cost function. Finally, the verification results are fairly carried out to demonstrate the effectiveness of the proposed algorithm. Note to Practitioners—The nonlinear state-space model (NSSM) is widely applied in industrial processes due to its ability to characterize the internal dynamics of the system. For probability-based identification of NSSM, the Gaussian or heavy-tailed distributions with symmetric statistical properties are commonly used, which may limit their application in practical industrial settings. Motivated by this problem, this paper proposes a lightweight robust identification algorithm of NSSM based on GHVG distribution. An improved PS method based on nearest neighbor idea is also introduced to reduce the computational efficiency of the algorithm while ensuring its performance. This means that the proposed method is not only robust to the outliers and the skewed output noise, but also greatly reduces the running time of the algorithm, which can extend its application in real industrial processes.}, archive = {J_TASE}, author = {Xin Liu and Yang Hai and Wei Dai}, doi = {10.1109/TASE.2025.3534849}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11968--11981}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lightweight approach for nonlinear state-space system identification subjected to skewed measurement noise}, volume = {22}, year = {2025}, }
• (2025). A practical micropipette-image calibration method for somatic cell microinjection. TASE, 22, 11955--11967. (www)

  @article{ , abstract = {This paper proposes a practical micropipette-image 2-D pixel and 3-D spatial coordination) calibration method indispensable to somatic cell microinjection, leveraging advancements in modern motorized micromanipulators. The method determines the depth information of the micropipette in the microscope field by assessing the contact between the micropipette tip and the bottom of the culture dish. It uses recoverable deformation upon contact as a criterion for precise positioning, ensuring the tip is on the dish’s bottom surface and in the microscope’s focus plane. Additionally, the paper introduces an on-the-spot method for breaking a micropipette tip and a preprocessing technique for somatic cells. The proposed micropipette tip-breaking method, using a low-cost acrylic ring, overcomes previous drawbacks and proves quick and user-friendly. The preprocessing technique converts fully adherent somatic cells into semi-adherent cells, increasing cell thickness for easier puncturing. Combining these techniques, the study validated the approaches through over 900 injections on human dermal fibroblast (HDF) cells, achieving a success rate of 53.3% and a survival rate of 95.8%. Note to Practitioners–This study was initiated to address the challenge of establishing accurate mapping relationships between observed 2-D coordinates of a micropipette tip in microscope vision and 3-D positional data obtained from an external controller of an XYZ micromanipulator. However, its relevance extends beyond somatic cell microinjection to other micromanipulation applications like micro-assembly. Common calibration methods involve motorized objectives or stereo microscopic vision, whereas our paper proposes a practical approach utilizing recoverable deformation upon tip contact with the dish bottom for precise positioning. The paper mathematically characterizes relationships between the pixel coordinate system and spatial coordinate systems, demonstrating problem-solving by ensuring the tip aligns with the dish’s bottom surface and the microscope’s focus plane. Although experiments affirm feasibility, refinement and testing in other micromanipulation applications are pending. Future research aims to enhance system modeling and calibration accuracy.}, archive = {J_TASE}, author = {Fei Pan and Shuxun Chen and Liushuai Zheng and Shaohua Zhi and Xi Chen and Dong Sun}, doi = {10.1109/TASE.2025.3533657}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11955--11967}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A practical micropipette-image calibration method for somatic cell microinjection}, volume = {22}, year = {2025}, }
• (2025). Optimal tracking control for cyber-physical systems under mixed attacks via game-theoretical Q-learning. TASE, 22, 11944--11954. (www)

  @article{ , abstract = {This paper investigates the optimal tracking control problem for cyber-physical systems (CPSs) under mixed attacks based on the Stackelberg game strategy. First, an improved value function is designed to meet performance criteria by considering the control signal, mixed attack signal, and tracking performance. Then, based on the Stackelberg game theory and the principle of optimality, the optimal control strategy and false data injection attack policy are derived by solving a coupled algebraic Riccati equation (CARE). The proposed control strategy can effectively alleviate the adverse impact of mixed attacks on the control performance of CPSs. Subsequently, sufficient conditions are provided to guarantee the existence of the solution to the CARE. Additionally, an improved Q-learning algorithm is proposed to learn the optimal control scheme through state reconstruction, which avoids the need for access to state vectors and facilitates data-based controller design. Using the Lyapunov stability theory, it is demonstrated that the presented algorithms are convergent and the output of CPSs can track the reference trajectory. Finally, the proposed approach is validated by numerical simulations. Note to Practitioners—In the engineering application scenarios, the sharing feature of network communication may expose the controlled system to malicious attacks (such as DoS attacks and FDI attacks). The majority of the current control methods focus on one-sided analyses. In this paper, the dynamic interaction between attackers and defenders is described using a Stackelberg game model. Within this model, an optimal tracking control algorithm is proposed to mitigate the impact of mixed attacks on CPSs. Moreover, sufficient conditions for tolerable probability of attack are derived, which enables practitioners to determine the conditions under which the attacked system’s stable tracking performance may still be maintained.}, archive = {J_TASE}, author = {Jinyan Li and Xiao-Meng Li and Guangdeng Chen and Xiao-Jie Peng and Hongyi Li}, doi = {10.1109/TASE.2025.3540401}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11944--11954}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal tracking control for cyber-physical systems under mixed attacks via game-theoretical Q-learning}, volume = {22}, year = {2025}, }
• (2025). SMC for networked non-homogeneous hidden semi-markov switching systems with cyber attacks and application to DC-DC buck converter circuit. TASE, 22, 11933--11943. (www)

  @article{ , abstract = {This study addresses the problem of sliding mode control (SMC) for networked non-homogeneous hidden semi-Markov switching systems under semi-Markov kernel and cyber attacks, in which the limited dwell time information is related to the probability density function. Firstly, the model of networked non-homogeneous hidden semi-Markov switching systems under cyber attacks is constructed. Considering that it is difficult to obtain all the real modes in actual systems, a hidden semi-Markov chain is utilised to determine the hidden mode switching of the underlying system, which is more realistic than semi-Markov chain. Based on restricted dwell time probability density function, the common assumption is relaxed under completely known probability density function. The main innovation is to build a suitable SMC scheme under denial-of-service attacks to achieve quasi-sliding mode that eliminates the effect of uncertain parameters. By means of the Lyapunov function depending on the system mode and the elapsed time, stability criteria are considered for the corresponding model. Finally, a DC-DC buck converter circuit model is introduced to validate the practicality of the proposed strategy. Note to Practitioners—Note that the control community has witnessed a tremendous development in networked control systems with wide applications in many practical models, enabling remote control in a sensitive manner through communication networks. Although networked control systems offer obvious advantages, they always suffer from particular difficulties associated with network transmission, such as cyber attacks and packet losses. With the rapidly developing network and communication technologies, networked non-homogeneous hidden semi-Markov switching systems have attracted much attention due to strong capabilities in modeling dynamical systems with abrupt changes in structures or parameters. It is worth emphasising that it is difficult to obtain all real mode information of networked semi-Markov switching systems. This paper offers a novel approach for researchers to investigate the SMC for networked non-homogeneous hidden semi-Markov switching systems under semi-Markov kernel and cyber attacks.}, archive = {J_TASE}, author = {Wenhai Qi and Feiyue Shen and Ju H. Park and Zheng-Guang Wu and Huaicheng Yan}, doi = {10.1109/TASE.2025.3540807}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11933--11943}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SMC for networked non-homogeneous hidden semi-markov switching systems with cyber attacks and application to DC-DC buck converter circuit}, volume = {22}, year = {2025}, }
• (2025). Event-triggered super-twisting fixed-time consensus control for networked nonlinear multi-agent systems with disturbance. TASE, 22, 11920--11932. (www)

  @article{ , abstract = {In this paper, the leader-follower fixed-time consensus problem of networked nonlinear multi-agent systems (NNMASs) with unknown disturbance is investigated. A new event-triggered super-twisting fixed-time sliding mode consensus control (ESFSMCC) method is proposed, enabling all agents to achieve consensus in a fixed time. Firstly, a terminal sliding mode variable is designed to eliminate the convergence time dependence on the initial values of the system and avoid singularity. Secondly, a distributed event-triggered mechanism is developed to effectively reduce inter-agent communication frequency in a networked environment. In contrast to the existing fixed-time consensus control method, the improved super-twisting algorithm is adopted to construct a control protocol to achieve both global fixed-time consensus robustness and alleviate the issue of high-frequency chattering. Thirdly, the Lyapunov theory is utilized without the piecewise sliding mode technique to derive sufficient conditions for establishing the fixed-time stability of NNMASs, which still avoids the singularity problem. In this paper, the non-segmented terminal sliding mode is employed to prove the global fixed-time stability of the system states, thereby avoiding computational complexity. Finally, the effectiveness and advantages of the proposed method are verified through numerical simulations. Note to Practitioners—With the flourishing development of networks today, networked control is poised to become a future research hotspot, especially when large-scale agents interact and collaborate. The efficient utilization of network resources has become increasingly urgent due to the proliferation of such interactions. Consequently, reducing energy consumption poses a significant challenge. Moreover, the high-frequency chattering of control inputs presents a hindrance to the practical application of SMC. To address these issue, this paper proposes an event-triggered super-twisting distributed control protocol for NNMASs. This protocol not only eliminates the influence of initial values on stability time, thereby enhancing its practical value, but also mitigates the impact of input chattering, providing robust support for practical applications. Finally, a dynamics model of a multi-robotic manipulator is employed to verify the correctness and effectiveness of the proposed method. Additionally, underwater autonomous vehicle formations are expected to become valuable tools for future ocean resource exploration, while drone formations will likely become the preferred choice for agricultural development, geological exploration, and transportation. Even space exploration will involve coordinated interaction between multiple spacecraft. These are typical applications of NNMASs.}, archive = {J_TASE}, author = {Guangxin Guo and Haoran Tan and Yun Feng and Yaonan Wang}, doi = {10.1109/TASE.2025.3540503}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11920--11932}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered super-twisting fixed-time consensus control for networked nonlinear multi-agent systems with disturbance}, volume = {22}, year = {2025}, }
• (2025). Event-driven control of hybrid systems using batches petri nets: Application to high throughput manufacturing systems. TASE, 22, 11906--11919. (www)

  @article{ , abstract = {The problem considered in this work concerns the control of hybrid systems, studied from the discrete event systems theory viewpoint. The objective is to compute a transient control trajectory for reaching a steady state from a given initial state. The hybrid systems are modeled by a class of hybrid Petri nets called controlled generalized batches Petri nets, which introduce variable delays on continuous flows for modeling and analysis of dynamical systems. The key contributions of this study are as follows: two event-driven control strategies are proposed for reaching a target steady state, offering advantages in terms of control actions and time performance. These strategies are based on ON/OFF state transitions, with the firing flow either limited to its steady flow or maximal flow; two algorithms are presented to implement these control strategies effectively; a case study of a bottling production line, a high throughput manufacturing system, is provided, demonstrating the application of the proposed methods and evaluating their efficiency in terms of time performance and the number of generated events. Note to Practitioners—In this work, two event-driven control strategies for reaching a steady state from a given initial state are presented. We use controlled generalized batches Petri nets that are suitable discrete event formalism for the modeling of complex systems such as transportation systems, high throughput production lines and, crowds behavior. This Petri net could model groups of moving entities with given forms (speed, density, length), which allows to consider variable delays of transfer elements due to congestion or groups merging. The objective is to reach a target steady state characterized by constant flows and also particular forms of the entities groups. This is a challenging issue because the existing formalisms and methods usually allow to reach a given steady setting without considering the forms of the entities groups. The proposed control strategies could be used, for example, during a recovery mode in order to reach the steady state after a system’s shutdown, failure or malicious attack. Their effectiveness is shown on the Perrier mineral water bottling production line.}, archive = {J_TASE}, author = {Ruotian Liu and Rabah Ammour and Leonardo Brenner and Isabel Demongodin}, doi = {10.1109/TASE.2025.3540744}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11906--11919}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-driven control of hybrid systems using batches petri nets: Application to high throughput manufacturing systems}, volume = {22}, year = {2025}, }
• (2025). Prediction-based method for micropipette approaching to optimal spindle removal position. TASE, 22, 11894--11905. (www)

  @article{ , abstract = {During somatic cell nuclear transfer (SCNT), precise removal of the oocyte genetic material is critical. However, due to the invisibility of the genetic material (spindle) under brightfield observation and the potential displacement caused by the movement of micropipettes, accurately positioning the micropipette at the spindle poses a significant challenge. This study introduces an approach for optimal spindle removal by predicting its position. Initially, the polarization imaging system visualized the oocyte spindle, while a Multi-Feature Adaptive Kernel Correlation Filter (MFAKCF) algorithm tracked the spindle with 92.84% accuracy. Subsequently, enhancements were made to the Nonlinear Mass-Spring-Damper (NMSD) model to simulate live oocyte mechanical characteristics. Adjustments to NMSD model parameters simulated spindle displacement variations under diverse experimental conditions. Finally, the optimal spindle removal position was determined using NMSD model to simulate the micropipette approach to the spindle and the resulting position of spindle displacement. Experimental validation showed that the predictive accuracy of this model was 97.26%, with an average positional error of $0.4~\mu $ m. Using this approach method can reduce cytoplasm loss to 4.5% and have a 100% enucleation success rate. Thus, the proposed prediction based optimal spindle removal position method can effectively anticipates spindle final positions, aiding in minimizing cytoplasmic loss during spindle removal. Note to Practitioners—Accurate oocyte nucleus removal is crucial for somatic cell nuclear transfer (SCNT) but challenging due to spindle invisibility under brightfield microscopy and displacement caused by micropipette movements. This study proposes a predictive method to address these issues. The polarization imaging system was used to visualize the spindle, and a MFAKCF tracking algorithm enabled real-time tracking. An enhanced NMSD model simulated spindle displacement, enabling accurate position prediction and reducing cytoplasmic loss during removal. This method offers a reliable tool for precise spindle removal in SCNT and broader applications in biomedical micromanipulation.}, archive = {J_TASE}, author = {Zuqi Wang and Yujie Zhang and Yidi Zhang and Detian Zhang and Zhaotong Chu and Qili Zhao and Mingzhu Sun and Maosheng Cui and Xin Zhao and Yaowei Liu}, doi = {10.1109/TASE.2025.3536036}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11894--11905}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prediction-based method for micropipette approaching to optimal spindle removal position}, volume = {22}, year = {2025}, }
• (2025). Fault diagnosis of labeled petri nets under attacks using integer linear programming. TASE, 22, 11881--11893. (www)

  @article{ , abstract = {This paper deals with the online fault diagnosis problem of discrete event systems under malicious external attacks. We consider a scenario where an attacker can intercept certain sensor measurements and alter them arbitrarily, potentially causing a diagnoser to malfunction. In the framework of labeled Petri nets, a novel integer linear programming problem is formulated by introducing binary variables to estimate the possible transition sequences of an observation that may have been tampered with by an attacker. The proposed approach makes two main contributions. The first one is that, by specifying two different objective functions to the integer linear programming problem, we can obtain the diagnosis results in the presence of attacks, which classic diagnosers may fail to achieve; the second is computational efficiency. In the absence of attacks, the proposed approach is experimentally verified to have lower computational overhead compared with the existing results that are based on integer linear programming and those using basis markings. Finally, the proposed approach is illustrated through a manufacturing system for assembling brake valves. Note to Practitioners—Fault diagnosis is critical for highly automated systems such as manufacturing systems, power plants and smart grids. Engineers are familiar with various fault diagnosis techniques in their community. However, malicious attacks in a system are not fully considered when performing fault diagnosis. This research elaborates upon a novel approach to fault diagnosis of discrete event systems modeled by labeled Petri nets, which can perform online fault diagnosis both in the presence and absence of attacks. The approach enables faults to be detected even after certain labels are tampered with by an attacker. The proposed diagnostic vehicle is presented in terms of integer linear programming problems, which facilitates its applications to real systems by practitioners.}, archive = {J_TASE}, author = {Tengbo Li and Huorong Ren and Ruotian Liu and Maria Pia Fanti and Zhiwu Li}, doi = {10.1109/TASE.2025.3533311}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11881--11893}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault diagnosis of labeled petri nets under attacks using integer linear programming}, volume = {22}, year = {2025}, }
• (2025). Three-fingered grasp detection in clutter with a unified grasping representation. TASE, 22, 11869--11880. (www)

  @article{ , abstract = {Robust grasp pose detection is crucial for robotic manipulation. Antipodal grasps are often inadequate for spatial objects. However, insufficient demonstrations and the higher complexity compared to parallel grippers make it challenging to implement learning-based robotic grasping with multi-fingered hands. To address these deficiencies, we propose a multi-modal grasp detection method tailored for three-fingered grippers. We utilize a simplified unified grasp representation, which can represent both precision and power grasping modes. Additionally, we introduce a pipeline to generate a multi-modal grasp dataset with a three-point contact grasp metric. We also present the concept of confness to quantify the probabilities of different grasp configurations within a given graspable area, guiding the predicted grasp poses to be concentrated around on centroidal region of objects. Furthermore, we develop QcMNet, an end-to-end network that is used to predict confness and infer grasp parameters. Experiments conducted with a Robotiq 3-Finger Adaptive Gripper, both in simulation and real-world scenarios, demonstrate that multi-modal grasping can increase the percent-cleared rate and enhance anti-disturbance capabilities in cluttered scenes. Note to Practitioners—This paper was motivated by the problem of finding the proper gripper configurations to ensure the stable grasping of objects across diverse tasks in industrial automation. Most existing approaches to robotic grasping detection utilize parallel grippers. However, it is insufficient to achieve stable grasping for spatial objects only with two frictional point contacts. This paper presents a learning-based grasp pose detection method tailored for three-fingered grippers, enabling robots to achieve a stable grasp on objects. We first propose a unified grasp representation that can represent different grasp modes, including precision and power grasp modes. Subsequently, a densely annotated dataset for three-fingered grippers is obtained by employing a robust grasp metric and an auto-annotation pipeline in simulation. This approach is designed to reduce labor consumption. The proposed pipeline not only enhances the efficiency of the data collection process but also ensures its diversity. Furthermore, we introduced a neural network to quantify the probabilities of different grasp modes within a graspable area and to infer grasp poses. Extensive physical robotic experiments validate the feasibility of this approach for grasping various objects. Significantly, most grasp poses concentrate around the centroid of objects, enhancing grasp stability. In future research, we will involve further optimizing the generated grasping candidates using local geometry information of objects. Additionally, we intend to generate online collision-free trajectories for robotic grasping to adapt to more complex automation environments.}, archive = {J_TASE}, author = {Guangbao Zhao and Yan Xing and Jianhua Wu and Zhenhua Xiong}, doi = {10.1109/TASE.2025.3541026}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11869--11880}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Three-fingered grasp detection in clutter with a unified grasping representation}, volume = {22}, year = {2025}, }
• (2025). Multi-branch multi-scale channel fusion graph convolutional networks with transfer cost for robotic tactile recognition tasks. TASE, 22, 11856--11868. (www)

  @article{ , abstract = {Inadequate consideration of tactile sensor features in existing algorithms can lead to insufficient prediction accuracy for robotic tactile recognition tasks. In this paper, we propose a spatial-temporal information based adjacency matrix construction method called transfer cost. On this basis, we propose a multi-branch multi-scale channel fusion graph convolutional network (MSCF-LSTM-GCN). This network combines the advantages of long short-term memory networks (LSTM) with graph convolutional networks (GCN) to extract spatial-temporal features. It is the first LSTM-GCN network with a multi-branch structure that utilizes multi-feature scale and multi-channel fusion. This structure provides diverse perceptual ranges and information levels. We introduce shortcuts combined with the multi-branch structure to reduce the loss of information caused by multiple layers of transmission. The Einstein summation calculation method and the parameters of binarizing the adjacency matrix are optimized to enable the established dynamic tactile graph to participate in batch training. The proposed method improves both prediction accuracy and F1 score by over 1% on the tactile grasp stability prediction task. In the tactile object recognition task, the proposed method improves both prediction accuracy and F1 score by over 3%.Note to Practitioners—Robotic tactile perception is an important research area that can improve the flexibility, accuracy and stability of robots in robotic tactile recognition tasks. However, the amount of accessible information depends on the resolution and number of tactile sensors, and the features extracted using conventional deep learning methods are limited. For this reason, we propose a spatial-temporal information based adjacency matrix construction method called transfer cost. Then, we propose the MSCF-LSTM-GCN model that extract spatial-temporal features from different feature scales and shortcuts. The network solves the problems of traditional deep learning methods in acquiringsingle features and losing deep information. The experiments on two public datasets show that our method performs due to other competing algorithms on the tasks of robotic grasp stability prediction and object recognition for multi-tactile sensor scenarios. In future research we will invest in the problem of 3D force regression.}, archive = {J_TASE}, author = {Yupo Zhang and Xiaoyu Li and Senlin Fang and Yiwen Liu and Jingnan Wang and Bo Yuan and Zhengkun Yi}, doi = {10.1109/TASE.2025.3541339}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11856--11868}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-branch multi-scale channel fusion graph convolutional networks with transfer cost for robotic tactile recognition tasks}, volume = {22}, year = {2025}, }
• (2025). Self-supervised production anomaly detection and progress prediction based on high-streaming videos. TASE, 22, 11843--11855. (www)

  @article{ , abstract = {Real-time production monitoring incorporating progress prediction and anomaly detection is essential for quality and efficiency. Traditional vision-based anomaly detection methods struggle to differentiate between production-related features and background noise, and fail to consider the heterogeneity of production stages. This paper introduces an integrated approach that merges progress prediction and anomaly detection, employing the Autoencoder Process Probability Embedding (APPE) method. APPE maps the distribution of images from normal production to a progress-related Gaussian Mixture Model (GMM), focusing on identifying production-relevant features while minimizing background interference through the proposed Spatial Activation Map (SAM). The proposed SAM improves the interpretability of the neural network by highlighting the specific features that influence the model’s decisions. The method is assessed through real-world datasets in the assembly of water valves and the production of commercial aircraft spoilers. The case study shows that our approach can achieve superior effectiveness compared to the benchmark, notably improving both task performances by integrating progress prediction with anomaly detection. Note to Practitioners—In many manufacturing settings, such as aircraft production, tasks that involve human-robot collaboration or high-precision manual assembly play a significant role. The ability to detect anomalies and monitor progress in real-time is critical for ensuring the quality and efficiency of production. The manual nature of these operations makes them challenging to monitor through in-situ embedded digital sensors, yet real-time operation videos are readily available. Vision-based production monitoring has been widely used in applications such as product surface inspection, but existing algorithms often face difficulties distinguishing between normal background variations and anomalies related to production. This paper introduces a new approach, called Autoencoder Process Probability Embedding (APPE), which integrates progress recognition and anomaly detection into a cohesive monitoring task, allowing the model to differentiate between background elements and features related to production. Although our method is demonstrated in production scenarios as case studies, the proposed SAM mechanism is versatile to be applied in other contexts with similar types of categorical labels.}, archive = {J_TASE}, author = {Yifan Li and Zhi-Hai Zhang and Jiaqi Xu and Xiaowei Yue and Li Zheng}, doi = {10.1109/TASE.2025.3538328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11843--11855}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-supervised production anomaly detection and progress prediction based on high-streaming videos}, volume = {22}, year = {2025}, }
• (2025). FMRT: Learning accurate feature matching with reconciliatory transformer. TASE, 22, 11826--11842. (www)

  @article{ , abstract = {Local Feature Matching, a pivotal component of numerous computer vision tasks (e.g., structure from motion and visual localization), has been effectively addressed by Transformer-based methods. Nevertheless, these methods solely incorporate long-range context information among keypoints with a fixed receptive field, which constrains the network from appropriately reconciling the importance of features with diverse receptive fields to realize complete image perception, hence limiting feature matching accuracy. In addition, these methods employ a conventional handcrafted encoding approach to incorporate positional information of keypoints into visual descriptors, which limits the capability of networks to extract effective positional encoding message. In this study, we propose FMRT, a novel detector-free method that reconciles local features with diverse receptive fields adaptively and utilizes parallel networks to realize reliable positional encoding. Specifically, FMRT proposes a dedicated reconciliatory transformer (RecFormer) that contains a global perception attention layer to identify visual descriptors with different receptive fields and integrate global context information under various scales, a perception weight layer to measure the importance of various receptive fields adaptively, and a local perception feed-forward network to extract deep aggregated multi-scale local feature representation. Moreover, we introduce a novel axis-wise position encoder (AWPE) that views positional encoding as two keypoints encoding tasks along the row and column dimensions, decouples the x- and y-coordinates of keypoints into two independent 1D vectors, and designs two parallel network branches to explicitly encodes geometric correlations among keypoints, hence realizing reliable positional encoding. Extensive experiments indicate that FMRT yields impressive performance on multiple tasks, including relative pose estimation, visual localization, homography estimation, and image matching. Besides, we integrate FMRT into a localization framework and conduct a visual localization experiment in a real scene, which further demonstrate the superiority of FMRT. Note to Practitioners—This paper presents a novel approach to enhancing the performance of local feature matching in computer vision tasks. Traditional methods often rely on fixed receptive fields for integrating context among keypoints, which can limit the perception of the complete image and, consequently, the precision of feature matching. Our work introduces a Reconciliatory Transformer that not only addresses these limitations by effectively reconciling the importance of features across varying receptive fields but also improves the integration of positional information into visual descriptors. The techniques developed here can be adapted to a wide range of systems, e.g., image matching for computer vision and visual localization for autonomous driving, offering practitioners a tool to significantly improve the fidelity of feature matching, which is foundational for accurate interaction with the surrounding environment.}, archive = {J_TASE}, author = {Li Wang and Xinyu Zhang and Zhiqiang Jiang and Kun Dai and Tao Xie and Lei Yang and Wenhao Yu and Yang Shen and Bin Xu and Jun Li}, doi = {10.1109/TASE.2025.3538919}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11826--11842}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FMRT: Learning accurate feature matching with reconciliatory transformer}, volume = {22}, year = {2025}, }
• (2025). Trajectory tracking control of wheeled mobile manipulators with joint flexibility via virtual decomposition approach. TASE, 22, 11808--11825. (www)

  @article{ , abstract = {Wheeled mobile manipulators (WMMs) involving a wheeled mobile platform and a serial manipulator are finding increasing applications in diverse fields, creating new challenges in performing high-precision operations in a spacious workspace. WMMs are challenging to control due to uncertainties in system parameters, coupled dynamics, and external disturbances, which make stability guarantees difficult. This paper proposes a virtual decomposition control (VDC)-based trajectory tracking controller for WMMs, addressing joint flexibility, external disturbances, etc. The proposed method uses a VDC-based iterative approach to manage the complex coupled dynamics and employs a separate adaptive controller to handle joint flexibility. The robotic system’s stability is validated using the specific features of VDC (proof of each subsystem’s virtual stability) according to the Lyapunov stability theory. The advantages and effectiveness of the proposed method are demonstrated through experiments.Note to Practitioners—This paper addresses the challenges faced in controlling WMMs, which are becoming increasingly common in various industrial and service applications due to their ability to perform tasks in large and dynamic environments. The coupling between the wheeled platform and the manipulator, as well as uncertainties in system parameters such as joint flexibility and external disturbances, make precise trajectory tracking difficult. To address these challenges, this paper presents a control approach based on VDC, which breaks down the complex system into manageable subsystems and ensures stability for each part individually. The control strategy also incorporates adaptive control to handle joint flexibility and unpredictable disturbances. The stability of the system is rigorously proven through Lyapunov theory, ensuring robust performance under real-world conditions. Practitioners working on autonomous mobile robots equipped with manipulators may find this approach useful for improving trajectory tracking performance in uncertain and dynamic environments. However, the practical implementation of this method will require careful tuning of controller parameters and real-time computational capabilities to ensure seamless operation in real applications.}, archive = {J_TASE}, author = {Hongjun Xing and Yuzhe Xu and Liang Ding and Jinbao Chen and Haibo Gao and Mahdi Tavakoli}, doi = {10.1109/TASE.2025.3540860}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11808--11825}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory tracking control of wheeled mobile manipulators with joint flexibility via virtual decomposition approach}, volume = {22}, year = {2025}, }
• (2025). Toward micro lip reading in the wild: Dataset, theory, and practices. TASE, 22, 11794--11807. (www)

  @article{ , abstract = {Micro lip reading, is characterized by tiny lip movements when someone is speaking. It is highly preferred in robotics and automation, such as a social robot for geriatric care, a patrol robot system for hospitals, and speech system for hearing impaired individuals, etc. However, it has not been well studied before. In this paper, we shed the light to this research topic. A labelled micro lip reading dataset (i.e. HUST-LMLR) of 399 video samples is first established. The samples are captured from the unconstrained movies. One key challenge for micro lip reading in the wild lies on extracting fine features of lip movements effectively and robustly. In this paper, we pay the research efforts to address this issue from two aspects: facial context attention and feature extraction. First, we propose a multi-task learning model for micro lip reading in the wild for the first time. It concerns micro lip reading and facial landmark detection jointly for capturing global face context with soft attention, to facilitate micro lip reading. Second, we propose that motion feature should be concerned and combined with appearance feature jointly to characterize tiny lip movements effectively. Finally, the experiments on HUST-LMLR demonstrate the challenges of our dataset and our proposed approach results in a remarkable improvement of the state-of-the-art by over 26% WER in HUST-LMLR. Furthermore, results in a well-known public dataset LRS2 also show the generalization and superiority of our approach. If accepted, we will publish our HUST-LMLR dataset and relative supporting matierials at https://hust-dpkw.github.io/Micro-lip-reading/.Note to Practitioners—This paper aims to construct a challenging public dataset and propose a novel vision-based approach for automatically micro lip reading under the unconstrained conditions. A labelled dataset HUST-LMLR was first established to reveal the “micro” characteristics, with 399 video samples captured from 40 unconstrained movies or documentaries. In addition, two key manners in perspective of algorithm are proposed to address the problem of micro lip reading at sentence level in the wild: multi-task learning with facial landmark detection, feature extraction with appearance and motion combined. The extensive experiments demonstrate the challenges of our dataset HUST-LMLR as well as the superiority of our approach in micro lip reading tasks. Nevertheless, it is still somewhat sensitive to the extremely tiny variation of lip movements and the dramatic variation on human posture. It is worth noting that our proposition can be applied not only for medical support, smart health care but also for human-computer interaction, special education, information security, assistant driving, and virtual reality etc.}, archive = {J_TASE}, author = {Ran Wang and Yingsong Cheng and Anqi Cao and Shengtao Zhong and Xiang Liao and Yanjiu He and Han Ye and Chutian Sun and Qinlao Zhao and Min Du and Yang Xiao}, doi = {10.1109/TASE.2025.3539220}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11794--11807}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward micro lip reading in the wild: Dataset, theory, and practices}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy event-triggered optimized consensus control for delayed unknown stochastic nonlinear multi-agent systems using simplified ADP. TASE, 22, 11780--11793. (www)

  @article{ , abstract = {This paper investigates the adaptive fuzzy event-triggered optimized consensus tracking control problem for uncertain stochastic nonlinear multi-agent systems (MASs) with unknown dynamic and time-delay. Typically, optimal control is derived by the solution of the Hamilton-Jacobi–Bellman (HJB) equation, but it is usually challenging to solve this equation since inherent nonlinearity and unknown dynamics. Specifically, the complexity of the MASs in controller design is further exacerbated by the issue of state interdependence. To achieve optimized consensus control, the adaptive dynamic programming strategy is derived using the negative gradient of a simple positive function. As a result, the designed optimized consensus tracking control is relatively simple and can eliminate the persistence excitation (PE) assumption. The fuzzy logic systems (FLSs) are utilized to approximate the current and delayed states of unknown nonlinear functions; the identifier is proposed to estimate the stochastic multi-agent dynamic; the critic and actor FLSs are designed to evaluate control performance and execute control behavior, respectively. Furthermore, the event-triggered control (ETC) method is developed to save transmission load and communication resources. Moreover, we demonstrate that all signals for the MASs are semi-globally uniformly ultimately bounded (SGUUB) in mean square, and that the states of the follower agents can reach consensus with the leader’s state. Finally, a numerical example is illustrated to demonstrate the effectiveness of the proposed method. Note to Practitioners—With environmental protection and energy conservation becoming dominant trends, improving efficiency is regarded as a fundamental principle in control design. In addition, MASs are widely affected by factors such as stochastic disturbance, model unknown, time-delay, and uncertain factors, which further increase the difficulty of controller design. In this work, the optimized control approach is developed to design the control strategy for MASs to achieve the control task with the minimum utilization rate of control resources. In practical application scenarios such as spacecraft control or aircraft operation, the designed optimized control scheme can not only minimize fuel consumption to complete the predetermined tasks but also eliminate the PE assumption and strong assumptions on the time-delay nonlinear functions. Furthermore, the proposed ETC strategy effectively reduces the communication resource and network transmission in practice. Therefore, the control strategy designed in this work is more easily applicable to practical engineering.}, archive = {J_TASE}, author = {Guoping Zhang and Chengbin Liang and Quanxin Zhu}, doi = {10.1109/TASE.2025.3540468}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11780--11793}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy event-triggered optimized consensus control for delayed unknown stochastic nonlinear multi-agent systems using simplified ADP}, volume = {22}, year = {2025}, }
• (2025). SF-pose: Semantic-fusion six degrees of freedom object pose estimation via pyramid transformer for industrial scenarios. TASE, 22, 11767--11779. (www)

  @article{ , abstract = {Object six degrees of freedom (6-DoF) pose estimation is the powerful vision algorithm for the robot-environment interaction. However, current robust pose estimation algorithms rely heavily on labeled real data with high-cost collection, making it difficult to apply the algorithm. Many studies discuss the use of synthetic data as a complement to real datasets. However, reducing the gap between synthetic and real data is still a challenging problem. Based on the consistency of object geometric characteristics between real data and synthetic data, we argue that multi-input, rather than image-only input, is more suitable for transfer from synthetic to real, because it strengthens the extraction of object geometric feature. Therefore, we propose a semantic-fusion 6-DoF object pose estimation method that effectively capture common features across various resolutions by employing the designed pyramid transformer feature-fusion module. Extensive experiments show that the proposed method performs better than the state-of-the-art (SOTA), indicating that the proposed method can effectively extract and fuse different representations. Furthermore, in response to the lack of industrial scene datasets, we also develop a synthetic pose dataset and conduct the human-robot collaboration experiment to verify the robustness of the proposed method. Note to Practitioners—The purpose of this paper is to bridge the gap between synthetic and real data for pose estimation of industrial tools. Our method can be trained only on synthetic data and accurately estimate pose parameters in real scenes. Combining physically-based renderer and industrial tools, such as hammers and screwdrivers, a synthetic dataset of industrial scenes can be produced using the data production pipeline proposed in this paper. In this case, the trained model can assist the robot vision system to understand object pose information in a real production workshop. Extensive dataset experiments and human-robot collaboration experiments demonstrate the effectiveness of the proposed method. In addition, based on the actual robot working environment, practitioners can produce industrial datasets from multiple angles, objects, and scenes. Sufficient datasets can enhance the model’s generalization and robustness.}, archive = {J_TASE}, author = {Jikun Wang and Yinlong Liu and Zhi-Xin Yang}, doi = {10.1109/TASE.2025.3529511}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11767--11779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SF-pose: Semantic-fusion six degrees of freedom object pose estimation via pyramid transformer for industrial scenarios}, volume = {22}, year = {2025}, }
• (2025). Geodesic distance field-based five-axis continuous sweep scanning method for the multi-entrance inwall surface. TASE, 22, 11753--11766. (www)

  @article{ , abstract = {Multi-entrance Inwall (MEI) surfaces are widely used in industrial applications, yet inspecting the MEI surfaces precisely remains a challenging task due to their complex multi-entrance topology and potential collision risks. The recently developed five-axis continuous sweep scanning technology offers significantly higher inspection efficiency compared to traditional point-by-point methods, presenting a valuable opportunity for accurate and efficient MEI surface inspection. However, planning a five-axis continuous sweep scanning process for general MEI surfaces still largely relies on human input. To address this challenge, this paper presents a novel set of methods for generating five-axis sweep scanning paths, specifically designed for the automatic and efficient inspection of MEI surfaces. Our methodology utilizes a sophisticated heat-induced geodesic distance field (GDF) to calculate guiding curves, which are used to generate the sweep scanning paths and partition the accessible regions of the MEI surface. This approach results in the creation of continuous five-axis sweep scan inspection paths that enhance both inspection efficiency and surface coverage rates. The proposed method has been validated through physical inspection experiments and computer simulations, with results confirming its feasibility and demonstrating its advantages over two benchmark approaches. Note to Practitioners—This article aims to generate an automatic, highly efficient inspection path for MEI surfaces using a five-axis coordinate measuring machine (CMM). While existing methods have addressed some issues in free-form surface inspection, they primarily focus on external and open surface inspection and do not adequately adapt to MEI surfaces, which are often occluded by challenging collision situations and complex topology. As a result, planning a five-axis continuous sweep scanning process for a general MEI surface still heavily relies on human interaction. To address this limitation, we propose an inspection path generation method that constructs a set of guiding curves considering the geometric information of both the entrances and collision situations. We utilize a heat-induced Geodesic Distance Field (GDF) to compute guide paths and assist in partitioning the accessible region. Through experiments and computer simulations, our proposed method demonstrates superior performance compared to traditional benchmarking methods in terms of both inspection efficiency and point accessibility rate. The generated inspection path conforms to the workpiece surface geometry, effectively overcoming challenges such as high interference and topological complexity in MEI, thus enabling efficient and comprehensive surface measurements.}, archive = {J_TASE}, author = {Yuzhu Ding and Zhaoyu Li and Dong He and Kai Tang and Pengcheng Hu}, doi = {10.1109/TASE.2025.3537873}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11753--11766}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Geodesic distance field-based five-axis continuous sweep scanning method for the multi-entrance inwall surface}, volume = {22}, year = {2025}, }
• (2025). Distributionally robust optimization for the multi-period multi-item lot-sizing problems under yield uncertainty. TASE, 22, 11731--11752. (www)

  @article{ , abstract = {Yield uncertainty is an important issue in various industries such as agriculture, food, and textile where the production output relies on uncontrollable factors and fluctuating raw material quality. To systematically leverage data to deal with uncertainty cost-effectively, distributionally robust optimization combines the strengths of stochastic programming and robust optimization by optimizing the expected costs against an ambiguity set that defines possible distributions. In this work, we leverage a data-driven robust optimization framework and formulate a mixed-integer distributionally robust multi-item lot-sizing model with uncertain production yield to determine a robust production plan. To this end, we use a scenario-wise formulation that partitions the available data into scenarios that define different patterns influencing the quality of the product and production process. In addition, we apply the proposed approach to real-world data of a case study to demonstrate the effectiveness of the proposed framework in dealing with yield uncertainty. Our experimental results show that distributionally robust plans lead to more effective cost-saving strategies and decreased risk of stock-outs. Additionally, our findings suggest that the proposed model exhibits lower sensitivity to variations in production yield realizations and it is more proficient in incorporating historical data into the decision-making process. This results in a more effective response to challenges encountered within the production system under yield uncertainty. Note to Practitioners–In a production context with various sources of uncertainty for which the mathematical estimation of the uncertain parameter can be complex or hard to perform, it would be better to use the proposed robust approach. Here, any information, accurate or not, new or historical, can be integrated into the system to improve the quality of the obtained production plan, yet still robust and mitigate nervousness. For the reduction of conservatism, the proposed approaches indicate how the manufacturer risk aversion could be integrated into the decision models to respond to the strategic need for robustness for production planning under uncertainty.}, archive = {J_TASE}, author = {Paula Metzker and Simon Thevenin and Yossiri Adulyasak and Alexandre Dolgui}, doi = {10.1109/TASE.2025.3535945}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11731--11752}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributionally robust optimization for the multi-period multi-item lot-sizing problems under yield uncertainty}, volume = {22}, year = {2025}, }
• (2025). V²-SfMLearner: Learning monocular depth and ego-motion for multimodal wireless capsule endoscopy. TASE, 22, 11717--11730. (www)

  @article{ , abstract = {Deep learning can predict depth maps and capsule ego-motion from capsule endoscopy videos, aiding in 3D scene reconstruction and lesion localization. However, the collisions of the capsule endoscopies within the gastrointestinal tract cause vibration perturbations in the training data. Existing solutions focus solely on vision-based processing, neglecting other auxiliary signals like vibrations that could reduce noise and improve performance. Therefore, we propose V2-SfMLearner, a multimodal approach integrating vibration signals into vision-based depth and capsule motion estimation for monocular capsule endoscopy. We construct a multimodal capsule endoscopy dataset containing vibration and visual signals, and our artificial intelligence solution develops an unsupervised method using vision-vibration signals, effectively eliminating vibration perturbations through multimodal learning. Specifically, we carefully design a vibration network branch and a Fourier fusion module, to detect and mitigate vibration noises. The fusion framework is compatible with popular vision-only algorithms. Extensive validation on the multimodal dataset demonstrates superior performance and robustness against vision-only algorithms. Without the need for large external equipment, our V2-SfMLearner has the potential for integration into clinical capsule robots, providing real-time and dependable digestive examination tools. The findings show promise for practical implementation in clinical settings, enhancing the diagnostic capabilities of doctors. Note to Practitioners—This paper is motivated by the problem of estimating the depth and ego-motion information for the wireless capsule endoscopy in the human gastrointestinal tract to realize accurate, efficient, robust, and real-time inspection. Our estimation method does not engage any external localization equipment. Instead, inspired by the existing research on integrating capsule endoscopy and inertial measurement units, we introduce vibration signals into vision-based depth and ego-motion estimation approaches, improving the accuracy and robustness of the estimation results based on multimodal learning methods. Research on capsule robots or computer vision can readily be combined with our framework for various clinical and industrial applications.}, archive = {J_TASE}, author = {Long Bai and Beilei Cui and Liangyu Wang and Yanheng Li and Shilong Yao and Sishen Yuan and Yanan Wu and Yang Zhang and Max Q.-H. Meng and Zhen Li and Weiping Ding and Hongliang Ren}, doi = {10.1109/TASE.2025.3530791}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11717--11730}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {V²-SfMLearner: Learning monocular depth and ego-motion for multimodal wireless capsule endoscopy}, volume = {22}, year = {2025}, }
• (2025). Discounted stable adaptive critic design for zero-sum games with application verifications. TASE, 22, 11706--11716. (www)

  @article{ , abstract = {In this paper, an adaptive critic design with performance guarantee is established based on the discounted value iteration algorithm to settle with the optimal regulation problem for discrete-time zero-sum games. Value iteration is implemented to obtain the approximate optimal solutions to the Hamilton-Jacobi-Isaacs equation for nonlinear systems and the game algebraic Riccati equation for linear systems. Then, we focus on system stability affected by the introduction of the discount factor and the admissibility of the policy pairs in the value iteration process. The appropriate selection range of the discount factor and the criteria for ensuring system stability are established to assist in obtaining the stabilized optimal policy pair, which not only makes the cost function converge to the optimal value, but also guarantees the asymptotic stability of the closed-loop system. Finally, practical examples for the power system and the ball-beam system are conducted to demonstrate the effectiveness of the presented method. Note to Practitioners—Since there exist a multitude of dynamic systems with uncertainty and interference, the zero-sum game problems are ubiquitous, especially when dealing with dynamic systems featuring antagonistic properties. As an important research direction in the field of optimal control, zero-sum games usually involve designing policy pairs that can optimize the system performance in the presence of adversarial disturbances. Due to the excellent adaptability, value iteration in adaptive dynamic programming is employed to deal with this kind of issues. In addition to focusing on the optimality of policies, the system stability during the control process is equally significance, where the stability is the premise of all operations. Therefore, we are dedicated to providing guidance on the optimal regulation of discrete-time zero-sum games with performance guarantee, which contributes to obtain the stable optimal policy pair. Theoretical analysis of the stability is provided and the asymptotic stability of the system is ensured, which improves the performance of the designed controller. Furthermore, simulation experiments for practical applications are conducted, which verify the feasibility and effectiveness of the proposed control design.}, archive = {J_TASE}, author = {Jin Ren and Ding Wang and Menghua Li and Junfei Qiao}, doi = {10.1109/TASE.2025.3539772}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11706--11716}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discounted stable adaptive critic design for zero-sum games with application verifications}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered adaptive tracking control for switched nonlinear system with vanishing control gains. TASE, 22, 11693--11705. (www)

  @article{ , abstract = {The asymptotic tracking control problem is investigated for nonstrict-feedback switched nonlinear systems by developing a dynamic event-triggered adaptive control technology. A key contribution is to design a hysteresis-type switching rule and an event-triggering-based switching controller to realize the asymptotically convergent target. First, by utilizing the Nussbaum-type function technique, the obstacle caused by unknown nonsmooth saturation nonlinearity is circumvented. Second, a novel gain-dependent switching signal is constructed, which ensures a time interval between any adjacent switching. Furthermore, a dynamic event-triggered schema is established, which allows the occurrence of asynchronous switching phenomenon generated by switching and event-triggering. An event-triggering-based switching controller is designed so that the adaptive asymptotic tracking control problem is solvable even when input saturation and vanishing control gains arise. Finally, simulation results with theoretical analysis are presented to validate the feasibility of the control algorithm. Note to Practitioners—With the development of artificial intelligence techniques, the studies of switched nonlinear systems have received widespread attention. This investigation is motivated by the vanishing control gains problem for switched nonlinear systems. We propose a co-design of a hysteresis-type gain-dependent switching rule and an event-triggering-based switching controller for switched nonlinear systems under input saturation and vanishing control gains. In existing results, the designed control signal is required to enter into the systems everywhere, while we relax the restriction. An improved dynamic event-triggered adaptive control framework has been established to obtain less conservative conditions further, where the control gain is allowed to vanish at some points. This study presents the method for practitioners interested in adaptive control design.}, archive = {J_TASE}, author = {Dong Yang and Yanrui Sun and Guangdeng Zong and Haibin Sun and Tao Liu}, doi = {10.1109/TASE.2025.3539861}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11693--11705}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered adaptive tracking control for switched nonlinear system with vanishing control gains}, volume = {22}, year = {2025}, }
• (2025). Secured bipartite consensus control for nonlinear multi-agent systems against hybrid attacks: A component-based WTOD protocol. TASE, 22, 11680--11692. (www)

  @article{ , abstract = {This article studies the secure bipartite consensus control issue for a nonlinear multi-agent system (MAS) subjected to hybrid attacks. The challenge is to simultaneously consider improving the bandwidth resource utilization of each agent and ensuring secure bipartite consensus. In response to the former, a component-based weighted try-once-discard (WTOD) protocol is proposed, which can save bandwidth resources by selecting the maximum component of each agent’s measurement output instead of only a single agent’s measurement output to transmit to the observer. Different from the previous work, which considered only one single type/channel attack, a unified hybrid attack model is constructed in which aperiodic denial-of-service (DoS) attacks and unknown false data injection (FDI) attacks occur simultaneously in the channels from sensor-to-observer and controller-to-actuator. At this point, a co-estimation strategy is given to estimate the state of agents and the FDI attack signals. Based on this, the observer-based compensation controller is developed to mitigate the impact of FDI attacks based on the feed-forward method. Meanwhile, a set of sufficient conditions is presented to guarantee the existence of such a resilient controller, which can further ensure the underlying system achieves ultimately bounded bipartite consensus. Finally, a simulation example verifies the feasibility and correctness of the proposed method. Note to Practitioners–Nonlinear MASs currently have numerous real-world applications, such as unmanned aerial vehicle formation control and intelligent transportation. Besides the collaborative behavior among agents, antagonistic interactions are also common in the real world. Thus, the research on the bipartite consensus problem would benefit the application of unmanned missions in complex environments. This paper investigates the secure bipartite consensus control problem for a class of nonlinear MAS under hybrid attacks. Notably, these systems are particularly susceptible to various types of cyberattacks. Therefore, it is crucial to mitigate the impact of cyber attacks on nonlinear MAS that rely on communication links for information transmission. To address the challenges posed by simultaneous DoS and FDI attacks occurring in two communication channels, this paper proposes a resilient compensation-based bipartite consensus control scheme for MASs to proactively mitigate the impact of attacks on system performance. Furthermore, a component-based WTOD protocol is proposed to reduce communication resource usage among agents, making the proposed control strategy easier to implement and more user-friendly for control engineers.}, archive = {J_TASE}, author = {Tingting Ru and Chenxiao Cai and Ju H. Park}, doi = {10.1109/TASE.2025.3534029}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11680--11692}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secured bipartite consensus control for nonlinear multi-agent systems against hybrid attacks: A component-based WTOD protocol}, volume = {22}, year = {2025}, }
• (2025). Dual-domain classification-aided high-quality PET synthesis with shared information maximization. TASE, 22, 11668--11679. (www)

  @article{ , abstract = {Positron emission tomography (PET) is widely applied in clinic for providing crucial diagnosis information. However, its inherent radiation exposure inevitably brings potential health risk for patient. To reduce radiation risk while also obtaining high-quality PET image, we plan to synthesize standard-dose PET (SPET) from low-dose PET (LPET). Since PET images can be represented in both projection domain and image domain (dual domains) emphasizing different information, considering dual domains in PET synthesis could contribute to better performance. In this way, we propose a novel dual-domain model for high-quality PET synthesis, named DCBi-GAN, by introducing a denoising network for the projection domain and an enhancing network for the image domain to effectively exploit dual-domain information. Concretely, the denoising network takes the LPET sinogram converted from LPET image to suppress noise and artifacts in the projection domain. Then, the enhancing network in the image domain takes the denoised LPET image (transferred back from the denoised sinogram) to enhance image quality. Notably, as LPET and SPET images come from the same subject, the abundant shared information between LPET and SPET can be used for boosting synthesis performance. Specially, we design a bi-directional contrastive generative adversarial network (GAN) to encourage maximal preservation of the shared information. Besides, we introduce a mild cognitive impairment (MCI) classification task to enhance clinical applicability of the synthesized PET. Evaluation on both Real Human Brain dataset and Phantom Brain dataset demonstrates effectiveness and superiority of our proposed model.Note to Practitioners—Positron emission tomography (PET) is a primary nuclear imaging technique for tumor detection and brain disorder diagnosis in the early stage of diseases, while the inherent radiation exposure inevitably raises concerns about potential health risk. This article proposes a novel PET image synthesis model to obtain clinically accepted PET image at low dose, namely DCBi-GAN, by taking account of the complementary multi-domain information and the modality shared content information, with a mild cognitive impairment (MCI) classification task to further boost clinical applicability of synthesized PET images. We experimentally validate the effectiveness of proposed DCBi-GAN on two datasets. Our proposed method could facilitate diagnosis and treatment of disease, to be used in the existing computer-aided medical systems.}, archive = {J_TASE}, author = {Yuchen Fei and Chen Zu and Xi Wu and Jiliu Zhou and Yan Wang and Dinggang Shen}, doi = {10.1109/TASE.2025.3539234}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11668--11679}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-domain classification-aided high-quality PET synthesis with shared information maximization}, volume = {22}, year = {2025}, }
• (2025). Neural-network-based event-triggered formation tracking for nonlinear multi-UAV systems with switching topologies under DoS attacks. TASE, 22, 11656--11667. (www)

  @article{ , abstract = {A time-varying formation tracking (TVFT) control method under denial-of-service (DoS) attacks is proposed for a class of multi-uncrewed aerial vehicle (UAV) system with unknown nonlinearity. The neural network approximation is utilized to mitigate the impact of nonlinear dynamics on the tracking performance. Meanwhile, the switching topologies are adopted to ensure the reliability of the communication topology. Considering the resource constraints of the system communication network, an improved integral event-triggered mechanism is devised to further decrease the triggering frequency in comparison to the traditional event-triggered one. By using the multi-Lyapunov function, the tracking error is proved to be bounded. Simulations convincingly demonstrate the effectiveness and merits of the developed TVFT and the integral event-triggered method. Note to Practitioners—Multi-UAV systems are widely utilized in both civil and military domains, with communication networks serving as crucial channels for information exchange among UAVs. However, these communication networks are often vulnerable to various types of attacks. Therefore, the control problems of multi-UAV systems under DoS attacks are of great importance. A switching event-triggered control protocol is developed for multi-UAV systems to facilitate effective formation tracking in the presence of DoS attacks. This protocol employs switching topologies to enhance the reliability and flexibility of the communication network. The improved event-triggered mechanism effectively balances the system performance and the communication consumption.}, archive = {J_TASE}, author = {Shuang Shi and Shuqing Wu and Bo Wei}, doi = {10.1109/TASE.2025.3539401}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11656--11667}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Neural-network-based event-triggered formation tracking for nonlinear multi-UAV systems with switching topologies under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Meal delivery routing problem with a hybrid fleet of riders and autonomous vehicles under dynamic environment. TASE, 22, 11642--11655. (www)

  @article{ , abstract = {Autonomous vehicles (AVs) are considered as next-generation delivery vehicles for logistics systems. This study proposes a dynamic Meal Delivery Routing Problem with a hybrid Rider-AV fleet (MDRP-RA). The hybrid fleet consists of riders and AVs. Each order must be fulfilled by an AV or a rider. Some orders can be delivered by riders or AVs only, while others can be delivered by both. The food of each order is one of three product segments (regular, frozen, and hot food), and each segment has a particular temperature need. An AV has multiple compartments, each of which needs to be cooled (heated) if it contains frozen (hot) food. Thus, AVs can deliver all kinds of food, while riders can deliver regular food only. A mathematical programming model is established for MDRP-RA, with the objective of minimizing the total cost, including the vehicle fixed cost, delivery fee to riders, energy consumption cost, and penalty cost for delay. An Adaptive Large Neighborhood Search based Approach (ALNS-A) is proposed to solve MDRP-RA. It involves a local search procedure with removal and insertion operators, where five operators are specifically devised for the problem. Experiments show that it can effectively solve MDRP-RA and outperforms comparative approaches.Note to Practitioners—AVs have great application potential in meal delivery since they have the advantages of saving labor costs, large capacity, and maintaining food temperature. The hybrid fleet of AVs and riders can meet diversified customer needs but brings challenges to the meal delivery route problem under a dynamic environment. This paper proposes a dynamic meal delivery routing problem with a hybrid rider-AV fleet, with the objective of minimizing the total cost. Some orders must be delivered by riders, some by AVs, and some by both. The food falls into three product segments, i.e., regular, frozen, and hot food. AVs can provide cooling or auxiliary heating to maintain the meal’s temperature. An adaptive large neighborhood search-based approach is proposed, which can provide high-quality solutions for problem instances. The approach can be embedded in takeout information platforms to realize intelligent scheduling of meal delivery.}, archive = {J_TASE}, author = {Zhishuo Liu and Xingquan Zuo and MengChu Zhou and Bin Jia and Chongyang Xin}, doi = {10.1109/TASE.2025.3534143}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11642--11655}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Meal delivery routing problem with a hybrid fleet of riders and autonomous vehicles under dynamic environment}, volume = {22}, year = {2025}, }
• (2025). Effective probabilistic neural networks model for model-based reinforcement learning USV. TASE, 22, 11625--11641. (www)

  @article{ , abstract = {Gaussian process (GP) offers a robust solution for modeling the dynamics of unmanned surface vehicles (USV) in model-based reinforcement learning (MBRL). However, the rapidly increasing computational complexity with a large sample capacity of GP limits its application in complex scenarios that require substantial samples to cover the state space. In this article, a novel probabilistic MBRL approach, probabilistic neural networks model predictive control (PNMPC) is proposed to tackle this issue. With an iterative learning framework, PNMPC properly models the USV dynamics using neural networks from a probabilistic perspective to avoid the computational complexity associated with sample capacity. Employing this model to effectively propagate system uncertainties, a model predictive control (MPC) policy is developed to robustly control the USV against external disturbances. Evaluated by position-keeping and multiple targets-tracking scenarios on a real USV data-driven simulation, the proposed method consistently demonstrates its significant superiority in both model accuracy and control performance compared to not only GP model-based approaches but also the probabilistic neural networks-based MBRL baselines, across various scales of external disturbances.Note to Practitioners—Modelling the system dynamics and maintaining computational efficiency with a large sample set has been challenging for MBRL in the USV domain. We propose a novel neural network modeling method to capture the dynamic features of USV within an RL loop and develop a robust MPC policy based on its uncertainty propagation. Our method achieves computational complexity independent of the sample capacity and outperforms related baselines in model accuracy and control performance.}, archive = {J_TASE}, author = {Wenjun Huang and Yunduan Cui and Huiyun Li and Xinyu Wu}, doi = {10.1109/TASE.2025.3539317}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11625--11641}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Effective probabilistic neural networks model for model-based reinforcement learning USV}, volume = {22}, year = {2025}, }
• (2025). Scheduling time-constrained cluster tools with non-identical parallel processing chambers. TASE, 22, 11613--11624. (www)

  @article{ , abstract = {As a computer-integrated manufacturing system, cluster tools are widely used for semiconductor manufacturing. To tackle their scheduling problems with parallel processing chambers, existing studies assume that the parallel processing chambers at each step are identical. However, due to machine aging, some differences among them could appear in reality. This work represents the first one to report the scheduling problems of cluster tools with non-identical parallel processing chambers. By focusing on time-constrained single and dual-armed cluster tools with non-identical parallel processing chambers, it derives the sufficient conditions under which a system is schedulable. Based on them, it develops two algorithms to obtain the feasible and optimal schedules. It also develops two linear programming models to find the feasible and optimal ones if the sufficient conditions are violated. The experimental results based on 5400 random instances show that the proposed scheduling method can be readily used to generate schedules for both time-constrained single and dual-armed cluster tools. Note to Practitioners—For cluster tools, due to machine aging, the wafer processing time at parallel processing chambers performing the same process tends to be different in practice. Such parallel processing chambers are called non-identical ones. It is tough to schedule time-constrained (with wafer residency time constraint) cluster tools with non-identical parallel processing chambers. The challenge is how to allocate robot waiting time at each step such that the wafer residency time constraint is satisfied. To ensure the high quality of finished wafers, their post-processing time at non-identical parallel processing chambers must be equal. This work offers readily applicable methods to address the problems, thereby generating a positive influence on the wafer quality and yield in the critically important field of semiconductor manufacturing.}, archive = {J_TASE}, author = {Fajun Yang and Rui Wang and Nan Chen and Naiqi Wu and MengChu Zhou}, doi = {10.1109/TASE.2025.3527479}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11613--11624}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling time-constrained cluster tools with non-identical parallel processing chambers}, volume = {22}, year = {2025}, }
• (2025). Automated strategy for tissue analysis in anatomic pathology: Fiducial marker integration and multisurface tissue comparison. TASE, 22, 11601--11612. (www)

  @article{ , abstract = {In anatomic pathology laboratories, several processes are still manual. Automated solutions can help standardizing the fabrication/processing of paraffin-embedded tissue blocks (PETBs) for a reliable and more effective diagnosis, in accordance with the requirement to work with unique and heterogeneous biological samples. We present a novel and automated approach to introduce a fiducial marker within PETBs, serving as a reliable reference point. This would assist the clinician in identifying specific regions of biological tissue in paraffin-embedded tissue blocks and paraffin-free tissue slices, in case the patient requires further laboratory analysis on that tissue portion to indicate a tailored treatment (e.g., oncological treatment). Two automated platforms are integrated into the conventional anatomic pathology workflow, according to the proposed strategy. The first one, named “Indexing”, involves the insertion of a fiducial marker into tissue-free areas within PETBs. The second one, named “Virtual marker reconstruction”, is based on image analysis and virtually reconstructs the fiducial marker on the paraffin-free tissue slice. An algorithm that analyzes the similarity of the tissue was also developed to assist in the traceability of the biological tissue along its processing from the embedding to the post-staining phase. Together, these platforms could assist the work of anatomopathologists avoid errors and support the final diagnosis in a future automated laboratory. Note to Practitioners—This research addresses a specific challenge encountered in anatomic pathology laboratories. The manual procedures involved in the fabrication and processing of paraffin-embedded tissue blocks highlight the pressing need for continuous verification and traceability at various stages of tissue processing and analysis. A lack of proper traceability poses challenges for anatomopathologists, leading to difficulties in managing the accurate identification of tissue blocks and slices, which contain unique biological tissues. This deficiency may lead to delayed or inaccurate diagnoses, with severe consequences for patients. We propose an innovative solution that focuses on two automated platforms that can seamlessly integrate into an anatomic pathology laboratory, and provide a way to correctly recognize and monitor biological tissue all along different procedures.}, archive = {J_TASE}, author = {L. Vannozzi and L. Guachi-Guachi and J. Ruspi and S. Ciancia and G. Baldi and D. Lunni and P. Scarlino and A. Poliziani and A. Zucca and M. Bellini and G. A. Pedrazzini and A. Cavazzana and L. Ricotti}, doi = {10.1109/TASE.2025.3538513}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11601--11612}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated strategy for tissue analysis in anatomic pathology: Fiducial marker integration and multisurface tissue comparison}, volume = {22}, year = {2025}, }
• (2025). Adaptive quantized control for a class of hysteresis nonlinear system with irregular constraints and its application to piezoelectric positioning stage. TASE, 22, 11589--11600. (www)

  @article{ , abstract = {This paper investigates the tracking control problem of hysteresis nonlinear systems with irregular constraint boundaries. By blending the constraint boundary with the barrier function and using the system transformation technique, an adaptive quantized tracking control scheme is developed, in which three significant advantages are obtained: 1) Compared with traditional constraint control methods, the constraints considered in this work only need to satisfy the condition of piecewise differentiability, which may be time-varying, asymmetric, and may appear in stages; 2) The designed controller can maintain the state within the boundary range where constraints are active, effectively managing situations where constraints and non-constraints alternate. 3) The quantizer used combines logarithmic and uniform quantizers, effectively reducing communication costs and maintaining acceptable quantization errors, which ultimately improves the overall system performance. The method’s effectiveness and superiority are validated through experiments conducted on a piezoelectric-driven motion system. Note to Practitioners—Smart materials are crucial components in high-precision positioning applications and indispensable tools for the precision manufacturing processes of advanced equipment. Current research on tracking control of smart material actuation systems typically focuses on eliminating the effects of hysteresis nonlinearity to achieve stable tracking. However, there is a lack of attention on ensuring that the desired trajectory and system states consistently meet predetermined constraints. It is observed that most existing works consider scenarios where constraints on system output or state are either uniformly present or uniformly absent. Consequently, the control methods proposed in these works become ineffective when constraint boundary gaps appear. On the other hand, to reduce the communication burden between system modules while ensuring system performance, it becomes necessary to quantify control signals. The existing quantizers employ a logarithmic distribution for quantization density, which significantly reduces communication load and quantization error. However, when the quantized signal’s value is large, it results in considerable quantization error, thereby affecting the system’s control performance. Therefore, this paper addresses the smart material actuation system with hysteresis characteristics by considering irregular boundary constraints and improving the traditional quantizer. This approach is further combined with system state transformation and backstepping techniques to construct an adaptive quantized control scheme. Additionally, a physical experimental platform based on a piezoelectric actuator within smart materials is developed to validate the effectiveness and superiority of the proposed control method.}, archive = {J_TASE}, author = {Heyu Hu and Shengjun Wen and Jun Yu and Changan Jiang}, doi = {10.1109/TASE.2025.3538918}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11589--11600}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive quantized control for a class of hysteresis nonlinear system with irregular constraints and its application to piezoelectric positioning stage}, volume = {22}, year = {2025}, }
• (2025). A lifelong multi-shuttle scheduling framework for the AS/RS system. TASE, 22, 11577--11588. (www)

  @article{ , abstract = {Efficient scheduling of multi-shuttle is crucial for optimizing the performance of Automated Storage/Retrieval Systems (AS/RS). Multi-Agent Path Finding (MAPF) techniques play a pivotal role in addressing scheduling challenges by guaranteeing collision-free paths for multiple agents, however the assumptions of MAPF cannot meet the practical applications, such as 1) agents move at constant speeds and change directions instantaneously, 2) agents remain stationary at their destinations, 3) agents have different destinations. In this paper, we propose a novel lifelong multi-shuttle scheduling framework (LMSSF) to fill this gap based on the AS/RS system implemented in Hunan, China. LMSSF can perform re-planning and execution occur simultaneously, ensuring robust performance even in dynamic and uncertain environment. To apply MAPF to practical scenarios, we propose task conflict resolution, an improved single-shot MAPF and a path constraint mechanism to ensure collision-free movement of shuttles and improve the throughput performance of AS/RS system. Empirical evaluations on throughput and occupancy rate of outbound stations demonstrate the superiority of our proposed algorithm. Finally, LMSSF is applied to a real-world system and the experimental results show a 31.8% improvement in throughput compared to conventional strategies.Note to Practitioners—The motivation of this article stems from the challenges posed by the assumptions inherent in most MAPF algorithms when applied to practical AS/RS environments. Traditional MAPF algorithms typically compute discrete collision-free paths based on predefined start and goal locations. However, in real-world scenarios, tasks are continually generated, shuttles are highly dynamic, and various environmental constraints exist. To address these challenges, we propose the LMSSF as a means to adapt existing MAPF algorithms for practical AS/RS systems. This framework incorporates task conflict resolution, path planning optimization, and efficient shuttle control mechanisms. The paths obtained from the MAPF solver can be seamlessly integrated into our framework with minor format conversions. Through comprehensive testing, our algorithm demonstrates significantly improved throughput performance compared to traditional strategies.}, archive = {J_TASE}, author = {Hongkai Fan and Bo Ouyang and Zheng Fang and Zhi Yan and Jiawen He and Zuozhi Zhang and Yaonan Wang}, doi = {10.1109/TASE.2025.3529847}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11577--11588}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A lifelong multi-shuttle scheduling framework for the AS/RS system}, volume = {22}, year = {2025}, }
• (2025). A projective weighted DTW-based monitoring approach for multi-stage processes with unequal durations. TASE, 22, 11564--11576. (www)

  @article{ , abstract = {Multi-stage processes, such as batch and transition processes, often have unequal operation duration due to differing conditions, posing significant challenges to process monitoring. Although dynamic time warping (DTW) has been applied for offline synchronization, it cannot adequately align an evolving, incomplete online batch with completed historical batches due to inherent inconsistencies in their progression. Moreover, traditional methods generally overlook time-scale faults in the operational progress of the process, which undermines overall monitoring performance. To address these issues, a novel projective weighted DTW (PwDTW)-based method is proposed to monitor multi-stage processes with unequal durations. First, the asymmetric weighted DTW is adopted to offline align the original training dataset with different lengths, incorporating the Itakura parallelogram constraint to restrict the region of the warping path. Then, the PwDTW with an open-ended strategy is proposed to handle the online asynchronization problem by assessing the progress and similarity of the ongoing trajectory against each training trajectory. Further, the k-nearest neighbor (KNN) is used to identify the most similar subsequences of the training dataset with the online trajectory. Leveraging these subsequences, two monitoring indices are designed to monitor the process in not only amplitude scale but also time scale. The two indices reflect both the strength and speed of the process. Finally, a benchmark Tennessee Eastman process and a practical semiconductor manufacturing case are introduced to prove the effectiveness of the proposed method. Note to Practitioners—This paper aims to solve the practical problem of on-line monitoring multi-stage processes with unequal durations from both amplitude and time perspectives. Traditional methods typically focus on offline alignment of processes with different durations and lack online applicability. Additionally, their focus on monitoring only the signal amplitude also limits the potential to monitor whether the process is progressing too fast or too slowly. This paper proposes a novel PwDTW-based monitoring method to address these issues. The approach begins by using asymmetric weighted DTW (wDTW) for offline alignment of historical data, synchronizing datasets with varying durations. When new samples arrive, this paper develops a PwDTW method with an open-ended strategy that evaluates process progress and identifies similar patterns for the online sample. Further, by assessing both the strength and the speed of the process, this paper provides dual-perspective monitoring that is more comprehensive than traditional methods. The effectiveness of this approach has been demonstrated through its application to the TE process and a real-world semiconductor manufacturing process. The proposed method is universal and can be adapted to other industries, such as energy production or pharmaceuticals, where varying durations are ubiquitous in the systems.}, archive = {J_TASE}, author = {Ying Zheng and Peiming Wang and Yang Wang and David Shan-Hill Wong}, doi = {10.1109/TASE.2025.3537687}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11564--11576}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A projective weighted DTW-based monitoring approach for multi-stage processes with unequal durations}, volume = {22}, year = {2025}, }
• (2025). Mask privacy preservation prescribed-time consensus control for nonlinear multi-agent systems. TASE, 22, 11554--11563. (www)

  @article{ , abstract = {In this study, we propose an innovative prescribed-time consensus control strategy for nonlinear strict-feedback multi-agent systems (MASs) with privacy protection requirements. Firstly, compared with the existing privacy protection strategies, the mask function adopted in this paper remains unknown to all agents, including the sender, thus greatly improving the security level of information transmission. Secondly, the existing related research results basically overlook prescribed-time control in the context of privacy preservation, based on the backstepping method, a prescribed time performance function is adopted in this paper, so that the systems can make the tracking error within the defined accuracy range within a user-defined time. Finally, through the verification of MATLAB simulation experiments, the proposed control strategy not only effectively realizes the privacy-preserving consensus control of multi-agent systems, but also shows better control performance compared with the existing schemes. Note to Practitioners—This paper aims to develop a mask privacy protection prescribed-time control algorithm for information transmission between multiple agents. In the automation industry, the demand for privacy protection in multi-agent systems is critical, necessitating the implementation of robust measures during agent collaboration and data sharing to safeguard data confidentiality. Employing advanced privacy-preserving technologies is essential to prevent the exposure of sensitive information, thereby ensuring the security of corporate secrets and operational integrity, in compliance with the evolving stringent privacy regulations. In addition, prescribed-time control enables users to achieve preset accuracy within a predefined time, reducing industrial resource consumption and improving resource utilization in the automation industry.}, archive = {J_TASE}, author = {Junhao Yuan and Wei Sun and Yougang Sun and Shun-Feng Su}, doi = {10.1109/TASE.2025.3535924}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11554--11563}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mask privacy preservation prescribed-time consensus control for nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Neural adaptive MPC with online metaheuristic tuning for power management in fuel cell hybrid electric vehicles. TASE, 22, 11540--11553. (www)

  @article{ , abstract = {In this paper, we present an advanced control framework for power management applications, named Neural Adaptive Model Predictive Control (NA-MPC), designed to provide an optimal power allocation among multiple energy sources, perform a multi-objective online adaptation of the optimal control policy, and ensure a fast real-time execution with low computational demand. NA-MPC augments general MPC problems with three key features: 1) an online metaheuristic tuning strategy adapts the MPC cost function weights, to attain multiple concurrent control objectives at once; 2) through neural emulation, the MPC control policy is replaced by an equivalent neural MPC controller, exhibiting universal approximation guarantees and ensuring real-time feasibility; 3) a neural black-box MPC prediction model is employed, identified only via noise-corrupted input-output measurements from the plant, which is assumed to be unknown. The general formulation and versatility of NA-MPC make it potentially applicable to several power management scenarios; in this work, we apply NA-MPC to the case study of power management in fuel cell hybrid electric vehicles (FCHEVs), a topic of growing interest within the frame of sustainable transportation, for which novel and efficient strategies are still lacking. The effectiveness of NA-MPC is thoroughly assessed via numerical simulations, demonstrating its capability to optimally attain multiple control objectives concurrently in real time; moreover, NA-MPC consistently outperforms the most prominent state-of-the-art HEV power management strategies. Note to Practitioners—The aim of this paper is to introduce an advanced online-adaptive optimal control strategy, named NA-MPC, and employ it as a novel power management strategy for FCHEVs, with the purpose of addressing several technical shortcomings of the existing state-of-the-art strategies. Specifically, the latter typically fail in performing effective trade-offs between accurate power tracking and supply consumption, proving a merely suboptimal control action. Such strategies have also very limited adaptation capabilities, being either offline-tuned or employing simple non-optimal adaptation policies. Moreover, only few basic optimal control strategies are proposed in the literature, with little focus on their real-time feasibility. By contrast, our NA-MPC strategy provides an optimal power allocation, effectively attains multiple concurrent control objectives, and, thanks to its neural embedding, is real-time feasible and easily implementable on hardware with limited computational resources. Furthermore, the general formulation and versatility of NA-MPC enable its potential application across a wide variety of different power management scenarios.}, archive = {J_TASE}, author = {Lorenzo Calogero and Michele Pagone and Francesco Cianflone and Edoardo Gandino and Carlo Karam and Alessandro Rizzo}, doi = {10.1109/TASE.2025.3534402}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11540--11553}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Neural adaptive MPC with online metaheuristic tuning for power management in fuel cell hybrid electric vehicles}, volume = {22}, year = {2025}, }
• (2025). Jerk-limited online trajectory scaling for cable-driven parallel robots. TASE, 22, 11529--11539. (www)

  @article{ , abstract = {Motion planning of cable-driven parallel robots (CDPRs) suffers from difficulties imposed by the flexibility and unilateral property of cables. Existing methods either rely on specific motion primitives or employ complex numerical search or optimization processes, which cannot be applied to real-time applications with arbitrary path constraints. Aiming to narrow this research gap, this paper proposes a look-ahead online scaling approach to generate feasible trajectories of CDPRs subject to cable velocity, acceleration, jerk and tension constraints. Firstly, based on a desired path, the constraint equations are converted into the equivalent bounds on the path states by a look-ahead bounds estimation module. Then the timing law is online scaled by three cascaded controllers to fulfill the estimated bounds. Finally, the scaled timing law and the desired path are combined to form the final trajectory. Comparative studies on a laboratory-developed CDPR prototype demonstrate that the proposed approach outperforms state-of-the-art methods in terms of solution quality and computation time. Note to Practitioners–This paper was motivated by the problem of generating feasible trajectories of cable-driven parallel robots (CDPRs) subject to cable velocity, acceleration, jerk and tension constraints. Existing approaches either rely on specific motion primitives or employ complex numerical search or optimization processes, which cannot be applied to real-time applications with arbitrary path constraints. This paper proposes a look-ahead online scaling approach to generate feasible trajectories of CDPRs subject to these constraints. The approach uses a look-ahead bounds estimation module to determine the constraint bounds on the path states and preserve the stability of the approach. And a cascaded trajectory scaling algorithm is designed to steer the constrained path states to track a reference signal. A theoretical proof of the convergence of the scaling algorithm is provided. Through comparative studies, we show that the approach outperforms state-of-the-art methods in terms of solution quality and computation time. Experiments on a real robot prototype indicate that our method effectively improves motion accuracy and alleviates robot vibrations by considering the jerk constraints.}, archive = {J_TASE}, author = {Ruobing Wang and Yangmin Li}, doi = {10.1109/TASE.2025.3533583}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11529--11539}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Jerk-limited online trajectory scaling for cable-driven parallel robots}, volume = {22}, year = {2025}, }
• (2025). Influence of visual-inertial sensor-to-segment calibration on upper limb joint angles estimation from multiple inverse kinematics methods. TASE, 22, 11519--11528. (www)

  @article{ , abstract = {This study aims to explore the potential for accurately estimating joint angles during upper limb rehabilitation tasks with different calibration procedures, inverse kinematics methods and measurement modalities. Affordable embedded visual-inertial measurement units offer a promising alternative to the costly and cumbersome gold standard marker-based optical motion capture systems. However, affordability comes with inherent sensors inaccuracies. Hence, prior to their application in a real clinical setting, it is important to demonstrate their ability for accurate joint angle estimation. Discrepancies in joint angles arise due to the inaccuracies of different sensing modalities but also to sensor-to-segment calibration procedures that significantly alter the joint offsets. Therefore, in this paper, the impact of functional and anatomical calibration procedures on joint angle estimation was compared among seven healthy young volunteers. When the same calibration procedures were applied with visual-inertial measurement units and optical motion capture systems data, a relatively small root mean square error of 7.9 deg and correlation coefficients exceeding 0.86 were observed. When different calibration procedures were applied with visual-inertial measurement units and optical motion capture systems data, higher root mean square superior to 10 deg were observed, highlighting the importance of consistency with the reference set when assessing accuracy. Furthermore, our analysis shows the benefit of using multi-body inverse kinematics procedure over treating inverse kinematics separately for each segment when dealing with inaccurate visual-inertial measurement units data. Note to Practitioners—This study addresses the practical challenge of accurately estimating upper limb joint angles in rehabilitation, using affordable Visual-Inertial Measurement Units (VIMUs) and cameras. The key finding for practitioners is the importance of consistent calibration procedures, either anatomical or functional, across both VIMUs and standard reference systems. This consistency significantly improves measurement accuracy, essential for effective rehabilitation assessment and planning. We also demonstrate that multi-body inverse kinematics (IK) methods are more reliable than single-body IK when using data from low-cost sensors. Multi-body IK better handles inaccuracies typical of affordable devices, making it a more suitable choice for clinical applications. While our results are promising, they are based on controlled conditions and do not encompass whole-body movements. Future research should focus on extending these findings to more diverse and challenging clinical scenarios, ensuring the practical applicability of this cost-effective technology in real-world rehabilitation settings.}, archive = {J_TASE}, author = {Mohamed Adjel and Raphael Dumas and Samer Mohammed and Vincent Bonnet}, doi = {10.1109/TASE.2025.3535857}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11519--11528}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Influence of visual-inertial sensor-to-segment calibration on upper limb joint angles estimation from multiple inverse kinematics methods}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered cluster consensus of multi-agent systems via PSO-GA co-design. TASE, 22, 11505--11518. (www)

  @article{ , abstract = {This paper is concerned with cluster consensus of multi-agent systems (MASs) with homogeneous and heterogeneous dynamics. Firstly, a resilient dynamic event-triggered (RDET) mechanism is proposed by introducing two parameters in the dynamic triggering threshold to describe busy and free networks of each cluster. An auxiliary function is configurated to describe the network status of agents in different clusters. Agents allocated in multiple clusters are classified into two categories: some agents are in the busy network environment and the rest are in the free network environment. The resulting consensus error system is modeled as an augmented system involving two time-varying delays with different bounds. Then, some sufficient criteria are derived for the stability analysis of homogeneous and heterogeneous MASs, respectively. Based on the obtained stability criteria, an algorithm consisting of particle swarm optimization (PSO) and a genetic algorithm (GA) is designed for a co-design of control gains and event-triggered parameters. Finally, a numerical example of satellite formation flying is simulated to illustrate the merits of the developed theoretical results.Note to Practitioners—The objective of this article is to design addresses the challenge of achieving consensus in MASs with diverse dynamics, such as those found in autonomous vehicle fleets or sensor networks. We’ve introduced a RDET mechanism that can handle both busy and free network environments by adjusting its triggering threshold based on network status. This allows agents to communicate efficiently, reducing the need for constant communication and conserving resources. Agents in multiple clusters are categorized into those in high-traffic (busy) and low-traffic (free) networks. We’ve developed a model for the consensus error system that accounts for time-varying delays, which is crucial for stability analysis in both homogeneous and heterogeneous MASs. To optimize performance, we propose an algorithm combining PSO and GA. This powerful tool co-designs control gains and event-triggered parameters, ensuring efficient and stable communication strategies for all agents regardless of their specific dynamics. In practical terms, this work provides a robust framework to improve inter-agent cooperation in complex industrial scenarios, such as large-scale robot swarms, intelligent transportation systems, or even satellite constellation management. By implementing the RDET mechanism with our optimization algorithms, practitioners can expect to achieve tighter synchronization, reduced communication costs, and better overall system resilience, as demonstrated through a simulation of satellite formation flying.}, archive = {J_TASE}, author = {Yanping Yang and Bo Shen and Xiaohua Ge and Shenrong Li and Qing-Long Han}, doi = {10.1109/TASE.2025.3536076}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11505--11518}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered cluster consensus of multi-agent systems via PSO-GA co-design}, volume = {22}, year = {2025}, }
• (2025). Deep reinforcement learning-based energy-conscious scheduling under time-of-use electricity price. TASE, 22, 11490--11504. (www)

  @article{ , abstract = {In the context of carbon peaking and carbon neutrality, green production scheduling that considers energy has attracted increasing attention. Reinforcement learning (RL) has emerged as a topic for developing efficient algorithms for solving complex combinatorial optimization problems, such as real-world shop scheduling problems. In this paper, we investigate the energy-conscious scheduling problem (ECSP) under time-of-use (TOU) electricity price with the goal of minimizing both the waiting time and extra electricity cost. A mathematical model is formulated, and an efficient deep RL (DRL)-based optimization method is proposed to solve the problem effectively. We design a novel ECSP network (ECSPNet) tailored to handle various ECSP scales based on the characteristics of the problem. Moreover, the Tchebycheff decomposition method is used to solve the multi-objective optimization problems, complemented by the application of the policy gradient method from reinforcement learning to train the ECSPNet without size limitations. Experiments verify that the proposed ECSPNet outperforms state-of-the-art methods and is computationally efficient, even on instances of larger scales unseen in training. Real-world case studies reveal that the proposed method can reduce annual total electricity cost by approximately 30% while effectively maximizing production efficiency. Note to Practitioners —Enhancing energy consciousness alongside improving production efficiency in manufacturing systems has increasingly become a focal point for both academia and industry, especially with the growing emphasis on environmental concerns and advancing industrialization. Time-of-use (TOU) electricity price policy is widely implemented to effectively balance electricity supply and demand. For business managers, appropriately responding to this policy by optimizing scheduling tasks can significantly reduce energy costs. This paper addresses a novel energy-conscious scheduling problem (ECSP) under TOU electricity price, specifically arising from the electrode graphitization production process in graphite material manufacturing. We propose a deep reinforcement learning-based energy-saving scheduling optimization method, which integrates considerations for both production efficiency and energy cost indicators. By combining the unique characteristics of ECSP with the decision-making capabilities of reinforcement learning and the perception capabilities of deep learning, our method excels in solution speed, efficiency, and adaptability. The effectiveness and practical applicability of our method have been demonstrated through experiments on actual enterprise cases of various scales. The rapidly generated optimization scheduling plans can assist enterprises in rationally arranging scheduling tasks while minimizing electricity cost. Looking ahead, our proposed method has the potential to address a wide range of energy-conscious scheduling problems under TOU electricity price policy.}, archive = {J_TASE}, author = {Caixia Wang and Wenxuan Fang and Xin Dai and Renchu He and Wei Du and Yang Tang}, doi = {10.1109/TASE.2025.3536198}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11490--11504}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep reinforcement learning-based energy-conscious scheduling under time-of-use electricity price}, volume = {22}, year = {2025}, }
• (2025). Distributed group consensus control for multi-agent systems based on combined state information. TASE, 22, 11483--11489. (www)

  @article{ , abstract = {This paper addresses group consensus control for a class of multi-agent systems (MASs) by proposing a novel distributed group consensus protocol that leverages the combined state information (including both current and outdated states) of MASs. Unlike the standard protocol that uses either only current states or only outdated states, our approach integrates combined state information to achieve consensus. We derive sufficient conditions for the MASs to achieve group consensus, eliminating the dependence on two conservative assumptions present in previous literature. Furthermore, our proposed protocol improves the upper bound of input time delays compared to the standard approach. Several examples are provided to illustrate the effectiveness of our results. Note to Practitioners—The motivation behind this paper stems from the imperative to tackle the challenge of achieving group consensus control in multi-agent systems (MASs). Specifically, it focuses on the complexities and practical considerations inherent in MASs, such as the impact of competition and the influence of outdated states and input delays on system performance and stability. To address this issue, the paper introduces a novel group consensus protocol designed for second-order multi-agent systems that closely resemble real-world scenarios. This protocol integrates both position and velocity information as control inputs, enabling consensus under competitive conditions. By combining real-time and outdated states, the paper establishes conditions for achieving group consensus convergence while improving the bounds on input delays. Additionally, the paper extends the traditional bipartite graph representation to a directed graph, relaxing conservative assumptions and validating the effectiveness of the proposed protocol through theoretical analysis, numerical simulations, and comparative analysis with existing literature.}, archive = {J_TASE}, author = {Shi Wang and Jie Chen and Jiacheng Su and Yaonan Wang}, doi = {10.1109/TASE.2025.3535829}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11483--11489}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed group consensus control for multi-agent systems based on combined state information}, volume = {22}, year = {2025}, }
• (2025). A resource-efficient decentralized sequential planner for spatiotemporal wildfire mitigation. TASE, 22, 11469--11482. (www)

  @article{ , abstract = {This paper proposes a Conflict-aware Resource-Efficient Decentralized Sequential planner (CREDS) for early wildfire mitigation using multiple heterogeneous Unmanned Aerial Vehicles (UAVs). Multi-UAV wildfire management scenarios are non-stationary, with spatially clustered dynamically spreading fires, potential pop-up fires, and partial observability due to limited UAV numbers and sensing range. The objective of CREDS is to detect and sequentially mitigate all growing fires as Single-UAV Tasks (SUT) while adhering to the physical constraints of UAV. CREDS minimizes biodiversity loss through rapid UAV intervention and promotes efficient resource utilization by avoiding complex multi-UAV coordination. CREDS employs a three-phased approach, beginning with fire detection using a search algorithm, followed by local trajectory generation using the auction-based Resource-Efficient Decentralized Sequential planner (REDS), incorporating the novel non-stationary cost function, the Deadline-Prioritized Mitigation Cost (DPMC). Finally, a conflict-aware consensus algorithm resolves conflicts to determine a global trajectory for spatiotemporal mitigation. The performance evaluation of the CREDS for partial and full observability conditions with both heterogeneous and homogeneous UAV teams for different fires-to-UAV ratios demonstrates a 100% success rate for ratios up to 4 and a high success rate for the critical ratio of 5, outperforming baselines. Heterogeneous UAV teams outperform homogeneous teams in handling heterogeneous deadlines of SUT mitigation. CREDS exhibits scalability and 100% convergence, demonstrating robustness against potential deadlock assignments, enhancing its success rate compared to the baseline approaches.Note to Practitioners—Practical wildfire scenarios often involve unknown clusters of rapidly evolving fires that exceed the available firefighting resources. Wildfire scenarios often involve vast areas and limited sensor capabilities of UAVs, resulting in partial information about the environment. When the number of fires exceeds the number of UAVs, decentralized sequential action becomes necessary. Early wildfire mitigation, focusing on containing fires within the quenching capability of a single UAV, is crucial for minimizing damage and efficient resource utilization by avoiding complex multi-UAV coordination. The approach of single UAV mitigation introduces physical constraints and deadlines for initiating mitigation efforts. The deadlines vary based on factors like fire area, spread rate, and quench rate. The computation of a quenching sequence with efficient prioritization of deadlines ensures mission success and reduction in the total destroyed area. The challenges involved in wildfire scenarios necessitate a three-phased framework: a search stage to locate fires using thermal sensors and cameras, a Resource-Efficient Decentralized Sequential planner to assign local trajectories (mitigation sequence) for each UAV, and a conflict resolution stage to ensure smooth operation by resolving potential conflicts between UAV trajectories. CREDS prioritizes deadlines and achieves successful missions even when fires outnumber UAVs by five times. Additionally, heterogeneous UAV teams with diverse quench and speed capabilities outperform homogeneous teams and traditional methods focused on execution time. This approach is well-suited for scenarios with spatially distributed, dynamic targets with diverse deadlines.}, archive = {J_TASE}, author = {Josy John and Shridhar Velhal and Suresh Sundaram}, doi = {10.1109/TASE.2025.3536356}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11469--11482}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A resource-efficient decentralized sequential planner for spatiotemporal wildfire mitigation}, volume = {22}, year = {2025}, }
• (2025). Vision-based nut quality classification using conditional GAN and CNN. TASE, 22, 11455--11468. (www)

  @article{ , abstract = {In this study, the quality of a nut is discussed by considering images of the internal thread, and an analysis is conducted using traditional machine-learning and deep-learning algorithms. Compared to the traditional contact methods, the vision-based method has the advantage of fast computing speed and is not affected by the conditions of tapping speed. The pitch and pitch diameter of the internal thread are the indicators that characterize nut quality classification. For one nut, 36 internal thread images are collected, one image per 10 degrees, by the self-designed laser triangulation measurement platform. Using the laser triangulation method, the information on both indicators can be obtained and analyzed. In the traditional machine-learning methods, the internal thread images undergo several preprocessing procedures to obtain the region of interest and calculate the depth between the crest and root. Subsequently, 33 handcrafted features are used to extract the features from the 36 processed images. Finally, the features are classified by three families of machine-learning algorithms, including support vector machines, k-nearest neighbors, and decision trees. In the deep-learning method, conditional generative adversarial network and convolutional neural network (CNN) are used for data augmentation and nut quality classification, respectively. The experimental results show that the proposed CNN model can achieve a higher classification accuracy rate. Furthermore, the proposed CNN model trained with the generated images is better equipped to detect the nut quality under different decision thresholds. Note to Practitioners—This study explores the assessment of nut quality, particularly focusing on the internal thread. Traditional methods for grading thread quality often involve contact-based detection, which may risk damaging the thread surface. Additionally, most non-contact methods tend to have prolonged classification processes. This study introduces a vision-based nut-quality classification method based on the Japanese Industrial Standard (JIS) specification, offering advantages such as rapid computation and independence from tapping speed conditions. Internal thread images are collected using a self-designed laser triangulation measurement platform, and various learning algorithms are employed for analysis. In traditional machine-learning approaches, we propose an image preprocessing method to extract 33 statistical features from internal thread images. Feature-importance analysis, calculated from random forest (RF) and gradient boosting (GB), reveals the significance of features such as crest-to-root depth variation between teeth and the brightness of the crest. By performing feature reduction, some machine-learning algorithms can improve the classification accuracy rate and AUC of the model. In the realm of deep learning, we utilize conditional generative adversarial network (CGAN) to generate internal thread images and employ CNN for nut quality classification. Experimental results demonstrate that the proposed CNN model achieves higher classification accuracy with fewer training parameters compared to VGG16, VGG19, ResNet50, and Xception. The deep-learning models, particularly the CNN, outperform traditional machine-learning methods without the need for manual feature extraction. CGAN successfully augments the training dataset, and the models can detect nut quality under different decision thresholds. Although this proposed nut-quality analysis method achieves non-contact detection, it has not yet attained full automation. Future research will endeavor to integrate automated mechanisms for placing nuts onto the inspection platform and to conduct real-time image analysis for continuous assessment of nut quality.}, archive = {J_TASE}, author = {Kuei-Jung Hung and Tzu-Chen Lee and Chiao-Sheng Wang and Tsung-Chun Lin and Chen-Wei Conan Guo and Der-Min Tsay and Jau-Woei Perng}, doi = {10.1109/TASE.2025.3533013}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11455--11468}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Vision-based nut quality classification using conditional GAN and CNN}, volume = {22}, year = {2025}, }
• (2025). Given-performance PID-SMC for 4-DOF tower crane systems under input constraints. TASE, 22, 11444--11454. (www)

  @article{ , abstract = {This paper focuses on the rapid jib and trolley positioning and payload sway suppression of the 4-degree of freedom (4-DOF) tower crane systems under uncertain system dynamics, external disturbances and control input constraints. A new adaptive proportional-integral-derivative sliding mode control (PID-SMC) method is proposed, which is model-free, does not need to linearize the system, and dispense with the uncertainty upper bound information required by traditional sliding mode control (SMC). In particular, a new time-varying scale function is used to constrain the system error to ensure the given-performance, that is, the actual transient-state and steady-state control performance of the system can be predetermined according to practical application requirements, and the quantified steady-state and transient-state properties of the 4-DOF tower crane system are obtained for the first time. In addition, although the actuator saturation upper bound is completely unknown the above mentioned given-performance can also be guaranteed by designing new parameter adaptive law. Finally, the effectiveness and superior performance of the control method are verified through rigorous theoretical analysis and experiments conducted on a 4-DOF tower crane platform. Note to Practitioners—The primary objective of this study is to tackle the challenges associated with rapid target positioning and effective suppression of swing in 4-DOF tower crane systems, amid presence of multiple practical problems. Traditional SMC strategies typically presume the prior knowledge of upper bounds for uncertainties—a condition that is seldom met in actual operations. Furthermore, their disregard for transient performance criteria further curtails their applicability in real-life scenarios. To overcome these limitations, we integrate an adaptive law for the estimation of previously indeterminable uncertainty boundaries. The incorporation of a performance function also allows us to impose stringent controls over both transient-state and steady-state performance, thereby enhancing operational efficiency and guaranteeing safety. Moreover, to address the issue of actuator saturation, this paper utilizes a parameter adaptive strategy that maximizes the utilization of the actuator’s potential without resorting to overly cautious presumptions regarding controller limits. In our future work, we plan to apply and validate this control methodology in actual tower crane applications.}, archive = {J_TASE}, author = {Yana Yang and Tengfei Zhang and Xiaoshuang Zhou and Changchun Hua and Junpeng Li}, doi = {10.1109/TASE.2025.3531463}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11444--11454}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Given-performance PID-SMC for 4-DOF tower crane systems under input constraints}, volume = {22}, year = {2025}, }
• (2025). DEED-ADMM: A scalable distributed algorithm for economic dispatch in multi-energy systems with energy storage. TASE, 22, 11431--11443. (www)

  @article{ , abstract = {Multi-energy systems with energy storage can coordinate various energy carriers to facilitate the integration of large amounts of distributed energy sources and promote the overall efficiency of energy use, which needs distributed dispatch with the requirement of security and privacy. This paper studies the distributed economic dispatch based on information from neighboring agents only. In order to handle the non-convexity due to the complementarity constraint of energy storage, it is proved that simultaneous charging and discharging is suboptimal for the multi-energy systems. Based on this, an equivalent convex problem is reformulated. A scalable distributed algorithm based on parallel ADMM and dynamic consensus mechanism, termed DEED-ADMM, is then proposed. It is shown that DEED-ADMM is scalable in terms of per-agent energy consumption and computational complexity with centralized methods. Moreover, under general convex cost functions, convergence properties of DEED-ADMM are theoretically analyzed by adopting the Lyapunov-based approach. It is proved that the primal problem and the dual problem can be simultaneously solved. Finally, case studies demonstrate the effectiveness of the proposed algorithm. Note to Practitioners—This paper is motivated by the problem of coordinating various energy carriers as well as energy storage in multi-energy systems to promote the overall efficiency of energy use. The coupling among different energy carriers and the complementarity constraint of non-simultaneous charging and discharging of battery storage make the problem non-convex. Existing distributed approaches require stringent assumptions on the cost functions, or suffer from a heavy computational burden. To address the above challenges, a fully distributed algorithm is developed, which is scalable and suitable for large-scale systems. Moreover, it is the first distributed algorithm that solves the economic dispatch problem and the dual problem simultaneously in multi-energy systems with general convex cost functions. Practitioners can easily adjust the coefficients of the proposed algorithm to guarantee convergence for the IEEE 30-bus or even 116-bus systems, as long as the economic dispatch problem is feasible. Our future work will focus on designing resilient mechanisms under potential attacks and considering more practical situations such as power loss.}, archive = {J_TASE}, author = {Shanying Zhu and Tao Ding and Cailian Chen and Mo-Yuen Chow and Xinping Guan}, doi = {10.1109/TASE.2025.3534171}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11431--11443}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DEED-ADMM: A scalable distributed algorithm for economic dispatch in multi-energy systems with energy storage}, volume = {22}, year = {2025}, }
• (2025). Automated surface patch extraction for 3D printing qualification. TASE, 22, 11419--11430. (www)

  @article{ , abstract = {Surface patches have been utilized to reduce shape complexity in qualification of product geometric quality. However, specifying surface patches case-by-case is impractical for qualifying 3D-printed products with complicated freeform designs. Automating patch extraction must overcome the issue of potentially infinite variety of surface patches. To achieve dimension reduction for automated qualification, this work defines and characterizes surface patches using Laplace-Beltrami (LB) operator and critical points to capture a finite number of patch deviation patterns. The deviation-pattern-driven characterization enables the automated determination of patch centroids through active landmark selection. The patch sizes are determined using the LB operator within a changepoint detection formulation. To verify the finite dimensionality of patch types, patches extracted from product designs are clustered based on geometric dissimilarity that is quantified by wave kernel and curvature signatures. To verify that extracted patches capable of capturing deviation patterns in printed products, we derive patch deviation signatures that are invariant to printing covariates to facilitate product qualification. Analysis of actual 3D-printed freeform products demonstrates the efficacy of the developed methodology by comparing with existing approaches. It also shows the potential of inferring deviation patterns of a new design using the surface patch characterization and extraction approach. Note to Practitioners—Qualification of geometric quality for products with complex geometries relies on specifying regions of interest in the form of features or surface patches. In 3D printing, the potentially infinite variety of product designs requires manual specification of context-dependent surface patches for new and unseen designs. This work establishes an automated framework to reduce the dimensionality of qualification by specifying finite types of surface patches and automating their extraction across an infinite variety of designs. This approach streamlines the qualification process and will support the inference and learning of geometric quality of previously unseen designs.}, archive = {J_TASE}, author = {Weizhi Lin and Qiang Huang}, doi = {10.1109/TASE.2025.3535900}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11419--11430}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated surface patch extraction for 3D printing qualification}, volume = {22}, year = {2025}, }
• (2025). Deadzone-modified robust adaptive learning bipartite consensus for heterogeneous nonlinear multiagent systems. TASE, 22, 11405--11418. (www)

  @article{ , abstract = {In this paper, the robust adaptive learning bipartite consensus problem for heterogeneous nonlinear multiagent systems with unknown control gains, external disturbances, and actuator constraints under signed directed graphs is investigated. A novel neural distributed protocol is proposed, whose key techniques lie in the introduction of deadzone-modified Lyapunov functions and integral Lyapunov functions into the command filtered backstepping design. The former enhances the robustness of the undertaken system by attenuating the impact of complex uncertainties and transforms the robustness problem into a convergence one by introducing the deadzones, while the latter efficiently tackles both the state-dependent control gain functions and the input saturation nonlinearities, simplifying the consensus design significantly. In addition, the requirement of the command filtered design for the control gain functions is relaxed by the appropriate system transformation. Furthermore, the incremental adaptive algorithm takes the place of the integral adaptation for parameter learning, avoiding numerical integration in implementation. The theoretical results of the performance analysis are presented in detail, in which the boundedness of all closed-loop variables is examined, and the asymptotic consensus is achieved, in the sense that the bipartite synchronization error converges to a pre-specified region asymptotically. Numerical results verify the feasibility of the presented scheme. Note to Practitioners—This paper is devoted to the problem of robust adaptive learning bipartite consensus for heterogeneous nonlinear multiagent systems (MASs). It is important to investigate heterogeneous MASs in practical engineering applications, and a typical example is the air-ground cooperative combat system consisting of unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs). In addition, the bipartite consensus allows MASs to accomplish more diverse tasks. However, uncertain nonlinearities, external disturbances, and actuator constraints are widespread in system dynamics. Moreover, the unknown control gains make consensus design challenging. These issues are well addressed by adopting the key techniques including deadzone-modified strategy, integral Lyapunov synthesis and incremental adaptive learning mechanism. Furthermore, the proposed scheme guarantees the asymptotic convergence of the bipartite synchronization errors with a pre-specified accuracy, and enhances the system robustness under complex heterogeneous nonlinearities. Therefore, the presented method contributes to practical applications.}, archive = {J_TASE}, author = {Shengxiang Zou and Mingxuan Sun and Guomin Zhong and Xiongxiong He}, doi = {10.1109/TASE.2025.3535516}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11405--11418}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deadzone-modified robust adaptive learning bipartite consensus for heterogeneous nonlinear multiagent systems}, volume = {22}, year = {2025}, }
• (2025). Long-term water quality prediction with transformer-based spatial-temporal graph fusion. TASE, 22, 11392--11404. (www)

  @article{ , abstract = {Over the past decades of rapid development, the global water pollution problem became prominent. Accurate water quality prediction can detect the trend and anomaly of water quality changes in advance, thereby taking timely measures to avoid water quality problems. Traditional statistical methods for water quality prediction tend to fail to capture the complex relationship among multiple water quality variables. Deep learning models face a challenge to capture both temporal dependence and spatial correlation of the water quality series data. To solve the above problems, this work proposes an adaptive and dynamic graph fusion water quality prediction model based on a spatiotemporal attention mechanism named Spatial-Temporal Graph Fusion Transformer (STGFT). It integrates a spatial attention encoder, a temporal attention encoder, an adaptive dynamic adjacency matrix generator, and a multi-graph fusion layer. Among them, the first two are adopted to capture the spatial correlations and temporal characteristics among different water quality monitoring stations, respectively. The generator can produce adaptive and dynamic adjacency matrices to reflect potential spatial relationships in a river network. Experimental results with real-life water quality datasets reveal that the prediction accuracy of STGFT outperforms the existing state-of-the-art models. Note to Practitioners—This paper is motivated by the problem of long-term water quality prediction. The highly volatile water quality data and the nonlinear characteristics of the time series greatly affect the accuracy of the forecasting task. Existing approaches fail to simultaneously capture spatial correlations and temporal characteristics among different water quality monitoring stations, affecting the accuracy of water quality predictions. This work proposes a water quality prediction method that captures the spatial correlations and temporal characteristics among different water quality monitoring stations. Moreover, it produces adaptive and dynamic adjacency matrices to reflect potential spatial relationships in a river network. Experimental results from three real-world datasets show that this approach is feasible and obtains more accurate prediction results. Furthermore, this method can also be applied to other areas of time series prediction, including finance, traffic, and smart manufacturing.}, archive = {J_TASE}, author = {Jing Bi and Ziqi Wang and Haitao Yuan and Xiangxi Wu and Renren Wu and Jia Zhang and MengChu Zhou}, doi = {10.1109/TASE.2025.3535415}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11392--11404}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Long-term water quality prediction with transformer-based spatial-temporal graph fusion}, volume = {22}, year = {2025}, }
• (2025). Sample-efficient deep reinforcement learning of mobile manipulation for 6-DOF trajectory following. TASE, 22, 11381--11391. (www)

  @article{ , abstract = {The whole-body control of mobile manipulators for the 6-DOF trajectory following task is the basis of many continuous tasks. However, traditional control strategies rely on accurate models and expert knowledge for solving the trajectory following task. Deep reinforcement learning (DRL) provides a promising model-free solution, but it is sample-inefficient. To this end, we propose Trajectory Following Hindsight Experience Replay (TF-HER), a sample-efficient DRL algorithm for the whole-body coupled trajectory following task with dense rewards. TF-HER builds a multi-trajectory state space, and relabels the low-reward data to generate informative high-reward experiences. Also, the distributional shift caused by the relabeling is corrected by estimating the density ratio of relabeled experiences. Extensive demonstrations on both nonholonomic and holonomic bases in simulation validate that our algorithm can accelerate the model convergence and significantly improve the sample efficiency. Furthermore, we present real-world experiments to demonstrate the effectiveness of our approach. The code is available: https://github.com/IRMV-Manipulation-Group/TF-HER.Note to Practitioners—The whole-body 6-DOF trajectory tracking capability for mobile manipulators is crucial in industrial automation, serving as the foundation for a wide range of continuous and precise operations including automated assembly, welding, and material handling. This paper proposes a reinforcement learning approach to enhance the efficiency and effectiveness of mobile manipulators, requiring no prior model information. Beyond this specific application, this research has promising applications in other robotic systems because it is model-free. By utilizing a gradient-descent-based relabeling method and an adaptive density ratio estimator, we address distributional shifts and mitigate hindsight bias, guiding the mobile manipulator to accurately follow complex 6-DOF trajectories with dense rewards. The experimental results highlight our method’s superior performance over existing model-free techniques, with improved sample efficiency and reduced tracking error across diverse robotic platforms in both simulated and real-world settings. The successful policy transfer from simulation to physical robots with fine-tuning further validates the robustness and practical applicability.}, archive = {J_TASE}, author = {Yifan Zhou and Qiyu Feng and Yixuan Zhou and Jianghao Lin and Zhe Liu and Hesheng Wang}, doi = {10.1109/TASE.2025.3530162}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11381--11391}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sample-efficient deep reinforcement learning of mobile manipulation for 6-DOF trajectory following}, volume = {22}, year = {2025}, }
• (2025). Physically informed synchronic-adaptive learning for industrial systems modeling in heterogeneous media with unavailable time-varying interface. TASE, 22, 11369--11380. (www)

  @article{ , abstract = {Partial differential equations (PDEs) are commonly employed to model complex industrial systems characterized by multivariable dependence. However, existing physics-informed neural networks (PINNs) barely perform well in heterogeneous media modeling compared to their achievement for a homogeneous medium. Unknown PDE parameters due to insufficient prior knowledge with respect to physical attributes and unavailable time-varying interface caused by heterogeneous media may weaken PINNs feasibility. To this end, physically informed synchronic-adaptive learning (PISAL) is proposed. First, PISAL is proposed for learning the solutions and interface satisfying PDEs, in which $Net_{1}$ , $Net_{2}$ , and $Net_{I}$ are constructed. $Net_{1}$ and $Net_{2}$ are for synchronically learning the solutions satisfying PDEs with diverse parameters; $Net_{I}$ is for adaptively learning the interface to decompose the domain with heterogeneous media. Then, a criterion combined with the output of neural networks is introduced to adaptively distinguish the attributes of measurements and collocation points. Furthermore, $Net_{1}$ , $Net_{2}$ , and $Net_{I}$ are integrated into a data-physics-hybrid loss function. Accordingly, a synchronic-adaptive learning (SAL) strategy is proposed to decompose the domain and optimize the three networks by iteratively erasing the training errors. Besides, we theoretically prove the proposed PISAL can iteratively approximate the fields with diverse physical attributes. Finally, extensive experimental results and comparisons with relevant state-of-the-art methods verify the feasibility and effectiveness of PISAL for industrial systems modeling in heterogeneous media. Note to Practitioners—The motivation behind this paper is to devise a method for industrial systems modeling, even in cases where unknown PDE parameters are caused by a lack of prior knowledge with respect to physical attributes and the unavailable time-varying interface is caused by heterogeneous media. Existing methods apply the domain decomposition technique under the assumption that the interface is available. To this end, a data-physics-hybrid method, PISAL, in which $Net_{1}$ , $Net_{2}$ , and $Net_{I}$ with SAL strategy are proposed. $Net_{1}$ , $Net_{2}$ , and $Net_{I}$ are first constructed. $Net_{1}$ and $Net_{2}$ are for synchronically learning the solutions satisfying PDEs with diverse parameters; $Net_{I}$ is for adaptively learning the unavailable time-varying interface. Subsequently, a criterion combined with the output of neural networks is introduced, which is to adaptively distinguish different physical attributes of measurements and collocation points. Additionally, the three neural networks are integrated into a data-physics-hybrid loss function. Accordingly, a SAL strategy is proposed to decompose the domain and optimize the three neural networks. Besides, the approximation capability of PISAL is theoretically proved based on well-posedness and denseness. To validate the efficacy of the proposed PISAL, the two-phase Stefan problem and the mixed Navier-Stokes problem are employed. Meanwhile, some comparisons with relevant state-of-the-art approaches are given. Results highlight the feasibility of our method in industrial systems modeling with the above-mentioned challenges. Thus, the proposed method is suitable for industrial automation applications. In future research, we intend to conduct the proposed PISAL on an experimental platform to further verify the feasibility in practical scenarios and more complex applications.}, archive = {J_TASE}, author = {Aina Wang and Pan Qin and Xi-Ming Sun}, doi = {10.1109/TASE.2025.3529578}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11369--11380}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Physically informed synchronic-adaptive learning for industrial systems modeling in heterogeneous media with unavailable time-varying interface}, volume = {22}, year = {2025}, }
• (2025). Resilient adaptive event-triggered control of nonlinear DC-microgrids under DoS attacks: Local stabilization approach. TASE, 22, 11356--11368. (www)

  @article{ , abstract = {This paper tackles the challenge of local stabilization control in nonlinear DC-microgrids (DC-MGs) under the threat of energy-constrained Denial-of-Service (DoS) attacks, using a dynamic resilient event-triggered mechanism (DRETM) and a fuzzy-basis-dependent Lyapunov approach. In contrast to previous studies, this paper introduces the operating range of system states, given when transforming the nonlinear DC-MG system into a Takagi-Sugeno (T-S) fuzzy model, as a pivotal constraint, aiming to prevent undesired system behaviors. Moreover, as part of an initial effort to analyze the impact of both DoS attacks and event-driven data transfer schemes on the local stabilization problem, this paper exploits the features of energy-limited DoS attacks and DRETM when deriving necessary conditions to ensure the local operation of the system state. Especially, to enhance the efficiency of the proposed method while minimizing conservatism and computational complexity, this paper eliminates certain unnecessary state constraints typically required when dealing with a closed operating range. Lastly, the validity of the proposed method is illustrated through two numerical examples. Note to Practitioners—As is widely recognized, the T-S fuzzy model has been extensively employed for efficiently capturing the nonlinear dynamics of DC-MG. However, the limitations of prior studies lie in their adherence to the system operating range assumption, rendering their findings unable to ensure the stability of the closed-loop system when this assumption is violated. To address this issue, this paper introduces a local stabilization control approach within the fuzzy-based Lyapunov function framework. This approach plays a crucial role in eliminating the need for operating range assumptions and achieving less conservative results. As a result, practitioners can now confidently utilize our improved approach to effectively stabilize nonlinear DC-MGs without the risks of unexpected behaviors. In addition, this paper confronts the task of devising an event-triggered controller while taking into account energy-limited DoS attacks and resource wastage in DC-MGs, such that: 1) the state trajectory of DC-MG T-S fuzzy systems remains in the operating region regardless of energy-limited DoS attacks, and 2) eventually converges to the equilibrium point without requiring any prior information about DoS attacks. Finally, two numerical examples demonstrate the validity of the proposed method.}, archive = {J_TASE}, author = {Gia Bao Hong and Sung Hyun Kim}, doi = {10.1109/TASE.2025.3532087}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11356--11368}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient adaptive event-triggered control of nonlinear DC-microgrids under DoS attacks: Local stabilization approach}, volume = {22}, year = {2025}, }
• (2025). Collaborative and distributed bayesian optimization via consensus. TASE, 22, 11343--11355. (www)

  @article{ , abstract = {Optimal design is a critical yet challenging task within many applications. This challenge arises from the need for extensive trial and error, often done through simulations or running field experiments. Fortunately, sequential optimal design, also referred to as Bayesian optimization when using surrogates with a Bayesian flavor, has played a key role in accelerating the design process through efficient sequential sampling strategies. However, a key opportunity exists nowadays. The increased connectivity of edge devices sets forth a new collaborative paradigm for Bayesian optimization. A paradigm whereby different clients collaboratively borrow strength from each other by effectively distributing their experimentation efforts to improve and fast-track their optimal design process. To this end, we bring the notion of consensus to Bayesian optimization, where clients agree (i.e., reach a consensus) on their next-to-sample designs. Our approach provides a generic and flexible framework that can incorporate different collaboration mechanisms. In lieu of this, we propose transitional collaborative mechanisms where clients initially rely more on each other to maneuver through the early stages with scant data, then, at the late stages, focus on their own objectives to get client-specific solutions. Theoretically, we show the sub-linear growth in regret for our proposed framework. Empirically, through simulated datasets and a real-world collaborative sensor design experiment, we show that our framework can effectively accelerate and improve the optimal design process and benefit all participants. Note to Practitioners—The proposed algorithm allows multiple clients to collaboratively distribute their trial-and-error efforts to fast-track and improve the optimal design process. In the algorithm, each client performs a test locally and then shares the results with an orchestrator. Using the information from all clients, the orchestrator then finds the best new experiment that each client should undertake and sends those back for the next round of experiments. Through this process, all clients can leverage each other’s strengths and optimize their designs with far fewer experiments than each client operating in isolation. This is confirmed through many simulation examples, along with a real-life sensor design experiment where multiple collaborating agents seqeuntially coordinate their experimentation efforts. The goal is to rapidly discover the biosensor design and measurement format parameters that find the maximum amount of captured target analyte.}, archive = {J_TASE}, author = {Xubo Yue and Yang Liu and Albert S. Berahas and Blake N. Johnson and Raed Al Kontar}, doi = {10.1109/TASE.2025.3529349}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11343--11355}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Collaborative and distributed bayesian optimization via consensus}, volume = {22}, year = {2025}, }
• (2025). Adaptive event-triggered control for wastewater treatment process using self-organizing fuzzy neural network. TASE, 22, 11333--11342. (www)

  @article{ , abstract = {In a wastewater treatment process (WWTP), which covers a variety of physical, chemical, and biological treatment steps, it is a challenging control task to maintain proper dissolved oxygen (DO) and nitrate nitrogen (NO) concentrations to meet effluent standards. This paper proposes an event-triggered adaptive control method using a self-organizing fuzzy neural network (ETSOFNN) to regulate DO and NO concentrations. Moreover, a self-organizing fuzzy neural network (SOFNN) based on the maximum correlation entropy-induced criterion identifies and approximates the nonlinear function, which further enables the dynamic adjustment of the controller structure, including the addition or deletion of parameters. In addition, a dynamic event-triggered mechanism with a relative threshold strategy is introduced into the controller, and the trigger conditions are designed according to the tracking error. Utilizing Lyapunov stability theory, the stability of the control system is demonstrated. Finally, simulations based on the benchmark simulation model 1 (BSM1) platform are conducted to verify the effectiveness of the ETSOFNN method. Note to Practitioners—In this paper, the aim is to develop an online tracking and control method for controlling dissolved oxygen (DO) concentration and nitrate nitrogen (NO) concentration in wastewater treatment process (WWTP) and to make it applicable to wastewater treatment plants under a wide range of operating conditions. Due to the dynamic and complex nature of WWTP, which involves multiple treatment units and control variables, as well as characteristics such as strong coupling, non-Gaussian characteristics, and significant nonlinearity, there are considerable challenges in accurately establishing mathematical models. Therefore, this paper investigates a dynamic event-triggered mechanism and utilizes a self-organizing fuzzy neural network (SOFNN) applied to the control method of WWTP. Firstly, a mathematical model of WWTP is established. Secondly, using the SOFNN based on correlation entropy compensation, the structure of the controller is automatically constructed, which enables more accurate adjustment depending on the actual working conditions to improve the robustness and adaptability of the controller. Next, a dynamic event-triggered strategy is applied in the controller to reduce the number of controller updates and thus reduce the communication cost. Finally, the results of simulation experiments show that this control method performs well, and the strategy will be extended to actual wastewater treatment plants in the future to optimize the operation results.}, archive = {J_TASE}, author = {Dong-Juan Li and Yi-Fan Yan and Dapeng Li}, doi = {10.1109/TASE.2025.3533020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11333--11342}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive event-triggered control for wastewater treatment process using self-organizing fuzzy neural network}, volume = {22}, year = {2025}, }
• (2025). Digital twin enabled flight control system testing: Design, development, and implementation. TASE, 22, 11318--11332. (www)

  @article{ , abstract = {Flight control system testing (FCST) is one of the most important process to check whether flight control surfaces can operate properly according to commands during aircraft assembly. Traditional testing method relies heavily on manual labor, leading to low efficiency and inconsistent quality. In this paper, we apply digital twin (DT) technology to the FCST process for the first time. We firstly design an architecture of DT-enabled FCST including four layers to support further development. Then, we present a triangular mesh alignment-based angle measurement (TMA-AM) algorithm to efficiently collect deflection angle data for DT-enabled FCST. Extensive experiments conducted on a aircraft wing subassembly platform show that the TMA-AM algorithm achieves an average angular measurement error of less than 0.1°, outperforming existing methods. Moreover, we develop a virtual experimental platform named DT-FCST aligned with a real aircraft wing subassembly platform. In addition, TMA-AM algorithm is integrated with the DT-FCST platform. By integrating real-time data from the cockpit, real-time physical-virtual interaction of aircraft control sticks and flight control surfaces are achieved, ensuring consistency between physical and virtual movements. The integration of DT technology with the TMA-AM algorithm enables real-time synchronization, monitoring, and unified data management, significantly enhancing the efficiency and accuracy of the FCST. Note to Practitioners—To address the inefficiencies and low monitoring quality associated with traditional manual testing methods in flight control system testing (FCST), we firstly introduce digital twin (DT) technology to this process. To support effective and accurate measurement during the FCST, we propose a vision-based method tailored to accurately measure deflection angles of flight control surfaces. This method replaces manual measurements with a non-contact approach, significantly improving measurement accuracy and efficiency. We provide a detailed description of the construction process of the DT-FCST platform including requirement analysis, DT model construction, and on-site experiments. This DT-based approach achieves real-time synchronization between virtual and physical testing processes, enhancing monitoring quality and overall testing effectiveness. Specifically, it can achieve a 90% reduction in the number of operators and shorten the single testing time to 16.7% of the traditional testing method.}, archive = {J_TASE}, author = {Cheng Ren and Jiaxin Xu and Cailian Chen and Shanying Zhu and Yehan Ma and Xinping Guan}, doi = {10.1109/TASE.2025.3532632}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11318--11332}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Digital twin enabled flight control system testing: Design, development, and implementation}, volume = {22}, year = {2025}, }
• (2025). Guided bayesian optimization: Data-efficient controller tuning with digital twin. TASE, 22, 11304--11317. (www)

  @article{ , abstract = {This article presents the guided Bayesian optimization (BO) algorithm as an efficient data-driven method for iteratively tuning closed-loop controller parameters using a digital twin of the system. The digital twin is built using closed-loop data acquired during standard BO iterations, and activated when the uncertainty in the Gaussian Process model of the optimization objective on the real system is high. We define a controller tuning framework independent of the controller or the plant structure. Our proposed methodology is model-free, making it suitable for nonlinear and unmodelled plants with measurement noise. The objective function consists of performance metrics modeled by Gaussian processes. We utilize the available information in the closed-loop system to progressively maintain a digital twin that guides the optimizer, improving the data efficiency of our method. Switching the digital twin on and off is triggered by our data-driven criteria related to the digital twin’s uncertainty estimations in the BO tuning framework. Effectively, it replaces much of the exploration of the real system with exploration performed on the digital twin. We analyze the properties of our method in simulation and demonstrate its performance on two real closed-loop systems with different plant and controller structures. The experimental results show that our method requires fewer experiments on the physical plant than Bayesian optimization to find the optimal controller parameters. Note to Practitioners—Industrial applications typically are difficult to model due to disturbances. Bayesian optimization is a data-efficient iterative tuning method for a black box system in which the performance can only be measured given the control parameters. Iterative measurements involve operational costs. We propose a guided Bayesian optimization method that uses all information flow in a system to define a simplified digital twin of the system using out-of-the-box methods. It is continuously updated with data from the system. We use the digital twin instead of the real system to perform experiments and to find optimal controller parameters while we monitor the uncertainty of the resulting predictions. When the uncertainty exceeds a tolerance threshold, the real system is measured, and the digital twin is updated. This results in fewer experiments on the real system only when needed. We then demonstrate the improved data efficiency of the guided Bayesian optimization on real-time linear and rotary motor hardware. These common industrial plants need to be controlled rigorously in a closed-loop system. Our method requires 57% and 46% fewer experiments on the hardware than Bayesian optimization to tune the control parameters of the linear and rotary motor systems. Our generic approach is not limited to the controller parameters but also can optimize the parameters of a manufacturing process.}, archive = {J_TASE}, author = {Mahdi Nobar and Jürg Keller and Alisa Rupenyan and Mohammad Khosravi and John Lygeros}, doi = {10.1109/TASE.2024.3454176}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11304--11317}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guided bayesian optimization: Data-efficient controller tuning with digital twin}, volume = {22}, year = {2025}, }
• (2025). A digital twin driven human-centric ecosystem for industry 5.0. TASE, 22, 11291--11303. (www)

  @article{ , abstract = {Industry 5.0 embodies the vision for the future of factories, emphasizing the importance of sustainable industrialization and the role of industry in society, through the key concept of placing the well-being of workers at the center of the production process. Building upon this vision, we propose a new paradigm to design human-centric industrial applications. To this end, we exploit Digital Twin (DT) technology to build a digital replica for each entity on the shop floor and support and augment interaction among workers and machines. While so far DTs in automation have been proposed for machine digitalization, the core element of the proposed approach is the Operator Digital Twin (ODT). In this scenario, biometrics allows to build a reliable model of those operator’s characteristics that are relevant in working contexts. Biometric traits are measured and processed to detect physical, emotional, and mental conditions, which are used to define the operator’s state. Perspectively, this allows to manage and monitor production and processes in an operator-in-the-loop manner, where not only is the operator aware of the state of the plant, but also any technological agent in the plant acts and reacts according to the operator’s needs and conditions. In this paper, we define the modeling of the envisioned ecosystem, present the designed DT’s blue-print architecture, discuss its implementation in relevant application scenarios, and report an example of implementation in a collaborative robotics scenario.Note to Practitioners—This paper was motivated by the problem of designing human-cyber-physical systems, where production processes are managed by concurrently taking into account operators, machines and plant status. This answers the needs of the novel Industry 5.0 paradigm, which aims to enhance social sustainability of modern factories. To this end, we propose an architecture based on digital twins that allows to develop a digital layer, detached from the physical one, where the plant can be monitored and managed. This allows the creation of a digital ecosystem where machines, operators, and the interactions among them are represented, augmented, and managed. We discuss how the proposed architecture can be applied to three relevant scenarios: remote training and maintenance, line operation and line supervision. Moreover, the implementation in a collaborative robotics scenario is presented, to provide an example of the proposed architecture can be implemented in industrial scenarios.}, archive = {J_TASE}, author = {Valeria Villani and Marco Picone and Marco Mamei and Lorenzo Sabattini}, doi = {10.1109/TASE.2024.3410703}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11291--11303}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A digital twin driven human-centric ecosystem for industry 5.0}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Engineering and operating digital twins for automated production or construction systems. TASE, 22, 11288--11290. (www)

  @article{ , archive = {J_TASE}, author = {Birgit Vogel-Heuser and Min-Hsiung Hung and Manuel Wimmer and Ilya Kovalenko and Xun Xu}, doi = {10.1109/TASE.2025.3550678}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11288--11290}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Engineering and operating digital twins for automated production or construction systems}, volume = {22}, year = {2025}, }
• (2025). Test automation for deterministic communication networks using chained orthogonal arrays. TASE, 22, 11276--11287. (www)

  @article{ , abstract = {arinc 664 is one of the most popular Ethernet-based deterministic network (dtn) technologies for safety-critical avionics systems. A typical arinc 664 network can support several thousands of Virtual Links (vls), each of which has hundreds of thousands of combinations for its configuration. Orthogonal Arrays (oas) method is a formalism to generate a test suite with a feasible size and a desired coverage of features for complex systems. Since the test space for dtn features is prohibitively large, a single application of oas may inadvertently eliminate many combinations of two-way interactions. In this paper, we propose to develop a test automation system that uses multiple and consecutive applications of oas to generate test configurations for dtn. The test factors representing dtn configuration parameters are divided into two groups, where the first group includes the number of vls passing through each port while the second group includes the configuration parameters for vls. Then, oas are constructed sequentially for each group such that an output element of the first oa is used as one of the inputs for the second oa. We present a system architecture for a test automation system capable of verifying real-time network operations, where the outputs of the chained oas are utilized to generate executable test scripts including dtn configurations and traffics. As a case study, a test automation system is successfully implemented for arinc 664switch. Note to Practitioners—Safety-critical avionics networks with a large number of configuration parameters need to be tested rigorously. Exhaustive testing of all possible configurations is not realistic since there are time and resource constraints. Therefore, a chained orthogonal array is proposed to create different sub-configuration spaces of the test parameters for dtn. An automation system is mandatory to generate the repeatable test cases fast and make them simpler to implement and run. In this article, a test automation system from generating the test cases using the chained orthogonal arrays to carrying out the tests on the prototype system is presented. We believe that the concepts presented in this article can be utilized by other researchers developing a test automation for complex systems having a large number of parameters.}, archive = {J_TASE}, author = {Muhammed Efe Altuntaş and İbrahim Hökelek and M. Selim Demir and Muhammed Şeker and Pınar Savaş and Eyüp Can Akpolat and Sabri Barış Yavuztürk and Erhan Akdoğan}, doi = {10.1109/TASE.2025.3532672}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11276--11287}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Test automation for deterministic communication networks using chained orthogonal arrays}, volume = {22}, year = {2025}, }
• (2025). LESO-based NMPC tracking control of climbing robot on large components with variable curvature. TASE, 22, 11264--11275. (www)

  @article{ , abstract = {Wheeled climbing robots have great application prospects in the machining of large components with variable curvature. However, its accurate motion control on variable curvature surfaces faces two fatal challenges. The varied contact states between the robot’s wheels and the variable curvature surfaces make it difficult to establish an accurate kinematics model. Additionally, there exists a different degree of robot slippage when the robot moves in different attitudes due to the dragging effect of gravity. To overcome the above problems, we first present a kinematics modeling method with instantaneous plane constraints on a variable curvature surface. Subsequently, a linear extended state observer (LESO)-based nonlinear model predictive control (NMPC) scheme is designed, in which the NMPC is used to calculate the nominal control inputs and the LESO is used to estimate and compensate for lumped disturbance brought by the robot slippage and surface constraints. Experiments on a real wind turbine blade with variable curvature show that the proposed control scheme can well eliminate the influence of lumped disturbance, and the climbing robots can achieve unbiased (AVG < 0.1 mm) and high-precision (RMSE < 2 mm) trajectory tracking. Note to Practitioners—Wheeled climbing robots, capable of adhering to curved surfaces of workpieces while in motion, provide a new approach for machining large components when equipped with machining actuators. This paper is motivated by the tracking control problem of wheeled climbing robots on curved surfaces with variable curvature, which is crucial for ensuring effective machining. The varying contact states between the wheels and the surfaces during the robot’s motion, along with the robot slippage introduced by the dragging effect of gravity, result in discrepancies between the ideal and actual motion speeds. These discrepancies introduce model uncertainties and thus pose challenges for achieving high-precision trajectory tracking. Additionally, current research on climbing robots primarily focuses on prototype development, with practitioners predominantly utilizing open-loop controller or basic model-free control schemes. This often leads to non-robust trajectory tracking of the climbing robot. To address these problems, this paper offers an accurate kinematics modeling method considering instantaneous plane constraints of the curved surfaces and a LESO-based NMPC scheme which can effectively reduce the impact of external disturbances. This approach provides a solution for automatic and high-precision trajectory tracking for climbing robot with localization system in factories, offering the potential for high-quality machining of large components.}, archive = {J_TASE}, author = {Ke Tan and Zeyu Gong and Bo Tao and Yuhao Zhang and Ying Shi and Zhenfeng Gu and Chong Wu and Han Ding}, doi = {10.1109/TASE.2025.3532418}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11264--11275}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {LESO-based NMPC tracking control of climbing robot on large components with variable curvature}, volume = {22}, year = {2025}, }
• (2025). A macro-micro vision integrated micromanipulation system for self-initialization and resilient control. TASE, 22, 11250--11263. (www)

  @article{ , abstract = {Robotic micromanipulation systems (RMS) enable precise and repeatable operations under a microscope. Traditional RMS rely solely on microscopic visual feedback, necessitating time-consuming manual positioning to bring the tool tip within the microscope field-of-view (Micro-FOV), which limits efficiency and heavily depends on operator skill. This paper proposes an innovative RMS that integrates macro and micro vision to automate the aforementioned tool tip positioning and facilitate resilient control. The system utilizes an external camera to obtain the macro field-of-view (Macro-FOV), containing the tool and fiducial markers, and estimates the tool tip’s 3D position by triangulation. Visual servoing is then used to guide the tool tip towards the Micro-FOV. Under the Micro-FOV, a tool-sweep detector based on partitioned difference images is used to sequentially locate the tool’s shaft and tip. After auto-focusing, the system executes tool tip and Petri dish resilient control based on our developed self-calibration and self-recalibration mechanisms. During the operation, the system provides an intuitive user interface that includes both macro and micro information, improving the visualization and productivity of micromanipulation. Experiments show that the self-initialization scheme can be implemented across different macro camera viewpoints, reducing the average tip positioning time from 65.70 to 50.08 seconds compared to manual operation, thereby decreasing manual labor intensity and improving efficiency. The self-recalibration mechanism achieves precision and resilient control, with an average error of $0.95~\mu $ m over 25 continuous trials. Additionally, the system exhibits robustness against vibration and visual interference, underscoring its potential for diverse biomedical applications.Note to Practitioners—In the biomedical field, robotic micromanipulation systems (RMS) are valued for their high precision and repeatability. Existing research on RMS typically focuses on achieving varying degrees of automation using visual feedback from the microscope, with the assumption that the tool tip is already within the Micro-FOV prior to the task. However, the initial step of moving the tool tip from the Macro-FOV to the Micro-FOV and eventually bringing it to focus usually requires time-consuming manual operation that highly depends on the skill of individual operators. To achieve automatic positioning and resilient control of the tool tip, this paper presents an RMS that integrates macro and micro vision. The system combines visual feedback from a macro camera and a microscope to complete the self-initialization of the system in three steps: macro tip positioning, micro tip positioning, and self-calibration. After self-initialization, the system provides the operator with an intuitive cursor-based user interface. During operation, the system uses visual feedback to detect the tip position and ensures the control accuracy through a self-recalibration mechanism. The proposed self-initialization method has the potential for a wide range of applications. It can be extended to micromanipulation systems with different types of end-effectors, thereby improving the efficiency and precision of micromanipulation.}, archive = {J_TASE}, author = {Tiexin Wang and Yun Long and Tianle Weng and Liangjing Yang}, doi = {10.1109/TASE.2025.3532214}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11250--11263}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A macro-micro vision integrated micromanipulation system for self-initialization and resilient control}, volume = {22}, year = {2025}, }
• (2025). Active diagnosis of time-interval automata: Time perspectives. TASE, 22, 11239--11249. (www)

  @article{ , abstract = {Language diagnosability captures the capability of the system to detect faults based on observations. When a system is not diagnosable, supervisory control can be used to enforce its diagnosability to prevent the faults from occurring silently, known as the active diagnosis problem. Note that in the context of timed discrete-event systems, an observation contains not only the sequence of the events but also their time information. Therefore, two sequences consisting of the same events but with different occurrence time instants can still reveal the occurrence of faults. This fact motivates us to consider enforcing the diagnosability of a timed discrete-event system by regulating the occurrence time instants of certain controllable events. In this paper, we first construct a verifier for a time-interval automaton to verify its diagnosability. Then, based on the verifier, we enforce the diagnosability of a time-interval automaton by restricting the time intervals of certain controllable events and disabling some controllable events. Note to Practitioners—Fault diagnosis and active diagnosis play a critical role in ensuring the reliability, safety, and efficiency of systems across various industries, ranging from automotive and aerospace to healthcare and smart grids. Discrete-event systems, as general models for complex man-made systems, are well-studied for modeling digital computer systems in the above scenarios. Early detection and correction of faults contribute to improved performance, reduced downtime, and enhanced overall system functionality. This work investigates the active diagnosis problem, i.e., design a supervisor to enforce diagnosability, for discrete-event systems modeled by time-interval automata. The control policy combines the time and logical information of the system, thereby allowing the closed-loop systems to retain more of the original system behavior. Time-interval automata is a model that is not complicated but is closer to actual engineering systems than finite automata, providing new insights for control engineers in modeling and control.}, archive = {J_TASE}, author = {Shaowen Miao and Jan Komenda and Aiwen Lai}, doi = {10.1109/TASE.2025.3531787}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11239--11249}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active diagnosis of time-interval automata: Time perspectives}, volume = {22}, year = {2025}, }
• (2025). A distributed framework for integrated task allocation and safe coordination in networked multi-robot systems. TASE, 22, 11219--11238. (www)

  @article{ , abstract = {Deploying a team of autonomous robots, operating collaboratively towards a common objective within dynamic environments, has the potential to improve the system efficiency across several fields. This paper proposes a distributed comprehensive framework enabling a networked multi-robot system to serve time-varying requests arising from different locations within the environment in a distributed and safe manner, i.e., by guaranteeing no collisions with possible obstacles and preserving connectivity among the robots. To this aim, a two-layer architecture is proposed where the top layer is in charge of distributively assigning new service requests to the robots by resorting to an auction-based algorithm, while the bottom layer is in charge of safely navigating the environment to serve the assigned requests by relying on Control Barrier Functions. However, the presence of connectivity constraints might affect the number of service requests that the multi-robot system can handle simultaneously and might lead to deadlock situations where robots cannot reach the designated locations due to loss of network connectivity. Hence, a distributed strategy based on consensus algorithms to detect and solve deadlocks in a distributed fashion is proposed. The completeness of the approach is proved. Simulation results in an agricultural setting and real-world laboratory experiments are provided to validate the effectiveness of the proposed approach.Note to Practitioners—This paper was inspired by the necessity to coordinate a team of robots to perform tasks within an unstructured agricultural field, including both the decision-making and navigation strategies, with no central control unit as envisioned by the European project CANOPIES. To this aim, a distributed approach is designed where robots only rely on local data and information from neighboring robots to assign and execute tasks effectively in a coordinated manner. In addition, as working under local communication constraints may prevent parallel execution of all tasks, potentially leading to deadlock situations, a distributed strategy is developed to enable each robot to detect and solve such situations. The proposed approach can be employed in several domains where the cooperation of multiple autonomous robots might be beneficial, ranging from logistics settings to search and rescue scenarios up to agricultural environments. Laboratory experiments with three robots demonstrate the effectiveness of the approach.}, archive = {J_TASE}, author = {Andrea Miele and Martina Lippi and Andrea Gasparri}, doi = {10.1109/TASE.2025.3532023}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11219--11238}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A distributed framework for integrated task allocation and safe coordination in networked multi-robot systems}, volume = {22}, year = {2025}, }
• (2025). Oral-anatomical knowledge-informed semi-supervised learning for 3D dental CBCT segmentation and lesion detection. TASE, 22, 11205--11218. (www)

  @article{ , abstract = {Cone beam computed tomography (CBCT) is a widely-used imaging modality in dental healthcare. It is an important task to segment each 3D CBCT image, which involves labeling lesions, bones, teeth, and restorative materials on a voxel-by-voxel basis, as it aids in lesion detection, diagnosis, and treatment planning. The current clinical practice relies on manual segmentation, which is labor-intensive and demands considerable expertise. Leveraging Artificial Intelligence (AI) to fully automate the segmentation process could tremendously improve the quality and efficiency of dental healthcare. The main hurdle in this advancement is reducing AI’s reliance on a large quantity of manually labeled images to train robust, accurate, and generalizable algorithms. To tackle this challenge, we propose a novel Oral-Anatomical Knowledge-informed Semi-Supervised Learning (OAK-SSL) model for 3D CBCT image segmentation and lesion detection. The uniqueness of OAK-SSL is its capability of integrating qualitative oral-anatomical knowledge of plausible lesion locations into the deep learning design. Specifically, the unique design of OAK-SSL includes three key elements, including transformation of qualitative knowledge into quantitative representation, knowledge-informed dual-task learning architecture, and knowledge-informed semi-supervised loss function. We apply OAK-SSL to a real-world dataset, focusing on segmenting CBCT images that contain small lesions. This task is inherently challenging yet holds significant clinical value as treating lesions at their early stages lead to excellent prognosis. OAK-SSL demonstrated significantly better performance than a range of existing methods. Note to Practitioners—This study tackles the challenges arising from a limited amount of labeled data due to the time-consuming manual segmentation of 3D dental cone beam computed tomography (CBCT) images. The scarcity of labeled data often impedes AI models from accurately segmenting periapical lesions. To overcome this, we introduce a novel semi-supervised learning algorithm that integrates the oral-anatomical knowledge about lesion location for 3D CBCT image segmentation. Our method effectively segments periapical lesions, including even small-sized periapical lesions, without solely relying on labeled data. The proposed method offers two significant benefits to clinicians. First, it reduces the necessity for large amounts of labeled data, particularly easing the burden of manually segmenting early-stage periapical lesions. Second, it helps reduce intra- and inter-observer disagreements and human errors by providing consistent and automated segmentation maps. These benefits not only simplify the segmentation process in dental imaging but also improve its reliability. As a result, our automated algorithm makes it easier and more trustworthy for practitioners. However, practitioners should be aware that the effectiveness of our method relies on the assumption of consistency between labeled and unlabeled data. When applying this method, it is crucial to carefully consider the characteristics of the unlabeled dataset. Significant differences in image quality, patient demographics, or acquisition parameters between labeled and unlabeled data might affect model performance.}, archive = {J_TASE}, author = {Yeonju Lee and Min Gu Kwak and Rui Qi Chen and Hao Yan and Muralidhar Mupparapu and Fleming Lure and Frank C. Setzer and Jing Li}, doi = {10.1109/TASE.2025.3530936}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11205--11218}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Oral-anatomical knowledge-informed semi-supervised learning for 3D dental CBCT segmentation and lesion detection}, volume = {22}, year = {2025}, }
• (2025). Concurrent-learning based relative localization in shape formation of robot swarms. TASE, 22, 11188--11204. (www)

  @article{ , abstract = {In this article, we address the shape formation problem for massive robot swarms in environments where external localization systems are unavailable. Achieving this task effectively with solely onboard measurements is still scarcely explored and faces some practical challenges. To solve this challenging problem, we propose the following novel results. Firstly, to estimate the relative positions among neighboring robots, a concurrent-learning based estimator is proposed. It relaxes the persistent excitation condition required in the classical ones such as the least-square estimator. Secondly, we introduce a finite-time agreement protocol to determine the shape location. This is achieved by estimating the relative position between each robot and a randomly assigned seed robot. The initial position of the seed one marks the shape location. Thirdly, based on the theoretical results of the relative localization, a novel behavior-based control strategy is devised. This strategy not only enables the adaptive shape formation of large groups of robots but also enhances the observability of inter-robot relative localization. Numerical simulation results are provided to verify the performance of our proposed strategy compared to the state-of-the-art ones. Additionally, outdoor experiments on real robots further demonstrate the practical effectiveness and robustness of our methods. Note to Practitioners—Shape formation has a broad potential for large groups of robots to execute certain tasks, such as object transport, forest firefighting, and entertainment shows. However, most of the existing approaches rely on external localization infrastructures, rendering them impractical in environments where such systems are not available. To address this issue, this article proposes an integrated strategy that can achieve shape formation for large groups of robots by using local distance and displacement measurements. This strategy consists of three main components. Firstly, a relative localization estimator is introduced to estimate the relative positions among neighboring robots. Secondly, a protocol for reaching a consensus on the desired shape’s position is proposed. Thirdly, a behavior-based controller is developed to achieve massive shape formation and enhance the observability of relative localization. More details of the proposed algorithms and swarm robotic systems are provided in this article.}, archive = {J_TASE}, author = {Jinhu Lü and Kunrui Ze and Shuoyu Yue and Kexin Liu and Wei Wang and Guibin Sun}, doi = {10.1109/TASE.2025.3533478}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11188--11204}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Concurrent-learning based relative localization in shape formation of robot swarms}, volume = {22}, year = {2025}, }
• (2025). Secure control of T–S fuzzy-based nonlinear active quarter-vehicle suspension systems under malicious attacks with experimental validation. TASE, 22, 11176--11187. (www)

  @article{ , abstract = {It is challenging to ensure the safety of vehicles in the presence of malicious attacks. For this reason, the security control problem of nonlinear active quarter-vehicle suspension systems (QVSSs) is investigated in this paper. Within transportation cyber-physical systems, a specific Takagi-Sugeno fuzzy representation is adopted to capture the nonlinear behavior of vehicle dynamics. Firstly, accurately capturing the randomness and variability of attacks within complex driving environments poses a challenge. To precisely model the randomness of attacks, a probabilistically uncertain denial of service (PUDoS) attack strategy is adopted, thereby providing a practical range for potential attacks. Secondly, a novel probability-dependent homogeneous polynomial non-quadratic control law (PHNQCL) is designed. On the one hand, the high-order character of the controller enables the introduction of more groups of gain matrices, enhancing the control flexibility and reducing the conservatism. On the other hand, the designed controller exhibits strong defense capabilities against PUDoS attacks. Subsequently, feasible criteria for the PHNQCL are established using a high-order Lyapunov function, which not only stabilizes the active QVSSs but also reduces the conservatism. Finally, hardware-in-the-loop tests serve to validate the feasibility of the proposed method. Note to Practitioners—As a critical part of automotive chassis, the active quarter-vehicle suspension systems (QVSSs) are crucial for improving ride comfort and ensuring handling stability. However, in real-world environments, external disturbances, uncertainties, and nonlinearities can degrade their performance. As networks evolve, the interconnections in active QVSSs increase. This makes communication channels more vulnerable to attacks, further compromising vehicle safety. This paper explores security control for nonlinear QVSSs, modeling them using T-S fuzzy logic to capture vehicle dynamics. To simulate a complex vehicle driving environment, a probabilistically uncertain DoS attack strategy is designed. The challenge is to design new control methods to mitigate the performance degradation caused by attacks while enhancing vehicle safety in diverse scenarios. Hardware-in-the-loop tests are performed to verify that the proposed control algorithm can provide strong operational stability, ride comfort, and protection of vehicle components.}, archive = {J_TASE}, author = {Mengni Du and Xiang-Peng Xie and Jian Wu and Heng Wang}, doi = {10.1109/TASE.2025.3531928}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11176--11187}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure control of T–S fuzzy-based nonlinear active quarter-vehicle suspension systems under malicious attacks with experimental validation}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered consensus control of multiagent systems under markov switching topology. TASE, 22, 11163--11175. (www)

  @article{ , abstract = {This paper studies the dynamic event-triggered consensus problem of multiagent systems (MASs) under Markov switching topology, where a new distributed dynamic event-triggered mechanism (DETM) is proposed and the Zeno behavior is eliminated. It adds an exponential function and anti-cotangent function which increases the threshold of the system error and slows down the attenuation rate of the dynamic variable. So the proposed DETM further effectively reduces the triggering times. In addition, considering the variability of the topology of MASs in practical applications, the DETM is applied to multiagent network with Markov switching topology. A new Lyapunov function is constructed to prove the stability of the MAS. Then the matrix equation and other conditions are used to ensure that the MAS can reach the mean square consensus. Finally, the effectiveness of the proposed strategy is verified through the simulations on a theory system and multi aircrafts system. Note to Practitioners—In the practical application of MASs, the communication topology which connects agents may be affected by external environments (such as external noise, system component failures, etc.), leading to changes of the topology structure. In other words, the topology structure of MASs may undergo jumps during operation, so the MASs with Markov switching topology considered in this paper is more practical. In the paper, the dynamic event-triggered consensus problem of MASs under Markov switching topology is considered, and it can be applied to some practical systems, e.g., multi aircrafts. Compared with the existing results, a new Lyapunov function is proposed for consensus analysis of MAS with Markov switching topology, and a novel DETM is proposed, which is more easier to be implemented.}, archive = {J_TASE}, author = {Fang Wang and Zhen Hu and Chao Zhou and Xiaonan Lin}, doi = {10.1109/TASE.2025.3531900}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11163--11175}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered consensus control of multiagent systems under markov switching topology}, volume = {22}, year = {2025}, }
• (2025). An intermittent volatile event-triggered synchronization approach for multi-layer networks with noise coupling under an almost sure framework. TASE, 22, 11150--11162. (www)

  @article{ , abstract = {In this paper, the almost sure synchronization of stochastic multi-layer networks with noise coupling is studied by means of pinning intermittent volatile event-triggered control. During the control interval, the determination of control updates follows an event-triggered mechanism with waiting time, which avoids continuous monitoring and also eliminates Zeno behavior. In contrast to the existing literature that focuses on moment synchronization, we study almost sure synchronization on stochastic multi-layer networks, where noise coupling with time-varying and nonlinear features plays an active role. In addition, the control gain displays sign-indefinite characteristics that follow some volatility patterns, including synchronizing control and desynchronizing inputs. A concept of average volatile control gain is presented to quantify the control gain. To cope with the challenges posed by volatile gain and intermittent control input, a generalization of Halanay-type inequality is proposed, its coefficients are time-varying and piecewise continuous, which shows that the results have a wider range of applications. Based on the stochastic analysis technique, graph theory, and Lyapunov method, the synchronization criteria are established. Finally, the feasibility is illustrated by simulation examples. Note to Practitioners—This paper was motivated by existing results on pinning intermittent control and event-triggered control about almost sure synchronization. The existing results mainly require that the control signals were updated in a consecutive manner during control activation intervals, which may hardly be implemented on digital computers. This paper designs a novel hybrid control method named pinning intermittent volatile event-triggered control, and therefore, the proposed approach is more friendly for control engineers. In addition, the volatile control gain is considered. The results obtained are applied to a spring-mass-damper system, demonstrating their effectiveness and it is expected that the proposed approach can be extended to more practical physical engineering systems.}, archive = {J_TASE}, author = {Dongsheng Xu and Shuting Song and Choon Ki Ahn and Huan Su}, doi = {10.1109/TASE.2025.3532350}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11150--11162}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An intermittent volatile event-triggered synchronization approach for multi-layer networks with noise coupling under an almost sure framework}, volume = {22}, year = {2025}, }
• (2025). Threshold-dependent secure event-triggered control for networked systems under denial of service attacks. TASE, 22, 11139--11149. (www)

  @article{ , abstract = {Event-triggered control addresses a communication resource constraint challenge for networked control systems (NCS) but it is vulnerable to denial of service (DoS) attacks. This paper is concerned with the secure event-triggered control problem for NCSs under DoS attacks. The novelty lies in the development of a novel threshold-dependent event-triggered transmission scheme and a co-design approach for the event-triggered scheme and the desired secure controller. First, two threshold concepts, namely the critical event-triggered threshold (CETT) and the secure event-triggered threshold (SETT), are proposed for the event-triggered scheme. More specifically, the CETT is defined by the maximum allowable event-triggered parameter which can guarantee stability of the NCS and the SETT is used to ensure timely transmissions when the NCS subject to DoS-resulted packet dropouts. Second, by virtue of the input delay approach, a comprehensive closed-loop system model is derived for the NCS, which accommodates simultaneously the event-based sporadic transmissions, packet delays, and DoS-resulted packet dropouts. Third, a co-design algorithm is presented to determine both the desired threshold parameter and secure state- and output-feedback secure controllers. Finally, numerical simulations on a CAN-based vehicle lateral control system are provided to demonstrate the efficacy of the proposed control method. Note to Practitioners—This paper was motivated by the secure control design problem of the NCSs subject to both limited communication resources and DoS attacks. Most existing works failed to establish the relation between event-triggered parameter and packet dropouts and thus lead to the difficulties in controller synthesis under DoS attacks. In this paper, we mathematically characterize the relation between event-triggered parameter and the number of DoS-resulted packet dropouts. The secure control and event-triggered communication co-design approach of the networked control systems under malicious attacks is carefully derived. To practitioners, this study provides a secure control design basis of the NCS subject to DoS attacks by analyzing the evolution of event-triggered threshold under DoS-resulted packet dropouts. However, such design method just mitigates the effects caused by DoS attacks but is less of ability to regulate the control actions during DoS attack intervals. In future research, the active and intelligent secure control compensation strategies will be addressed.}, archive = {J_TASE}, author = {Hong-Tao Sun and Chen Peng and Xiaohua Ge}, doi = {10.1109/TASE.2025.3531944}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11139--11149}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Threshold-dependent secure event-triggered control for networked systems under denial of service attacks}, volume = {22}, year = {2025}, }
• (2025). AutoLfD: Closing the loop for learning from demonstrations. TASE, 22, 11124--11138. (www)

  @article{ , abstract = {Over the past few years, there have been numerous works towards advancing the generalization capability of robots, among which learning from demonstrations (LfD) has drawn much attention by virtue of its user-friendly and data-efficient nature. While many LfD solutions have been reported, a key question has not been properly addressed: how can we evaluate the generalization performance of LfD? For instance, when a robot draws a letter that needs to pass through new desired points, how does it ensure the new trajectory maintains a similar shape to the demonstration? This question becomes more relevant when a new task is significantly far from the demonstrated region. To tackle this issue, a user often resorts to manual tuning of the hyperparameters of an LfD approach until a satisfactory trajectory is attained. In this paper, we aim to provide closed-loop evaluative feedback for LfD and optimize LfD in an automatic fashion. Specifically, we consider dynamical movement primitives (DMP) and kernelized movement primitives (KMP) as examples and develop a generic optimization framework capable of measuring the generalization performance of DMP and KMP and auto-optimizing their hyperparameters. Evaluations including peg-in-hole, block-stacking and pushing tasks on a real robot evidence the applicability of our framework. Note to Practitioners—The paper is motivated by the demand to transfer human skills to robots. While the problems of ‘what to learn’ and ‘how to learn’ have been long-standing research topics, the solutions for evaluating the quality of such skill transfer remain largely open. We introduce a novel closed-loop framework towards transferring human skills to robots in an automatic manner. Specifically, we collect a training dataset that reflects user preference for trajectory adaptation and train a trajectory encoder network using the dataset. With the encoder network, we design a robust metric to measure the skill transfer quality and subsequently employ the metric to guide imitation learning of human skills. By using our framework, unseen robotic tasks can be tackled by adapting the demonstrations straightforwardly, where relevant hyperparameters involved in skill transfer are optimized automatically.}, archive = {J_TASE}, author = {Shaokang Wu and Yijin Wang and Yanlong Huang}, doi = {10.1109/TASE.2025.3532820}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11124--11138}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AutoLfD: Closing the loop for learning from demonstrations}, volume = {22}, year = {2025}, }
• (2025). An intelligent microscope vision system for hooking multi-thread flexible microelectrode based on few-shot segmentation. TASE, 22, 11112--11123. (www)

  @article{ , abstract = {Multi-thread flexible microelectrodes hold the promise of high-quality and long-term neural signal recording. To implant fragile microelectrodes into brain, a microneedle is usually used as a shuttle to increase stiffness. However, hooking microelectrodes with a microneedle is error-prone and time-consuming, due to the difficulty in accurately obtaining micro-object coordinates through low-quality microscope images. To solve this problem, this paper proposes an intelligent microscope vision system based on few-shot segmentation method. Firstly, a stereo vision system is designed to enable cameras to focus on all microelectrode threads simultaneously. Secondly, a few-shot instance segmentation method is proposed to obtain key-point coordinates for vision guidance, which consists of data augmentation, improved Mask-RCNN and anchor-free clustering loss. The data augmentation method synthesizes low-quality images to strengthen model’s generalization ability. To improve the proposal quality, a residual region proposal module is introduced into Mask-RCNN. By clustering instance features in metric space, the anchor-free clustering loss enhances the model capability of predicting hard instances and avoids the intra-class bias of fixed anchor. Experimental result shows that the proposed method achieves 99.72% and 90% correct rate for microelectrode and microneedle segmentation with 5-shot training. Code is available at https://github.com/NamingIsEasy/FSS_implantation. Note to Practitioners—Multi-thread flexible microelectrodes are an emerging brain-computer interface platform. Compared with traditional electrodes, they have the advantages of low elastic modulus, micron-level width and material biocompatibility. However, the low elastic modulus also makes it difficult to be inserted into brain. Thus, researchers proposed “sewing machine” paradigm, in which a microneedle is used to hook microelectrodes and drive them to be inserted into brain. The hooking process is threading the microneedle into engaging holes at the end of microelectrodes. However, undesirable conditions in microscope images, such as out-of-focus blur and occlusion, make it difficult to accurately obtain microneedle and microelectrode coordinates. Towards automated hooking operation, this paper proposed an intelligent microscope vision system based on few-shot segmentation. Firstly, a stereo vision system is designed to focus cameras on all microelectrode threads simultaneously. Secondly, a few-shot segmentation method is proposed to reduce time of image acquisition and annotation process. It consists of data augmentation, improved Mask-RCNN and anchor-free clustering loss. The data augmentation method synthesizes undesirable-condition images to reduce the bias of training dataset. The classical region proposal module in Mask-RCNN is replaced with an improved residual version. To make the model focus more on hard instances during training, an anchor-free clustering loss is introduced. The proposed method achieves 99.72% and 90% correct rate for microelectrode and microneedle segmentation in 5-shot experiments. In future, we will explore automated hooking control and subsequent implantation process based on the proposed vision system.}, archive = {J_TASE}, author = {Bo Han and Hanwei Chen and Chao Liu and Xinjun Sheng}, doi = {10.1109/TASE.2025.3532674}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11112--11123}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An intelligent microscope vision system for hooking multi-thread flexible microelectrode based on few-shot segmentation}, volume = {22}, year = {2025}, }
• (2025). Time consistent surface mapping for deformable object shape control. TASE, 22, 11099--11111. (www)

  @article{ , abstract = {Shape control involves deforming objects to achieve a desired shape. One of the main challenges of this task is to define a suitable control reference, especially when 3D objects that lack distinctive visual texture and geometric features are involved. This paper addresses the problem of generating a suitable shape control reference using surface maps for 3D texture-less objects. The proposed surface mapping method is based on functional maps and ensures time consistency with robustness to non-isometries. Our time consistent method is validated within a shape control strategy, where local exponential stability analysis is provided. The effectiveness of the framework is illustrated through simulations and experiments. Note to Practitioners—We present a method for shape comparison to analyse the geometric similarities between the shape of an object and a desired target shape. The goal is to generate a point map between both surfaces so that, with the use of robots that grasp the object, a control strategy can deform the object towards the desired shape. The point map we compute, while adapting to the changing shape of the deforming object, remains stable enough to enable the shape control task. This can be of use for the automation of processes that involve shape control (e.g., object packaging, moulding, etc.). Our proposed shape control framework combines computer graphics, computer vision, and automation techniques to create a system that is well-suited for industrial setups equipped with commonly used range sensors like RGB-D cameras. Regarding the experiments presented in this paper, controlled lighting conditions are recommendable in order to perform the colour-based object segmentation (e.g. avoidance of abrupt light variations, uniform illumination). Automating the grasping process is not within the scope of this paper, therefore, in each experiment we manually defined the object grasping points according to the shape control task involved. The proposed framework has the potential to increase productivity, reduce costs and, improve safety in hazardous tasks by automating the manipulation of 3D objects that lack distinctive visual texture and geometric features.}, archive = {J_TASE}, author = {Ignacio Cuiral-Zueco and Gonzalo López-Nicolás}, doi = {10.1109/TASE.2025.3529180}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11099--11111}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time consistent surface mapping for deformable object shape control}, volume = {22}, year = {2025}, }
• (2025). An initial-rectifying-constraint optimization scheme for repetitive motion planning by using prescribed-time zeroing neural networks. TASE, 22, 11087--11098. (www)

  @article{ , abstract = {To realize repetitive motion planning (RMP) for redundant robot manipulators in the presence of initial shifts, this paper introduces an initial-rectifying-constraint optimization scheme that reformulates the problem as a constrained time varying quadratic programming task. Using the initial-rectifying zeroing neural network (IRZNN) proposed in this article, the initial-rectifying-constraint optimization problem can be solved, and a repeatable solution with prescribed-time convergence can be obtained. The polynomial rectifying functions are developed to maintain smooth operation of manipulators during rectification. Convergence of the IRZNN model in solving the initial-rectifying-constraint optimization scheme is analyzed. The proposed scheme takes the initial shift problem into account, realizes RMP for redundant manipulators, and achieves prescribed-time convergence of the end-effector position error. The simulation and experiment results validate the effectiveness and practicality of the proposed approach. Note to Practitioners—The motivation of this article is the joint-angular-drift phenomenon found in closed trajectory tracking of redundant manipulators. This phenomenon, where a closed path of the end-effector may result in non-closed joint motion, is regarded as a potential menace in the operation of redundant manipulators. Existing solutions assume the initial state of joints aligns with the desired one, which is too idealistic and usually requires additional adjustment to meet the assumption. This article suggests a novel RMP scheme to alleviate the joint-angular-drift phenomenon in the presence of initial joint shifts. The proposed scheme enables the specification of the settling time for the end-effector position error in accordance with task requirements. To be more practical, polynomial rectifying functions are formed to facilitate smooth operation of robotic manipulators.}, archive = {J_TASE}, author = {Guomin Zhong and Liming Wang and Mingxuan Sun}, doi = {10.1109/TASE.2025.3531002}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11087--11098}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An initial-rectifying-constraint optimization scheme for repetitive motion planning by using prescribed-time zeroing neural networks}, volume = {22}, year = {2025}, }
• (2025). Human-in-the-loop robot learning for smart manufacturing: A human-centric perspective. TASE, 22, 11062--11086. (www)

  @article{ , abstract = {Robot learning has attracted an ever-increasing attention by automating complex tasks, reducing errors, and increasing production speed and flexibility, which leads to significant advancements in manufacturing intelligence. However, its low training efficiency, limited real-time feedback, and challenges in adapting to untrained scenarios hinder its applications in smart manufacturing. Introducing a human role in the training loop, a practice known as human-in-the-loop (HITL) robot learning, can improve the performance of robots by leveraging human prior knowledge. Nonetheless, the exploration of HITL robot learning within the context of human-centric smart manufacturing remains in its infancy. This study provides a holistic literature review for understanding HITL robot learning within an industrial context from a human-centric perspective. A united structure is presented to encompass different aspects of human intelligence in HITL robot learning, highlighting perception, cognition, behavior, and notably, empathy. Then, the typical applications in manufacturing scenarios are analyzed to expand the research landscape for smart manufacturing. Finally, it introduces the empirical challenges and future directions for HITL robot learning in the next industrial revolution era. Note to Practitioners—This review is motivated by the emergence of the next generation of smart manufacturing, which emphasizes the coexistence of humans and robotics in the manufacturing workstation to mitigate inherent limitations of each. It presents an overview of HITL robot learning-related works to identify state-of-the-art and significant focuses for human-centric smart manufacturing. It classifies representative studies into detailed sub-categories based on various facets of human intelligence, highlighting perception, cognition, behavior, and empathy, providing a complete and detailed survey of this field. The applications in manufacturing scenarios are analyzed, and we discuss the possible challenges and future directions. This paradigm has the potential to revolutionize manufacturing operations, enhancing flexibility, and resilience in supply chains, and efficiency for self-organizing collaborative intelligence and cyber-physical systems toward human-robot coevolution. The goal is to attract scholars in broader research fields to contribute to the development of HITL robot learning for smart manufacturing.}, archive = {J_TASE}, author = {Hongpeng Chen and Shufei Li and Junming Fan and Anqing Duan and Chenguang Yang and David Navarro-Alarcon and Pai Zheng}, doi = {10.1109/TASE.2025.3528051}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11062--11086}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human-in-the-loop robot learning for smart manufacturing: A human-centric perspective}, volume = {22}, year = {2025}, }
• (2025). A fast and precise piezoelectric-stage positioning scheme using a novel modified lemniscate scanning trajectory with shunt-based model predictive control and long short-term memory. TASE, 22, 11050--11061. (www)

  @article{ , abstract = {Piezoelectric stages are widely used in the field of high-precision measurement where controllers are designed for the stages to track reference trajectories. Most existing scanning trajectories tend to trigger mechanical vibrations or yield distorted images resulting in mapping errors. In addition, the performance of many existing controllers in tackling nonlinearities are, to a large extent, not satisfactory. To improve the situation, this paper aims to develop a modified lemniscate scanning trajectory combined with a shunt-based model predictive control and long short-term memory neural network to alleviate mechanical vibrations yielding a higher scanning rate and tracking accuracy. Experimental results are presented to validate the design and verify its effectiveness in which the proposed scheme outperforms the proportional-integral-derivative and model predictive control in terms of significant tracking error reductions. Quantitatively, vibration improvement of the modified lemniscate waveform over the raster’s triangular waveform in power spectrum analysis in terms of average decibel level is reduced from −49.67dB to −73.18dB. Percentage wise, performance improvement of the proposed controller over MPC in the same trajectory tracking at 10Hz the root-mean-square errors are reduced from 9.217% and 9.362% to 8.064% and 8.395% in X-axis and reduced from 7.431% and 7.153% to 5.557% and 4.694% in Y-axis, respectively. The root-mean-square errors at 50Hz are reduced from 43.015% and 46.071% to 35.394% and 37.746% in X-axis and reduced from 35.659% and 35.215% to 33.497% and 30.252% in Y-axis, respectively. Note to Practitioners—This paper was motivated by the problem of scanning speed and accuracy of piezoelectric stage-type measurement systems, such as atomic force microscopy. First, a novel modified lemniscate trajectory is presented to replace conventional trajectories that can minimize the image distortion thanks to reduction of the mapping errors of scanning paths as well as alleviation of mechanical oscillations of high-frequency resonances. Furthermore, the paper develops a model predictive control combined with a long short-term memory neural network that can efficiently deal with the nonlinear behavior of the piezoelectric stage to reduce unwanted path tracking errors. Experiments are conducted to validate the method and assess the effectiveness of the proposed trajectory and the control scheme. The preliminary results support the feasibility of the proposed approach, but since the proposed trajectory uses two curved lines at corners, it does require a more extended stroke stage.}, archive = {J_TASE}, author = {Jim-Wei Wu and Jui-Tse Weng and Ting-Kuei Hsu and Shih-Hao Chien and Jen-Te Yu}, doi = {10.1109/TASE.2025.3531109}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11050--11061}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A fast and precise piezoelectric-stage positioning scheme using a novel modified lemniscate scanning trajectory with shunt-based model predictive control and long short-term memory}, volume = {22}, year = {2025}, }
• (2025). STDSD: A spacecraft target detection framework considering similarity and diversity. TASE, 22, 11038--11049. (www)

  @article{ , abstract = {Benefiting from the development of deep learning and computer vision, object detection performance has witnessed remarkable improvements in various domains recently. Nonetheless, several challenges persist for spacecraft detection, including challenging space environment, domain gaps, and real-time application. This paper proposes a Spacecraft Target Detection framework considering Similarity and Diversity strategy (STDSD), specifically designed to adapt and excel in the special requirements of space object detection. The STDSD framework integrates multiple feature extraction networks (FNets) and overlapping vectors prediction networks (ONets). Initially, the original images are cropped into several segments, which are processed through the FNets to extract critical features. These features are subsequently input to the ONets, where adversarial learning and random dropout techniques are implemented to enhance network diversity. This strategy of incorporating both similarity and diversity significantly enhances the precision and robustness of target detection, especially in bridging domain gaps. STDSD was evaluated against the existing detectors using the SPEED, SPEED+, SwissCube, Web-scraped, and Hardware-in-loop datasets. Specifically, STDSD achieved an F-score of 0.9989 on the SPEED dataset, 0.9876 on the “Lightbox” subset, and 0.9523 on the “Sunlamp” subset of the SPEED+ dataset. Additionally, on the SwissCube, Web-scraped, and Hardware-in-loop datasets, the F-scores were 0.9805, 0.8848, and 0.9879, respectively. Furthermore, STDSD exhibited a remarkable capability of over 10 frames per second (FPS) @GPU on a laptop. These results demonstrate that STDSD achieves high prediction accuracy and processing speed, underscoring the robustness capabilities across different datasets and the practicality for real-time applications in space environment. Note to Practitioners—This paper was motivated by the lack of robustness and accuracy in spacecraft object detection, which is also applicable to target detection in other fields. Existing approaches typically employ end-to-end networks that directly output target positions without considering the complex backgrounds and the requirements of real-time and reliability for space missions. In this paper, a spacecraft target detection framework considering similarity and diversity strategy is proposed. It utilizes the uncertainty and randomness outputted by neural networks, where different networks exhibit diversity in suspicious regions and similarity in target regions. By combining the results from multiple networks, the method leads to improved prediction accuracy. In this paper, we analyze the impact of parameters such as dataset distribution, thresholds, and the number of networks on the accuracy of object detection, and compare it with state-of-the-art methods. Experimental results demonstrate that the proposed method has better performance and the detection accuracy and proceeding speed might be balanced by adjusting different network configurations.}, archive = {J_TASE}, author = {Yanfang Liu and Rui Zhou and Zheng Yao and Jiayu She and Naiming Qi}, doi = {10.1109/TASE.2025.3531720}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11038--11049}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {STDSD: A spacecraft target detection framework considering similarity and diversity}, volume = {22}, year = {2025}, }
• (2025). Optimal stochastic injection attack strategy against discrete time-varying system via moment matrices. TASE, 22, 11026--11037. (www)

  @article{ , abstract = {This paper investigates injection attacks incorporating stochastic noise against linear discrete time-varying system, which is more general but also more challenging to defend than deterministic injection attacks. Based on the optimization theory and novel key defining matrices, an optimal stochastic injection attack strategy is proposed. Unlike existing strategies, this strategy is stochastic, making it harder for defenders to predict. Therefore, the newly designed attack is expected to be widely used to disrupt system performance. By leveraging estimated data from the observer and the attack input, a virtual residual system is established, which more accurately reflects changes in the system error before and after the attack than the traditional error system. Using the state and output residuals along with the stochastic attack input, two performances are defined to ensure the stealthiness and effectiveness of the attack, respectively. Subsequently, an optimal attack problem with non-convex objective function and constraint is formulated. The key to acquiring the designed optimal stochastic injection attack strategy is to apply a semi-definite relaxation involving moment matrices for transforming this non-convex optimization problem into a convex optimization problem and solving it. Finally, the effectiveness of the proposed attack strategy is validated through numerical simulation of a networked mass-spring-damper system and a V-formation experiment involving three quadrotors. Note to Practitioners—The primary objective of this paper is to focus on the cyber security of discrete time-varying systems from the perspective of the attacker, which provides insight into the way of generating attack strategies under the stochastic case. The majority of existing injection attack strategies against intelligent systems are deterministic, and this can make the attacks less effective as the attacked system reconstructs and compensates for the attack signals, and the attack strategies tend to fail or are mostly ineffective. This paper synthesizes state estimation, semi-definite programming, optimization principles, and control theory to propose an optimal stochastic injection attack strategy. Compared with the deterministic injection attack strategies, the latter is harder to defend. Specifically, the operation of the attacker is divided into two phases: initial data eavesdropping and strategy generation. In the initial data eavesdropping phase, the attacker continuously eavesdrops and stores the initial traffic data of the system for a virtual state residual data estimation. In the strategy generation phase, the attacker generates the attack strategy using semi-definite programming with the help of the measured initial data. The mathematical analytical form of the proposed optimal stochastic injection attack strategy is given in detail. Subsequently, the effectiveness is verified by the numerical simulation and experiment based on a leader-following form consisting of three quadcopters, but not yet tested in production. In the future, we plan to design stochastic injection attack strategies with a data-driven framework.}, archive = {J_TASE}, author = {Sheng Gao and Dennis Gramlich and Hao Zhang and Huaicheng Yan and Christian Ebenbauer}, doi = {10.1109/TASE.2025.3531365}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11026--11037}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal stochastic injection attack strategy against discrete time-varying system via moment matrices}, volume = {22}, year = {2025}, }
• (2025). Degradation modeling and prognostic analysis under unknown failure modes. TASE, 22, 11012--11025. (www)

  @article{ , abstract = {Operating units often experience various failure modes in complex systems, leading to distinct degradation paths. Relying on a prognostic model trained on a single failure mode may result in poor generalization performance across multiple failure modes. Therefore, accurately identifying the failure mode is of critical importance. Current prognostic approaches either ignore failure modes during degradation or assume known failure mode labels, which can be challenging to acquire in practice. Moreover, the high dimensionality and complex relations of sensor signals make it challenging to identify the failure modes accurately. To address these issues, we propose a novel failure mode diagnosis method that leverages a dimension reduction technique called UMAP (Uniform Manifold Approximation and Projection) to project and visualize each unit’s degradation trajectory into a lower dimension. Then, using these degradation trajectories, we develop a time series-based clustering method to identify the training units’ failure modes. Finally, we introduce a monotonically constrained prognostic model to predict the failure mode labels and Remaining Useful Life (RUL) of the test units simultaneously using the obtained failure modes of the training units. The proposed prognostic model provides failure mode-specific RUL predictions while preserving the monotonic property of the RUL predictions across consecutive time steps. We evaluate the proposed model using a case study with the aircraft gas turbine engine dataset.Note to Practitioners—The paper aims to develop an unsupervised method for identifying potential failure modes based on multiple sensor signals during the degradation process. After obtaining the failure mode labels, we introduce a monotonically constrained prognostic model for jointly predicting the failure modes and the RUL for the test units. Implementing this method involves four steps: First, collecting multiple sensor signals, and the failure times of historical units. Second, applying the UMAP dimension reduction technique to transform the high-dimensional multisensor data into low-dimensional representations and then visualizing the degradation trajectories for each unit. Third, utilizing a time series-based clustering method to get the failure mode labels of each training unit. Fourth, constructing a joint prognostic model to predict the failure modes and provide monotonically constrained RUL predictions for the test units. The proposed data-driven method can capture complex data relationships and has a flexible model structure, allowing it to adapt to different types of sensor data and operating conditions. In addition, the proposed model provides RUL predictions that are consistent with prior domain knowledge. As a result, the proposed model can be applied in various practical scenarios that involve degradation systems characterized by complex structures and unknown failure modes, providing a valuable tool for practitioners in real-time decision-making and maintenance planning.}, archive = {J_TASE}, author = {Ying Fu and Ye Kwon Huh and Kaibo Liu}, doi = {10.1109/TASE.2025.3530845}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11012--11025}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Degradation modeling and prognostic analysis under unknown failure modes}, volume = {22}, year = {2025}, }
• (2025). FTLS: Fragments-based twofold learning strategy for target recognition on the lunar surface. TASE, 22, 10997--11011. (www)

  @article{ , abstract = {In lunar exploration missions, tasks such as target recognition or obstacle detection always encounter significant challenges due to the quality of images and the limited availability of training samples. Typically, existing methodologies can only recognize and locate a limited number of targets within simple scenarios, and models developed within simulated environments lack practical applicability. To alleviate above issues, an image fragments-based twofold learning strategy has been conceptualized. This strategy facilitates the decoupling of target features from environmental attributes through processes such as image fragmentation, confidence evaluation, and domain transfer, thereby achieving enhanced recognition accuracy across diverse lighting conditions. To provide more accurate localization and identification of a wider variety of targets, $\mathrm {\omega }$ -CIoU loss is introduced to address the anomalies in prediction box scale and target count induced by flare and shadow features. Moreover, the establishment of the Real Chang’e Lunar Landscape Dataset, the most extensive public dataset for lunar surface target recognition in terms of sample quantity and diversity, provides an invaluable experimental foundation for future research in this field. Comprehensive experimental results on the RCLLD and public dataset ALLD demonstrate that the FTLS can significantly boosts recognition precision and generalization capabilities without escalating model complexity, outperforming prevailing target recognition methodologies for lunar exploration rovers.Note to Practitioners—This research is inspired by the need for high-precision, robust target recognition and avoidance systems for lunar rovers during lunar surface exploration missions. These systems are crucial for deployment on rovers with limited computational resources and energy, providing a convenient platform for future lunar terrain surveys, lunar base site selection, and mineral collection tasks. However, existing target recognition systems usually require training on extensive actual lunar datasets and often suffer from significantly reduced accuracy in areas with severe lighting changes, which decreases reliability on the lunar rover platform—an unacceptable risk for the cost-intensive lunar exploration missions. To address these challenges, this paper proposes a twofold learning strategy based on image fragments that trains on limited actual lunar data, offers resistance to flare and shadow interference, and its effectiveness is validated on both actual lunar and simulated datasets.}, archive = {J_TASE}, author = {Yanbo Wang and Ting Yuan and Chuankai Liu and Edmond Q. Wu and Jiuchao Qian}, doi = {10.1109/TASE.2025.3532019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10997--11011}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FTLS: Fragments-based twofold learning strategy for target recognition on the lunar surface}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy prescribed-time formation control for nonlinear multi-agent systems. TASE, 22, 10986--10996. (www)

  @article{ , abstract = {This paper discusses the issue of achieving the prescribed-time formation (PTF) over a directed topology for nonlinear multi-Agent systems (NMASs). A novel PTF protocol framework is proposed through the incorporation of a time-varying function for NMASs. Fuzzy Logic Systems (FLSs) are used to approximate potentially unknown nonlinear functions within the system. It is crucial to note that incorporating the adaptive control technique into the proposed protocol framework eliminates the adverse impact stemming from fuzzy approximation errors. Consequently, the formation errors of each agent converge to zero within the prescribed time. Additionally, through the introduction of a novel adaptive law, the protocol framework is further expanded to the NMASs with disturbances. Both the benefits and efficacy of the presented protocol are shown through numerical examples. Note to Practitioners—This paper delves into the pivotal issue within the realm of NMASs concerning prescribed-time formation control. This paper is motivated by the observation that current practices utilizing fuzzy controllers may not achieve the convergence of system formation errors to zero within the user-defined time frame. To solve this, the introduction of a prescribed-time control methodology is advocated, poised to expedite the formation convergence. Furthermore, the incorporation of an adaptive fuzzy controller is proposed to address challenges stemming from inaccurate system modeling, thereby ensuring the stringent control accuracy. The proposed framework harbors considerable potential for application across diverse industrial contexts, encompassing the realms of mobile robotics, unmanned aerial vehicles, and vehicular traffic management systems. Subsequent research endeavors can delve deeper into refining this approach and investigating methodologies for attaining prescribed time control within switched multi-agent systems.}, archive = {J_TASE}, author = {Haiying Zhang and Chen Chen and Zhengrong Xiang}, doi = {10.1109/TASE.2025.3531846}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10986--10996}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy prescribed-time formation control for nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). MINER-RRT*: A hierarchical and fast trajectory planning framework in 3D cluttered environments. TASE, 22, 10973--10985. (www)

  @article{ , abstract = {Trajectory planning for quadrotors in cluttered environments has been challenging in recent years. While many trajectory planning frameworks have been successful, there still exists potential for improvements, particularly in enhancing the speed of generating efficient trajectories. In this paper, we present a novel hierarchical trajectory planning framework to reduce computational time and memory usage called MINER-RRT*, which consists of two main components. First, we propose a sampling-based path planning method boosted by neural networks, where the predicted heuristic region accelerates the convergence of rapidly-exploring random trees. Second, we utilize the optimal conditions derived from the quadrotor’s differential flatness properties to construct polynomial trajectories that minimize control effort in multiple stages. Extensive simulation and real-world experimental results demonstrate that, compared to several state-of-the-art (SOTA) approaches, our method can generate high-quality trajectories with better performance in 3D cluttered environments (https://youtu.be/fXuuMRX19q0). Note to Practitioners—The motivation is the problem of planning trajectories for quadrotor autonomous flight in 3D cluttered and complex scenarios such as wild forest exploration and subterranean environment search-and-rescue. Sampling-based path planning methods are suitable for dealing with the complexity of the physical environment but are not convenient for computing dynamics and their differentials. Optimization-based trajectory generation methods are appropriate for handling various high-order constraints but rely on high-quality initial path solutions. Therefore, this paper combines the advantages of the two methods to propose a novel trajectory planning framework that can generate high-quality trajectories for quadrotors faster than many previous algorithms. We conduct numerous simulations and real-world experiments to verify that our method can be effectively deployed in real scenarios and empower quadrotors for complex autonomous tasks in the future.}, archive = {J_TASE}, author = {Pengyu Wang and Jiawei Tang and Hin Wang Lin and Fan Zhang and Chaoqun Wang and Jiankun Wang and Ling Shi and Max Q.-H. Meng}, doi = {10.1109/TASE.2025.3531504}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10973--10985}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MINER-RRT*: A hierarchical and fast trajectory planning framework in 3D cluttered environments}, volume = {22}, year = {2025}, }
• (2025). Spatio-temporal regularized stochastic configuration network for supervised and semi-supervised soft sensor development. TASE, 22, 10960--10972. (www)

  @article{ , abstract = {Stochastic configuration network (SCN) has become a favorable soft sensor model due to its innovative random parameter construction approach under inequality constraint. However, it may suffer from the disadvantage of model overfitting with the increase of hidden nodes. To handle this issue, an improved Regularized SCN based on Spatio-Temporal nearest neighbors (ST-RSCN) is presented for better nonlinear soft sensor development. Different from regularized SCN only with L2 regularization, a dually-regularized SCN optimization framework is designed, where the L2 regularization term and manifold regularization (MR) term are applied to enforce the constraints from the perspectives of model parameters and data structure. Peculiarly, considering the data dynamic property, the traditional spatial nearest neighbor selection method is upgraded by integrating temporal searching strategy. Two spatio-temporal neighbor searching strategies are formed by designing different neighbor determination orders in spatial and temporal domains. The efficiency of the developed ST-RSCNs is finally demonstrated by two industrial cases, including a debutanizer column process and a continuous stirred tank reactor. The outcomes in both supervised and semi-supervised scenarios indicate that ST-RSCNs have better prediction performance compared with several models with respect to stability and generalization. Note to Practitioners—Data structure characteristic is of great importance for soft sensor model performance in practice. However, this is omitted in the existing SCN models. This paper designs a new enhanced regularized SCN for soft sensing. Specifically, a unified optimization objective is established involving both L2 regularization for constraining output weight magnitude and manifold regularization for mining the underlying geometric structure information. Further, two spatio-temporal nearest neighbor strategies are introduced to better fit the dynamic data structure. ST-RSCN related principles and proofs are described in detail. The simulation results of DCP and CSTR cases show the proposed ST-RSCNs significantly reduce the model prediction root mean squared errors (RMSEs) and are applicable for supervised and semi-supervised domains. Additionally, the key parameters’ selection methods are discussed, which contributes to a sound understanding of this paper.}, archive = {J_TASE}, author = {Yue Zhao and Xiaogang Deng and Jing Zhang and Ping Wang}, doi = {10.1109/TASE.2025.3531850}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10960--10972}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Spatio-temporal regularized stochastic configuration network for supervised and semi-supervised soft sensor development}, volume = {22}, year = {2025}, }
• (2025). Log-scale quantization in distributed first-order methods: Gradient-based learning from distributed data. TASE, 22, 10948--10959. (www)

  @article{ , abstract = {Decentralized strategies are of interest for learning from large-scale data over networks. This paper studies learning over a network of geographically distributed nodes/agents subject to quantization. Each node possesses a private local cost function, collectively contributing to a global cost function, which the considered methodology aims to minimize. In contrast to many existing papers, the information exchange among nodes is log-quantized to address limited network-bandwidth in practical situations. We consider a first-order computationally efficient distributed optimization algorithm (with no extra inner consensus loop) that leverages node-level gradient correction based on local data and network-level gradient aggregation only over nearby nodes. This method only requires balanced networks with no need for stochastic weight design. It can handle log-scale quantized data exchange over possibly time-varying and switching network setups. We study convergence over both structured networks (for example, training over data-centers) and ad-hoc multi-agent networks (for example, training over dynamic robotic networks). Through experimental validation, we show that (i) structured networks generally result in a smaller optimality gap, and (ii) log-scale quantization leads to a smaller optimality gap compared to uniform quantization. Note to Practitioners—Motivated by recent developments in cloud computing, parallel processing, and the availability of low-cost CPUs and communication networks, this paper considers distributed and decentralized algorithms for machine learning and optimization. These algorithms are particularly relevant for decentralized data mining, where data sets are distributed across a network of computing nodes. A practical example of this is the classification of images over a networked data centre. In real-world scenarios, practical model nonlinearities such as data quantization must be addressed for information exchange among the computing nodes. This work emphasizes the importance of handling log-scale quantization and compares its performance over uniform quantization. By exploring these quantization methods, we aim to determine which is more accurate in terms of optimality gap and learning residual. Moreover, we study the impact of the structure of the information-sharing network on reducing the optimality gap and improving the convergence rate of distributed algorithms. As contemporary distributed and networked data mining systems demand highly accurate algorithms with fast convergence for real-time applications, our research emphasizes the benefit of structured networks under logarithmic quantization information-exchange. Our findings can be extended to different machine learning algorithms, offering pathways to more accurate and faster data mining solutions.}, archive = {J_TASE}, author = {Mohammadreza Doostmohammadian and Muhammad I. Qureshi and Mohammad Hossein Khalesi and Hamid R. Rabiee and Usman A. Khan}, doi = {10.1109/TASE.2025.3526967}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10948--10959}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Log-scale quantization in distributed first-order methods: Gradient-based learning from distributed data}, volume = {22}, year = {2025}, }
• (2025). Continuous nonsingular fast terminal sliding mode control for a free gyro stabilized mirror system. TASE, 22, 10937--10947. (www)

  @article{ , abstract = {Optical sensor-based systems necessitate a pointing and stabilization mechanism to separate the line-of-sight (LOS) from the operating environment. A gyro-stabilized mirror system (GSMS) can fulfill this requirement. Although this configuration is cost-effective, structured and unstructured nonlinear disturbances, dynamic imbalances, and carrier vibrations deteriorate the system performance. Therefore, it is indispensable to design a control system that has a fast response, high control accuracy, and insensitivity to perturbations. To acquire these attributes for GSMS applications, an improved nonsingular fast terminal sliding mode-based control is proposed in this paper. The prime advantage of the investigated control scheme is that its realization requires only available measurements of the system and control gains without acquiring a priori knowledge of the bounds on system perturbations. Moreover, the developed controller alleviates the main drawback, chattering, in conventional sliding mode control while preserving its robustness to disturbances. The stability of the closed-loop system is proven via the Lyapunov method, and the effectiveness of the proposed control strategy is elucidated through comparative numerical studies. Note to Practitioners—This paper proposes an improved control system for GSMS used in optical sensors. These systems are susceptible to vibrations and disturbances that can degrade their performance. The proposed control method offers several advantages for practical applications: (i) Fast Response and High Accuracy: The controller achieves fast settling times and minimizes pointing errors, leading to improved overall system performance; (ii) Insensitivity to Disturbances: The controller is designed to handle external vibrations and other uncertainties without requiring precise knowledge of their characteristics. This makes it robust and adaptable to real-world operating conditions; (iii) Reduced Chattering: Unlike traditional sliding mode control, this method minimizes chattering, a high-frequency control output oscillation that can damage mechanical components. Implementing this control strategy requires only angular position measurements. Furthermore, unlike previous methods that typically employ artificial neural networks or fuzzy logic to obtain an approximate model for unknown disturbances, we use simple adaptive laws for adjusting control gains. These eliminate the need for complex system modeling, making it practical for real-world engineering applications. While the paper explores the effectiveness through simulations, real-world implementation can further illustrate closed-loop features. The focus of this paper is on GSMS. While the control method has broader applicability, its adaptation to other systems might require adjustments. Overall, this paper suggests that the proposed control scheme offers a promising approach for engineers working with GSMS or similar positioning systems that require high precision, fast response, and robustness to disturbances. Future research will focus on designing an event-triggered quantitative prescribed performance control scheme based on the proposed method for a GSMS.}, archive = {J_TASE}, author = {Siyavash Jamshidi and Alireza Safa and Saeed Mirzajani}, doi = {10.1109/TASE.2025.3530679}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10937--10947}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Continuous nonsingular fast terminal sliding mode control for a free gyro stabilized mirror system}, volume = {22}, year = {2025}, }
• (2025). GraspMixer: Hybrid of contact surface sampling and grasp feature mixing for grasp synthesis. TASE, 22, 10920--10936. (www)

  @article{ , abstract = {The capability of robots to rapidly adapt to new tasks without extensive reprogramming offers significant flexibility in reconfiguration of manufacturing processes to cope with unforeseen events. In modern manufacturing environments where numerous hardware and software systems exchange data with each other to perform a myriad of tasks, modularizing sub-systems and reusing commonly available information like product CAD models can increase robustness and efficiency of the reconfiguration. Yet, current approaches for robotic grasping tend to focus on standalone vision-based learning that often require either retraining to adapt to new object categories or massive dataset not available in manufacturing environments, making generalization challenging. This paper addresses the problem of exploiting available information, like CAD models, in manufacturing settings to efficiently generate a tractable set of grasps for known rigid objects, which can be directly applied to a wide class of robotic manipulations. In order to quickly produce diverse grasp configurations for arbitrary geometric models, we present GraspMixer, a combination of (1) an efficient offline sampler that utilizes specifications of a parallel-jaw gripper, and (2) a mapping function that fuses multiple features of a grasp to output a binary quality metric. During evaluation using physics-based simulations, a robotic gripper successfully executes 92.9% of all grasp configurations for 12 novel objects selected by GraspMixer. Among five different grasp sampling methods, GraspMixer also achieves the highest grasp success rate when performing table-top single object grasping under object pose uncertainty. The computation of this offline pipeline takes less than 1.0 minutes for each object without GPU hardware acceleration, which is comparable to or outperforms most of the benchmarks in the evaluation. Importantly, our framework exhibits impressive simulation-to-reality adaptation, achieving over 95% grasp success rate on previously unseen novel objects. All of these results are achieved with fewer than 10% of the samples typically used by other learning-based grasping techniques. Note to Practitioners—Modularization is a major theme in current manufacturing systems to increase efficiency. In this work, we introduce a new framework called GraspMixer, which is part of a larger manipulation and decision making architecture to enable versatile robotic manipulation in a manufacturing environment. The framework decomposes the task of reasoning about graspable local surfaces on object 3D models into sequentially connected sub-components. GraspMixer leverages information about objects and grippers, including their 3D models, materials, and inertial properties, which are available in a manufacturing environment. This enables our framework to automatically precompute grasping points on new objects that can be shared among multiple robots equipped with parallel-jaw grippers. GraspMixer synergizes with Internet of Things (IoT) and Cloud Computing platforms to efficiently scale up advanced robotic automation in manufacturing. Such a combination could provide greater flexibility in deploying advanced perception systems in a manufacturing environment to accelerate adaptation of the automation while saving computational resources of onboard processors within robots.}, archive = {J_TASE}, author = {Hojun Lee and Tyler Toner and Dawn M. Tilbury and Kira Barton}, doi = {10.1109/TASE.2025.3530795}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10920--10936}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GraspMixer: Hybrid of contact surface sampling and grasp feature mixing for grasp synthesis}, volume = {22}, year = {2025}, }
• (2025). Deep reinforcement learning based transmission scheduling for sensing aware control. TASE, 22, 10905--10919. (www)

  @article{ , abstract = {Massive field data is wirelessly transmitted to the edge side to facilitate sensing and control in the emerging Industrial Internet of Things (IIoT) systems. Under the expanding transmission scheduling space and dynamic network conditions, balancing control performance and limited transmission resources is a fundamental challenge. For this problem, we propose a novel deep reinforcement learning (DRL)-based transmission scheduling method (DTSM), where sensing performance guarantee is introduced for its criticality in ensuring complete system observation and effective control. Specifically, taking system observability as the key metric, the time slots for multi-sensor data transmission under different control demands are properly reserved with theoretically guaranteed performance. Then, the primal-dual DRL framework is adopted to further improve the overall performance of system control and resource utilization by dynamically scheduling the transmission number of each sensor. The scheduling is based on the real-time states of sensing and wireless network, and the action space is determined according to our reserved time slots. Besides, after primal-dual updates, the scheduling results can satisfy the estimation error-evaluated constraint imposed for the ultimate control effect. Finally, the proposed method is applied to the industrial laminar cooling process and its effectiveness is fully demonstrated. Note to Practitioners—This paper is motivated by the requirement of balancing control performance and scarce transmission resources in industrial automation fields such as steel manufacturing, where massive sensor data is transmitted to the edge side through wireless networks. The expanding transmission scheduling space and dynamic network conditions have led to increased interest in advanced deep reinforcement learning (DRL) methods. However, few previous works have explored the impact of control demands on intelligent transmission scheduling design. For these issues, we propose a novel DRL-based transmission scheduling method (DTSM), where the time slots for multi-sensor data transmission are delicately reserved according to different control demands and dynamic scheduling is realized based on real-time states of sensing and wireless network. The overall performance of system control and resource utilization is improved, and practitioners can easily adjust method parameters to achieve the desired balance between the two aspects according to practical demands. Case studies in the industrial hot rolling process demonstrate the superiority of DTSM. Our future work will consider the joint scheduling of uplink-downlink transmissions and design the collaboration among multiple edge computing nodes (ECNs) to address the limitations of centralized learning methods. Besides, the proposed method can be extended to other industrial applications such as flight control system testing.}, archive = {J_TASE}, author = {Tiankai Jin and Cailian Chen and Yehan Ma and Xinping Guan}, doi = {10.1109/TASE.2025.3530409}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10905--10919}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep reinforcement learning based transmission scheduling for sensing aware control}, volume = {22}, year = {2025}, }
• (2025). Predictive inspection and maintenance optimization for partially observable semi-markov deteriorating systems. TASE, 22, 10893--10904. (www)

  @article{ , abstract = {Inspection and maintenance strategies are crucial for ensuring system serviceability. Due to stochastic deterioration and limited observability, sequential decision-making often relies on established approaches such as Partially Observable Markov Decision Processes (POMDPs). However, such approaches impose constraints on periodic decision epochs and exponentially distributed sojourn times, leading to redundant early-stage inspections and increased overall costs. To address these limitations, we propose a novel Predictive Semi-Markov Decision Process (PSMDP) that leverages Remaining Useful Life (RUL) predictions to reduce unnecessary inspections. Our approach incorporates an online scheme that utilizes inspection data for RUL prediction and proactively schedules subsequent maintenance decisions. When applied to infrastructure management, such as bridge maintenance, our PSMDP demonstrates an average cost reduction of 8.9% across various deterioration scenarios compared to other sequential decision-making models. Note to Practitioners—Effective infrastructure maintenance is essential for ensuring public safety and economic growth. However, limited budgets necessitate more efficient strategies. Current periodic inspection strategies often lead to wasteful expenditures due to unnecessary inspections during the early stages of deterioration. To address this challenge, we propose a predictive and proactive approach that leverages Remaining Useful Life (RUL) estimation to schedule inspections and optimize sequential maintenance decisions. By intelligently planning inspections, practitioners can reduce redundant inspections by 89.5% and save overall costs by 8.9% while retaining the required reliability. This approach facilitates a paradigm shift from condition-based maintenance to predictive maintenance.}, archive = {J_TASE}, author = {Chunhui Guo and Zhenglin Liang}, doi = {10.1109/TASE.2025.3530014}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10893--10904}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predictive inspection and maintenance optimization for partially observable semi-markov deteriorating systems}, volume = {22}, year = {2025}, }
• (2025). Learning graph dynamics with interaction effects propagation for deformable linear objects shape control. TASE, 22, 10881--10892. (www)

  @article{ , abstract = {Robotic manipulation of deformable linear objects (DLOs) has broad application prospects, e.g., manufacturing and medical surgery. To achieve such tasks, a critical challenge is the precise control of the DLOs’ shapes, which requires an accurate dynamics model for deformation prediction. However, due to the infinite dimensionality of the DLOs and the complexity of their deformation mechanism, dynamics models are hard to theoretically calculate. In this paper, for representing the DLO, we use multiple particles being uniformly distributed along the DLO. For learning the dynamics model, we adopt Graph Neural Network (GNN) to learn local interaction effects between neighboring particles, and use the attention mechanism to aggregate the effects of these interactions for the purpose of effect propagation along the DLO (called GA-Net). For manipulation, the Model Predictive Control (MPC) coupled with the learned dynamics model is used to calculate the optimal robot movements, which can also generalize to unseen DLOs. Simulation and real-world experiments demonstrate that GA-Net shows better accuracy than existing methods, and the proposed control framework is effective for different DLOs. Specifically, for model prediction (150 steps), the prediction performance of GA-Net is 14.14% better than the strong baseline (IN-BiLSTM). Videos are available at https://parkergu.github.io/work_dlo/. Note to Practitioners—This paper was motivated by the problem of shape control of DLOs (e.g., ropes, cables) but it also applies to other deformable objects. Robotic manipulation of DLOs has broad application prospects across various industries, including medical surgeries and manufacturing. Existing approaches to manipulate DLOs, such as reinforcement learning, suffer from sample inefficiency and challenges in generalization. To alleviate these issues, we propose a model-based framework. We adopt GNN and attention mechanism to learn DLOs’ dynamics. Then we use MPC coupled with the learned dynamics model for manipulation of DLOs. The framework is sample-efficient for manipulation, and can generalize to unseen DLOs. Previous works on GNN-based dynamics model do not consider instantaneous propagation of interaction effects, which leads to a false prediction. To alleviate this issue, we adopt GNN to learn interaction effects between neighboring particles, and use the attention mechanism to propagate local interaction effects along the DLO. Simulation and real-world experiments demonstrate that our dynamics model shows better accuracy than existing methods, and also demonstrate the effectiveness of the proposed control framework.}, archive = {J_TASE}, author = {Feida Gu and Hongrui Sang and Yanmin Zhou and Jiajun Ma and Rong Jiang and Zhipeng Wang and Bin He}, doi = {10.1109/TASE.2025.3530957}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10881--10892}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning graph dynamics with interaction effects propagation for deformable linear objects shape control}, volume = {22}, year = {2025}, }
• (2025). Practical variable exponent fixed-time nonsingular sliding mode control for nonlinear systems and its applications. TASE, 22, 10871--10880. (www)

  @article{ , abstract = {For a class of nonlinear systems with disturbances, a new fixed time nonsingular sliding mode control (SMC) method is proposed in this paper. A new fixed time stability lemma is proposed firstly. Compared with traditional fixed time stability, the exponent of the proposed lemma can switch to a positive constant greater than or less than 1 when the system states are far from or near the equilibrium point. On the basis of the proposed lemma, a practical fixed time stability lemma is proposed, which ensures that the system states converge to a domain within a fixed time. Based on the proposed lemma, this paper proposes a novel terminal SMC (TSMC), which can be adjusted adaptively in different stages to make the system converge fastly. In addition, this article uses a saturation function to solve the singular problem and a hyperbolic tangent function to reduce chattering. And the proposed method ensures that the system states are fixed time stable. Finally, the proposed control method is verified through simulation and experiments to have advantages such as fast convergence. Note to Practitioners—The motivation of this paper is to study the trajectory tracking problem of aircraft, but it is also applicable to the trajectory tracking problem of other nonlinear systems, such as robotic arms. The exponents of existing fixed time control methods are positive constants greater than 1 and less than 1, which cannot transition when the system states are far away from or close to the equilibrium point. This article proposes a new method that enables the exponent to switch to a positive constant greater than 1 or less than 1 when the system states are far away from or close to the equilibrium point. This paper combines sliding mode control with fixed time control to address the trajectory tracking problem of aircraft, taking into account the issues of chattering and singularity, and designing a suitable controller. Finally, the feasibility of the proposed method was verified through simulation and experiments.}, archive = {J_TASE}, author = {Xiaohua Liu and Chao Jia and Haozhe Wang}, doi = {10.1109/TASE.2025.3529832}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10871--10880}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical variable exponent fixed-time nonsingular sliding mode control for nonlinear systems and its applications}, volume = {22}, year = {2025}, }
• (2025). Parallel inspection route optimization with priorities for 5G base station networks. TASE, 22, 10860--10870. (www)

  @article{ , abstract = {5G base station networks generate numerous alarms daily. With the increasing demand for digital services, it is vital to inspect and rectify anomalies to uphold user satisfaction. This study explores the potential of unmanned aerial vehicle (UAV) empowered opportunistic inspection based on alarm data. We formulate the inspection routing problem as a prioritized traveling salesman problem (PTSP) encompassing two categories of base stations. Priority is assigned to stations generating more alarms, while others are subject to opportunistic inspection. To expedite large-scale opportunistic inspection routes, we introduce a novel transformer-based parallelizable routing algorithm (TPRA). TPRA is an intelligent optimization that orchestrates multiple parallelized constrained reinforcement learning algorithms. Through balancing spectral clustering, the large-scale graph is segmented into manageable subgraphs. For each subgraph, the prioritized inspection routing problem is formulated as a constrained Markov decision process and optimized by transformer-based reinforcement learning in parallel. The optimized subgraphs are then merged using an adaptive large neighborhood search approach. Through parallel computing, our approach achieves as much as 75% reduction in computation time, while concurrently generating shorter routes. The approach is implemented in real-world cases to validate its efficacy. Note to Practitioners—The rapid expansion of 5G infrastructure underscores the critical need for advanced technology and maintenance strategies. Base stations are often placed at high altitudes to ensure line-of-sight connectivity, which poses difficulties for maintenance, particularly in challenging terrains. UAVs offer a promising solution for faster and safer inspection and rectification. The designed approach utilizes reinforcement learning in parallel to optimize UAV inspection routes in an opportunistic manner. This method strategically prioritizes inspection routes based on the real-time base station alarm data, ensuring a swift and effective response to potential issues. Trained in simulated scenarios, the model requires few adjustments for real-world deployment, making it readily implementable in 5G networks. Beyond the potential of the 5G network, the approach also unlocks new value across various types of service in the low-altitude economy.}, archive = {J_TASE}, author = {Xiangqi Dai and Zhenglin Liang}, doi = {10.1109/TASE.2025.3530425}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10860--10870}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Parallel inspection route optimization with priorities for 5G base station networks}, volume = {22}, year = {2025}, }
• (2025). Server failure detection for robustness analysis in automated manufacturing systems with arbitrary resource failures. TASE, 22, 10845--10859. (www)

  @article{ , abstract = {In the paradigm of Petri nets (PNs), this paper focuses on the maximally permissive robustness analysis for automated manufacturing systems (AMSs) with arbitrary resource failures through server failure detection. First, unreliable system of sequential systems with shared resources (U-S4Rs) are constructed by adding detection subnets to the original PNs for each unreliable place. Based on different failed statuses of unreliable resources, the reachable marking set of a U-S4R is partitioned into one safe and multiple failed subsets, where there exists a one-to-one relationship between all markings in the safe subset and those of the original system. Second, subnet systems corresponding to part types are categorized to provide the definitions of the robust, strongly robust, and weakly robust markings. Third, we provide algorithms to compute the sets of strongly connected, weakly connected, and dangerous markings in each failed zone. When resources are in a failed status ${\mathcal {X}}_{f}$ , we establish an ${\mathcal {X}}_{f}$ -failure-dependent marking set. The robustness of all markings within that set can be evaluated by examining the connectedness of the markings derived from those in that set. Moreover, for all markings not in the ${\mathcal {X}}_{f}$ -failure-dependent marking set but in the safe subset, we provide conditions further to determine the robustness of these markings. Finally, examples are presented to illustrate the proposed methods and show the advantage over the existing ones.Note to Practitioners—In practical manufacturing scenarios, it is crucial to analyze and control automated manufacturing systems (AMSs) to ensure uninterrupted production in the case of resource failures. Failure to handle these failures effectively can result in unexpected blockages, reduced throughput, and compromised product quality. Most existing works have been dedicated to robust control policies in order to guarantee the robustness. However, there are still some states that do not satisfy the robust control policies despite their robustness. This paper specifically investigates the maximally permissive robustness analysis of AMSs with arbitrary resource failures using Petri nets. Some significant results are proposed to determine the robustness of all markings for the original system when the unreliable resources are in different failed statuses. Therefore, based on the results of the maximally permissive robustness analysis, researchers can design optimal robust controllers to enhance the flexibility of manufacturing processes considerably.}, archive = {J_TASE}, author = {Junyan Li and Hesuan Hu}, doi = {10.1109/TASE.2025.3528996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10845--10859}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Server failure detection for robustness analysis in automated manufacturing systems with arbitrary resource failures}, volume = {22}, year = {2025}, }
• (2025). Model-free tracking control for linear stochastic systems via integral reinforcement learning. TASE, 22, 10835--10844. (www)

  @article{ , abstract = {This paper investigates optimal tracking control of linear stochastic systems with multiplicative state-dependent and input-dependent noise via a novel model-free integral reinforcement learning algorithm. We have conquered two major obstacles in this work. Firstly, the model-free tracking control on linear stochastic systems has scarcely been pivoted academically and it demands innovative methods and mathematical strategies to progress. An augmented stochastic differential equation of Itô’s type has been constructed while the control objective has been equated to minimization of the expected quadratic cost function. A model-dependent algorithm is shown to solve the stochastic algebraic Riccati equation. Second, control of linear stochastic systems essentially values for its sophisticated formation and sustained application. Inspired by optimal stationary control for linear stochastic systems, an integral reinforcement learning algorithm utilizing the adaptive dynamic programming method has been developed to scrap the reliance on the complete knowledge of system dynamics. The convergence of the core matrix and the sequence of control policies and the system stability has been rigorously studied following. Finally, numerical simulations are performed to demonstrate the efficacy of the proposed integral reinforcement learning methodology. Note to Practitioners—This paper focuses on the problem of optimal tracking control on linear stochastic systems, which is motivated by optimal stationary control and robust adaptive dynamic programming. Stochastic multiplicative noises are pervasive in modern control systems and engineering fields, such as power systems, aerospace systems, and industrial processes, thus important in modeling the random perturbation in system parameters and coefficients. However, research on tracking control on linear stochastic systems are challenging and languishing. First, we need to eradicate the dependence on complete knowledge system dynamics since machinery parameters in industrial design are constantly varying through time and space which builds a lasting impact on its structure, performance, and growth trajectory. Second, we innovatively apply integral reinforcement learning techniques on linear stochastic systems, which can be seamlessly integrated into networked control systems with noisy communication channels or neuronal networks to constructively solve stochastic multiplayer differential games online. Numerical simulations demonstrate that the proposed algorithm is feasible, which has also facilitated the compounding of robust adaptive dynamic programming and optimal tracking control on the industrial society. This study exhaustively presents the proposed approach to iteratively find optimal policies of the optimal tracking control problem directly from input and state data without explicitly identifying any system dynamics, and thoroughly unveils the explicit mathematical proof of convergence and system stability.}, archive = {J_TASE}, author = {Kun Zhang and Yunjian Peng}, doi = {10.1109/TASE.2025.3529895}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10835--10844}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model-free tracking control for linear stochastic systems via integral reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Fixed time control for interlayer synchronism under DDMNNs and application in secure communication. TASE, 22, 10827--10834. (www)

  @article{ , abstract = {This paper mainly focuses on fixed time control for interlayer synchronism problem under voltage-flux–time (VFT) delay deplux memristive neural networks (DDMNNs). Distinct from most previously reported achievements, continuous memristive VFT neural networks (MMNs) field with $2^{2{n^{2}} + n}$ variables are considered, where the memristor is considered as a continuous time-varying uncertain parameter, and n is the number of states in DDMNNs. This memristor is a type of continuous system built on the HP memristor. To synchronize VFT DDMNNs, a fixed time control policy is proposed for interlayer synchronism of flux and voltage values under the VFT DDMNNs. According to the second method of Lyapunov, selecting appropriate Lyapunov functionals and using inequality techniques, the structure and parameters of a kind of fixed time controller are designed, and interlayer synchronism criteria for DDMNNs are obtained. Finally, a secure communication scheme based on interlayer synchronism is designed. It is found that the fixed settling time is relative to the constructive of DDMNNs and the parameters of the fixed time controller. Note to Practitioners—This work solves the fixed time interlayer synchronism control problem of VFT DDMNNs, which can be applied to some practical scenarios, such as information security communication and image encryption. When transmitting signals in computer networks, for security, encryption mechanisms are usually added to network communication protocols to achieve secure communication. In some encryption applications, there are special requirements for speed of encryption and decryption, which requires reducing encryption and decryption time and improving efficiency of encryption. This work provides a fixed time control strategy which can achieve synchrony for VFT DDMNNs under controllable time, thereby achieving encryption under controllable time. Meanwhile, improve encryption efficiency in secure communication.}, archive = {J_TASE}, author = {Fei Tan and Lili Zhou and Guangdeng Zong and Zhen Wang and Guangming Zhuang and Xingchen Shangguan}, doi = {10.1109/TASE.2025.3528957}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10827--10834}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed time control for interlayer synchronism under DDMNNs and application in secure communication}, volume = {22}, year = {2025}, }
• (2025). A weighted deep learning-based predictive control for multimode nonlinear system with industrial applications. TASE, 22, 10814--10826. (www)

  @article{ , abstract = {In response to the challenge of strongly nonlinear and multimode systems control, this paper introduces a weighted deep learning based adaptive predictive control method. This approach integrates LSTM networks for different operating modes using a set of weighting coefficients. These coefficients are dynamically updated during online control via an error-guided scheduling strategy to adapt to changing operation modes. Compared to offline identification based methods, the proposed method eliminates the need for mode recognition or model switching strategies and can adapt to drifted operation modes. In contrast to online methods, it achieves rapid model convergence and reduced computational cost, requiring only minimal data to update the weighting coefficients without necessitating the retraining of the LSTM networks. Theoretical convergence and stability analysis ensure the reliability of the proposed method. Numerical simulations and industrial control experiments demonstrate that the proposed approach exhibits favorable control performance across both known and drifted operation modes. Note to Practitioners—Considering the changing operation modes in complex industrial processes and the detrimental effect of slow or unstable control during system operation, this paper proposes a weighted LSTM based predictive control method for strongly nonlinear and multimode systems. Extensive experiments demonstrate that compared to other state-of-the-art methods, this method can rapidly adapt to changes in operating modes with a small amount of data, meeting both real-time and stability requirements for online control.}, archive = {J_TASE}, author = {Keke Huang and Wenpu Cao and Yishun Liu and Dehao Wu and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2025.3529124}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10814--10826}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A weighted deep learning-based predictive control for multimode nonlinear system with industrial applications}, volume = {22}, year = {2025}, }
• (2025). Bidirectional agent-map interaction feature learning leveraged by map-related tasks for trajectory prediction in autonomous driving. TASE, 22, 10801--10813. (www)

  @article{ , abstract = {Accurate prediction of the future trajectories of surrounding agents is essential for safely autonomous vehicles. However, it is quite challenging due to the dynamics of driving environments caused by frequent agent-agent and agent-map interactions. Most existing methods mainly focus on modeling the dynamics of agents but overlook the inherent dynamics of the usable map at different driving moments. To address this limitation, this paper proposes a bidirectional interaction network (DyMap) that takes into account the bilateral constraints of the map on the agents and the impact of the agents on the map, using a unique bidirectional attention module. By employing multiple stages of bidirectional interaction modules, the features of agents and map nodes are continuously updated to adapt to changing environments. To facilitate network learning, we further incorporate map-related auxiliary tasks into the learning objectives, including traffic flow prediction, map node occupancy prediction, and road direction prediction. These tasks are designed to help the network learn map features that can reflect the dynamics of the driving environment, leading to a comprehensive feature representation of the agents. This representation encodes not only the agent’s historical motions but also the latent map information triggered by other agents. This proposed network offers a new approach to modeling environmental dynamics in driving scenarios, and achieves the state-of-the-art performance on the Argoverse 1 and Argoverse 2 trajectory prediction benchmarks. Note to Practitioners—Autonomous vehicles are of particular interest to practitioners in the field of automation science and technology. Trajectory prediction is of critical importance to the safety of autonomous driving, which aims to forecast the future positions of a given traffic participator (called the agent in this paper, such as a vehicle, pedestrian, etc.) based on the observations of itself and other agents’ historical trajectories as well as the map information. This paper proposes a trajectory prediction network based on a novel bidirectional interaction module that enables bidirectional information flow between the features of map nodes and agents so as to simultaneously learn their feature representations being aware of the timing-vary property of traffic scenes. The proposed DyMap contains three blocks of bidirectional interaction modules, which are interleaved with self-interactions of the agents and map nodes. Three different kinds of map-related tasks are developed to facilitate network training by leveraging on maximizing the map features for predicting traffic flow, occupancy, and lane direction. The superiority of the proposed method has been confirmed by the widely used benchmarks in the community. Besides modelling the complex interaction between agents and map nodes in dynamic traffic scenarios, the proposed bidirectional interaction network can be beneficial to address other problems requiring symmetric modelling of interactions between different components, such as human-robot interaction. The proposed map-related tasks can be directly used in other applications requiring to extract map features while considering the time-varying conditions, including but not limited to robot navigation and embody intelligence.}, archive = {J_TASE}, author = {Bin Fan and Haifeng Yuan and Yinke Dong and Zhengyu Zhu and Hongmin Liu}, doi = {10.1109/TASE.2025.3529736}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10801--10813}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bidirectional agent-map interaction feature learning leveraged by map-related tasks for trajectory prediction in autonomous driving}, volume = {22}, year = {2025}, }
• (2025). Real-time caustic ratio prediction in alumina digestion process for closed-loop operation: A cloud-edge deep learning approach. TASE, 22, 10787--10800. (www)

  @article{ , abstract = {The current control of alumina digestion process (ALDP) relies on infrequent manual assays for caustic ratio, leading to open-loop operation between assays and suboptimal performance. This paper proposes a novel cloud-edge collaboration (CEC) architecture utilizing deep learning for real-time caustic ratio prediction, enabling closed-loop operation at all times. Our method combines slow-sampled assays with fast-sampled process measurements. An autoregressive moving average with extra inputs (ARMAX) model is used for feature extraction, while an improved bidirectional gated recurrent unit (I-BiGRU) that incorporates a multi-head attention (MHA) mechanism bridges data gaps through prediction. A soft-collaboration mechanism ensures smooth model updates at the edge, enabling operators to make informed control decisions. Theoretical analysis guarantees the convergence of prediction errors. Real-world data from a large-scale alumina plant demonstrates the superior performance of our prediction method compared to the baseline methods. Additionally, industrial experiments validate the effectiveness of the CEC architecture in supporting human operators for ALDP closed-loop operation. Note to Practitioners—This paper was motivated by the challenges of achieving closed-loop operation in the alumina digestion process. The concept of cloud-edge collaboration (CEC) architecture with deep learning offers a powerful solution for various industrial scenarios that rely on infrequent manual measurements, where real-time monitoring of key performance indexes is essential. CEC can continuously predict these values, bridging the gaps between slow, accurate assays and frequent, less precise sensor data. This approach allows for closed-loop operation based on real-time predictions, potentially improving production efficiency and yield. The ability to handle data gaps makes our method applicable to processes with inherent measurement challenges. By implementing CEC in various industries, practitioners can transition from open-loop control to data-driven closed-loop systems, optimizing processes and achieving substantial benefits.}, archive = {J_TASE}, author = {Liyi Yu and Wen Yu and Yao Jia and Tianyou Chai}, doi = {10.1109/TASE.2025.3529719}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10787--10800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time caustic ratio prediction in alumina digestion process for closed-loop operation: A cloud-edge deep learning approach}, volume = {22}, year = {2025}, }
• (2025). Voltage-sensitivity-based attack strategy against remote state estimation in power systems: Attack design and mitigation. TASE, 22, 10775--10786. (www)

  @article{ , abstract = {This paper investigates the attack-defense framework for power systems, a novel voltage-sensitivity-based (VSB) attack schedule method is proposed to maximize the attack destructiveness of the remote estimator of the target bus in the power system, and a mitigation mechanism is also constructed to make up for the negative effect of the proposed attacks. Firstly, the relationship between voltage and input current at different buses within the power system is analyzed, and the voltage sensitivity of target buses to other buses is calculated. Based on this, a VSB attack strategy is proposed to maximize the attack disruption on the target bus by targeting those with higher voltage sensitivity. Secondly, considering the clear periodicity of voltage signals, an enhanced local outlier factor algorithm is proposed to mitigate the proposed attack strategy. This algorithm can identify abnormal situations by detecting outliers and compensate for the attack by establishing a voltage model. Finally, the effectiveness of both the attack design method and mitigation strategy is validated through extensive simulation experiments conducted on the IEEE 14-bus and 118-bus systems. Note to Practitioners—This paper is motivated by the escalating challenges of attack-defense problems in power systems. From the perspective of attackers, existing DoS attack strategies often lack discrimination. They operate without considering the specific characteristics of the target, proceeding with attacks based on predetermined sequences while disregarding the varying importance of information carried by different targets. In response, this paper proposes a novel VSB attack strategy that utilizes readily available information to evaluate the voltage sensitivity of each bus within the power system. By targeting buses with high sensitivity, this strategy aims to magnify its destructive impact. From the perspective of defenders, this paper proposes utilizing the periodicity of voltage signals to mitigate the aforementioned attack strategy. Through simulation, the effectiveness of the proposed attack-defense framework is demonstrated from both attacker and defender perspectives.}, archive = {J_TASE}, author = {Guowei Liu and Engang Tian and Huwei Chen}, doi = {10.1109/TASE.2025.3529881}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10775--10786}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Voltage-sensitivity-based attack strategy against remote state estimation in power systems: Attack design and mitigation}, volume = {22}, year = {2025}, }
• (2025). Automated sperm tracking and immobilization with a clinically-compatible XYZ stage. TASE, 22, 10764--10774. (www)

  @article{ , abstract = {Automated positioning systems play a pivotal role in micro-scale cell manipulation. In clinical intracytoplasmic sperm injection (ICSI) for infertility treatment, a motile sperm needs to be immobilized by glass micropipette tapping for subsequent surgical steps. The process requires accurate tracking of the target sperm and precise alignment between the sperm tail and the micropipette. Manual sperm immobilization suffers from inconsistent success rates, and current robotic systems developed for the task fail to comply with the standard clinical setup. Instead of using a motorized micromanipulator as in existing robotic systems, this paper presents an automated and compact three-dimensional (3-D) positioning stage for sperm immobilization that can be seamlessly integrated into standard clinical platforms. To tackle the challenge of accurately tracking the target sperm with degraded detection quality due to the complex 3-D motion of the positioning stage, a multi-stage sperm tracking scheme is designed for detection-to-tracklet association. To prevent physical contact between the sperm head and the micropipette, an adaptive tail-tapping planning strategy based on the sperm head orientation analysis is established. A visual servo controller equipped with a dynamic sperm motion observer is further employed to achieve precise positioning of the target sperm during the immobilization process. Experimental results demonstrated that the proposed system achieved a sperm tracking accuracy of 88.12%, and a sperm positioning accuracy of $2.3~\pm ~1.2~\mu $ m. Further experiments revealed the system achieved a success rate of 93.5% and a time cost of 5.5 s for automated sperm immobilization. Note to Practitioners—This work presents an automated positioning system for the robotic immobilization of live human sperm in the intracytoplasmic sperm injection (ICSI) process. Conventional robotic systems developed for sperm immobilization use motorized stages which only provide two degrees of freedom (DOF) and disturb standard clinical setups with bulky sizes and additional equipment. Leveraging the advantages of piezoelectric positioners, a compact 3-D positioning system is developed to perform robotic sperm immobilization in an all-in-one manner. A sperm immobilization tracker is developed to track the target sperm during the immobilization process. The proposed multi-stage data association metric can effectively resist the noisy detection results due to occlusion and out-of-focus blur introduced by the 3-D movement of the positioning stage. Based on the analysis of the sperm head orientation, an adaptive tail-tapping planning strategy is established to avoid the risk of contacting the sperm head where DNA is contained. The developed positioning system can be easily integrated into a standard microscopy operation platform for biological cell manipulation. The proposed tracking scheme is applicable in various microscopy cell analysis scenarios where the detection results are inevitably affected by the degraded image quality.}, archive = {J_TASE}, author = {Haocong Song and Wenyuan Chen and Guanqiao Shan and Changsheng Dai and Steven Yang and Aojun Jiang and Hang Liu and Zhuoran Zhang and Yu Sun}, doi = {10.1109/TASE.2025.3529283}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10764--10774}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated sperm tracking and immobilization with a clinically-compatible XYZ stage}, volume = {22}, year = {2025}, }
• (2025). Total least squares algorithm for errors-in-variables systems: Iterative algorithm or two-step algorithm. TASE, 22, 10753--10763. (www)

  @article{ , abstract = {The total least squares (TLS) algorithm is a superior identification tool for low-order errors-in-variables (EIV) systems, where the estimate can be obtained by solving an eigenvector of the minimum eigenvalue of an augmented matrix. However, the TLS algorithm demonstrates inefficiency when applied to high-order EIV systems. This study introduces two innovative TLS algorithms: an iterative TLS algorithm, offering superior performance for low-order EIV models, and a two-step TLS algorithm, designed to effectively handle high-order EIV models. In comparison to the conventional TLS algorithm, these proposed methodologies present noteworthy advantages, including: 1) reduced computational costs, 2) the utilization of an iterative technique to calculate the inverse, and 3) the diversification of EIV identification methods. Simulation bench test examples are selected to show the efficacy of the proposed algorithms and transparent procedure for applications. Note to Practitioners—This paper was motivated by the problem of identifying network systems which are contaminated by noises. For network systems, the input and output data are usually contaminated by noises. Existing approaches to estimating such systems have the assumption that the noises are in little level scenarios or only the output data are contaminated by noises. This paper suggests two new total least squares approaches which can deal with systems contaminated by noises in medium level scenarios or whose input and output are both contaminated by noises. These two algorithms, using iterative technique and two-step technique, can: 1) avoid the matrix inverse calculation; 2) reduce the computational efforts; 3) increase the convergence rates. The proposed algorithms can also be extended to various fields such as inverse scattering, pattern recognition, image restoration, and computer vision.}, archive = {J_TASE}, author = {Jing Chen and Jing Na and Junhong Li and Quanmin Zhu}, doi = {10.1109/TASE.2025.3528532}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10753--10763}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Total least squares algorithm for errors-in-variables systems: Iterative algorithm or two-step algorithm}, volume = {22}, year = {2025}, }
• (2025). Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase continuous contact tasks. TASE, 22, 10741--10752. (www)

  @article{ , abstract = {This research unveils a novel Multi-Critic Reinforcement Learning framework designed to navigate the multifaceted challenges associated with multi-phased garment handling tasks, notably marked by persistent contact and erratic deformations between textiles and solid bodies. These tasks, ubiquitous in domestic and industrial environments, encompass activities such as dressing, fabric printing, and pressing, and are complicated by the unpredictability of textile states and the intricacy of devising control strategies. Our reinforcement learning model combines multiple time-sequenced Critic networks with traditional Deep Deterministic Policy Gradient (DDPG) techniques, thereby equipping the system to adapt to the diverse effects of fabric distortions throughout various stages. The effectiveness of this approach is demonstrated through a multi-phase pre-printing operation and further validated by real-world implementations, showing significant improvements in coverage and a substantial reduction in wrinkle formation, with its versatility further confirmed by a complex vertical dressing task. We anticipate future applications of this framework in a range of complex problems, not just garment handling. The model used in this paper can be found at https://github.com/jkk5454/multiddpg.git. Note to Practitioners—This paper addresses garment handling challenges where deformable clothes are in continuous contact with objects, such as pulling a T-shirt over a print bench before silk printing and dressing in a vertical hanger. Unlike previous research focusing on handling clothes in the air (like folding or unfolding), we tackle the complexities introduced by continuous contact, which can alter a garment’s shape and affect the task. We segment these tasks into distinct phases and employ Multi-Critic Reinforcement Learning to evaluate each phase, enabling us to predict their overall impact. Specifically, we divide a task like pulling a T-shirt over a workbench of the pre-printing tasks in garment printing into three phases and use Multi-Critic DDPG to generate control trajectories for a flat, correctly positioned surface. The practical applicability of our algorithm was further validated through experiments involving the dragging task on a realistic printing board and the vertical dressing task using an ironing board. This approach aims to facilitate garment handling tasks like dragging, dressing, and ironing, involving multiple phases and continuous contact with obstacles. However, the current simulation environment significantly differs from the real world, challenging policy transfer. Future work will concentrate on narrowing this simulation-to-reality gap.}, archive = {J_TASE}, author = {Yukuan Zhang and Dayuan Chen and Weizan He and Alberto Elías Petrilli Barceló and Jose Victorio Salazar Luces and Yasuhisa Hirata}, doi = {10.1109/TASE.2025.3527003}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10741--10752}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase continuous contact tasks}, volume = {22}, year = {2025}, }
• (2025). Compensatory data-driven networked iterative learning control with communication constraints and DoS attacks. TASE, 22, 10728--10740. (www)

  @article{ , abstract = {Considering the three critical factors of data quantization, channel fading, and denial of service (DoS) attack introduced by the networked control systems (NCSs) simultaneously, we propose a novel compensatory data-driven networked iterative learning control (COMP-DDNILC) method for nonlinear repetitive NCSs under a model-free design and analysis framework. By reformulating the iterative input-and-output (I/O) dynamics of the nonlinear NCS as an iterative linear data model (iLDM), an iterative linear predictive data model (iLPDM) is developed to predict the missing data arisen from DoS attacks. Then, a relationship is built to describe the coupling effects of the three critical factors, based on which the COMP-DDNILC is designed by involving the compensatory mechanism of DoS attacks and the fading coefficient inversion to improve the control performance. The COMP-DDNILC also involves an iterative adaption mechanism to update the iLPDM to enhance the robustness against uncertainties. The data-driven nature of COMP-DDNILC makes it applicable to practical NCSs without model information available. The simulation study verifies the results. Note to Practitioners—Networked control systems (NCSs) have becoming an important topic in practical processes owing to the shareable communication among the sensors, actuators, and controllers. However, the communication constraint and the cyber attack are inevitably induced to NCSs where the former is mainly involved with finite channel capacity and channel fading. Among different cyber attacks, denial of service (DoS) attack is most common and reachable. Further, it is worth noting that data quantization is an effective method to deal with channel capacity, but may degrade control performance by compressing I/O data. Therefore, we are motivated to propose a compensatory data-driven networked iterative learning control method by considering the three factors of data quantization, channel fading and DoS attack simultaneously to compensate the negative effect of the communication constraints and cyber attacks. An iterative linear data model is developed to formulate the iteration-based I/O dynamics of nonlinear NCS, and is also used to predict the missing data. The proposed method is data-driven and includes an iterative adaptation mechanism to enhance the robustness, therefore becoming much applicable to realistic nonlinear NCSs.}, archive = {J_TASE}, author = {Huimin Zhang and Ronghu Chi and Biao Huang and Zhongsheng Hou}, doi = {10.1109/TASE.2025.3528462}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10728--10740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Compensatory data-driven networked iterative learning control with communication constraints and DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Two-stage integrated planning of energy-saving operations of metro trains using MOJS and GWO algorithms. TASE, 22, 10713--10727. (www)

  @article{ , abstract = {In recent years, with a remarkable increase in urban rail transit operations, the issue of energy efficiency in train operations has attained increasing attention. In this study, a two-stage optimization model is proposed to optimize driving strategies and schedules. We comprehensively consider the optimization of train running curves, running time allocations to a whole line, and utilization of regenerative braking energy, to reduce the net energy consumption of train operations. In the first stage, a multi-objective jellyfish search (MOJS) optimization algorithm is used to optimize a switching sequence at each inter-station, and Pareto fronts are obtained corresponding to energy-saving train running curves. In the second stage, a grey wolf optimizer (GWO) is adopted to optimize running times between adjacent stations, dwell times at stations, and headway time. This stage aims to coordinate the operations of multiple trains, to reduce the traction energy consumption of a whole line, and to increase the utilization of regenerative braking energy. The optimality is discussed for the proposed two-stage optimization processes. Numerical experiments are conducted based on train and infrastructure data of the Beijing Yizhuang metro line. The results show that the proposed optimization model and solution algorithms have a considerable energy-saving effect. Note to Practitioners—The motivation of this work is to reduce the energy consumption of a metro line by optimizing control profiles and scheduling schemes of multiple trains, including three main steps. Firstly, a multi-objective optimization model is constructed with inter-station running time and traction energy consumption as optimization objectives, and the time-energy Pareto fronts between stations are obtained by optimizing the inter-station running curves of trains. Secondly, an objective function considering multi-train regenerative-energy synergistic utilization is established, with derived regenerative-energy utilization formulas employed to calculate saved energy. Finally, based on the obtained Pareto fronts between stations, running times between stations, dwell times at stations, and headway time are optimized to comprehensively reduce whole-line traction energy consumption and improve regenerative energy utilization. After these holistically optimized processes, the preferable energy-saving schemes can be attained for metro train operations.}, archive = {J_TASE}, author = {Xiangmeng Jiao and Yonghua Zhou and Hamido Fujita}, doi = {10.1109/TASE.2025.3527973}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10713--10727}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Two-stage integrated planning of energy-saving operations of metro trains using MOJS and GWO algorithms}, volume = {22}, year = {2025}, }
• (2025). Morphology transformation of underwater self-reconfigurable modular robots via heterogeneous decomposition and distributed control. TASE, 22, 10698--10712. (www)

  @article{ , abstract = {This paper addresses the morphology transformation problem of an underwater self-reconfigurable modular robotic system. Morphology decomposition and reconnections are reduced to mitigate transformation failures and the overhead of underwater wireless communication, giving rise to subgraph matching problems. We propose an efficient probabilistic decomposition method by constraining the search depth of maximal common subgraphs of the initial and goal morphologies. The computational complexity reduces from $O(n^{2})$ to $O(n)$ . The decomposition yields a swarm of heterogeneous clusters, which are interconnected modular robots of varying quantities. The heterogeneity makes the exchange of clusters’ designated positions in the goal morphology not immediately feasible. Subsequently, we present Distributed Control with minimal In-situ task Refinement (DCIR). DCIR is proven to ensure collision-free and deadlock-free morphology transformation. The numerical simulations involving up to 641 modular robots and experiments on 6 robots have shown that DCIR scales well with the number of modular robots, runs in real time, and reduces traveling distances by at least 14% and communication costs by about half, compared to the distributed control with homogeneous task exchange and the modified surface sliding method. Note to Practitioners—This paper presents a distributed control approach to transform the morphologies. Considering the limited communication bandwidth, the disconnections between modular robots are minimized. The proposed distributed control approach refines tasks locally to transform the morphologies, and it scales well to the number of modular robots. This effort is orthogonal to the existing studies on the structures of the modular system. However, the positioning of the underwater robots in this study was assumed known or given by an underwater motion capture system, and it should be further investigated.}, archive = {J_TASE}, author = {Wenjie Lu and Manman Hu}, doi = {10.1109/TASE.2025.3528757}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10698--10712}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Morphology transformation of underwater self-reconfigurable modular robots via heterogeneous decomposition and distributed control}, volume = {22}, year = {2025}, }
• (2025). Non-parametric neuro-adaptive formation control. TASE, 22, 10684--10697. (www)

  @article{ , abstract = {We develop a learning-based algorithm for the distributed formation control of networked multi-agent systems governed by unknown, nonlinear dynamics. Most existing algorithms either assume certain parametric forms for the unknown dynamic terms or resort to unnecessarily large control inputs in order to provide theoretical guarantees. The proposed algorithm avoids these drawbacks by integrating neural network-based learning with adaptive control in a two-step procedure. In the first step of the algorithm, each agent learns a controller, represented as a neural network, using training data that correspond to a collection of formation tasks and agent parameters. These parameters and tasks are derived by varying the nominal agent parameters and a user-defined formation task to be achieved, respectively. In the second step of the algorithm, each agent incorporates the trained neural network into an online and adaptive control policy in such a way that the behavior of the multi-agent closed-loop system satisfies the user-defined formation task. Both the learning phase and the adaptive control policy are distributed, in the sense that each agent computes its own actions using only local information from its neighboring agents. The proposed algorithm does not use any a priori information on the agents’ unknown dynamic terms or any approximation schemes. We provide formal theoretical guarantees on the achievement of the formation task. Note to Practitioners—This paper is motivated by control of multi-agent systems, such as teams of robots, smart grids, or wireless sensor networks, with uncertain dynamic models. Existing works develop controllers that rely on unrealistic or impractical assumptions on these models. We propose an algorithm that integrates offline learning with neural networks and real-time feedback control to accomplish a multi-agent task. The task consists of the formation of a pre-defined geometric pattern by the multi-agent team. The learning module of the proposed algorithm aims to learn stabilizing controllers that accomplish the task from data that are obtained from offline runs of the system. However, the learned controller might result in poor performance owing to potential data inaccuracies and the fact that learning algorithms can only approximate the stabilizing controllers. Therefore, we complement the learned controller with a real-time feedback-control module that adapts on the fly to such discrepancies. In practise, the data can be collected from pre-recorded trajectories of the multi-agent system, but these trajectories do need to accomplish the task at hand. The real-time feedback-control is a closed-form function of the states of each agent and its neighbours and the trained neural networks and can be straightforwardly implemented. The experimental results show that the proposed algorithm achieves greater performance than algorithms that use only the trained neural networks or only the real-time feedback-control policy. Our future research will address the sensitivity of the algorithm to the quality and quantity of the employed data as well as to the learning performance of the neural networks.}, archive = {J_TASE}, author = {Christos K. Verginis and Zhe Xu and Ufuk Topcu}, doi = {10.1109/TASE.2025.3528501}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10684--10697}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Non-parametric neuro-adaptive formation control}, volume = {22}, year = {2025}, }
• (2025). Behaviorally-aware multi-agent RL with dynamic optimization for autonomous driving. TASE, 22, 10672--10683. (www)

  @article{ , abstract = {This study presents a novel Multi-Agent Reinforcement Learning (MURL) architecture for autonomous vehicle (AV) navigation in complex urban traffic environments. By integrating a Social Value Orientation (SVO) model into a model-free SARSA reinforcement learning framework, our approach effectively balances individual agents’ social preferences with safety and performance objectives. A logistic regression-based risk assessment module evaluates collision probabilities in real time by analyzing spatiotemporal dynamics such as distances and velocities. Additionally, a dynamic optimizer adapts the learning rate and exploration strategies of the SARSA algorithm to provide efficient convergence to optimal policies. Extensive simulation experiments demonstrate that the proposed method significantly enhances safety and efficiency, achieving a 55.6% reduction in collision risk and increasing average rewards per episode by 2.1 compared to traditional SARSA without SVO. Furthermore, the optimized policy reduces average episode length, indicating the framework’s effectiveness in providing robust decision-making and adaptability across various traffic scenarios. Note to Practitioners—The proposed framework in this paper is driven by the demand for comprehensive navigation systems in the rapidly evolving field of connected and autonomous vehicles (CAVs), especially within complex and unpredictable urban environments and mixed traffic scenarios. As AVs are getting more and more attention, the capacity to navigate effectively among many road users, including other AVs, pedestrians, and human-driven vehicles, is essential. Our framework builds upon the SARSA algorithm to produce an optimal policy for the AV and integrates a dynamic optimization method that represents the concept of risk as the inverse logistic of potential collisions. Distinctive to our proposed model is a finely-tuned Social Value Orientation (SVO) that captures the nuanced social dynamics between multiple autonomous agents, spanning a continuum from self-interested to entirely cooperative behaviors. This allows AVs to make decisions socially and cooperatively. This framework significantly influences the AV navigation sector by contributing to the development of secure, human-centric, and reliable transportation systems. Its multi-agent focus and the incorporation of dynamic optimization emphasize its potential to facilitate a network of AVs that interact with diverse road users, thus improving scalability. The robustness and adaptability of this machine learning-powered solution are crucial for navigating the varied scenarios that characterize urban driving, ensuring that AVs can adapt to changing conditions and make decisions that benefit all road users.}, archive = {J_TASE}, author = {Hamid Taghavifar and Chuan Hu and Chongfeng Wei and Ardashir Mohammadzadeh and Chunwei Zhang}, doi = {10.1109/TASE.2025.3527327}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10672--10683}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Behaviorally-aware multi-agent RL with dynamic optimization for autonomous driving}, volume = {22}, year = {2025}, }
• (2025). Adaptive dynamic programming-based finite-time optimal backstepping Force/Position control of reconfigurable robot manipulators via pareto optimal. TASE, 22, 10660--10671. (www)

  @article{ , abstract = {To address the force/position control challenges in transitioning from free-space motion to tasks involving environmental contact, this paper proposes an Adaptive Dynamic Programming (ADP)-based finite-time optimal backstepping force/position control method for Reconfigurable Robot Manipulators (RRMs), which ensures rapid convergence of state errors under external constraints while maintaining system stability. By integrating robust control, the proposed method enhances both convergence speed and robustness against uncertainties. Furthermore, the parameters related to robustness are optimized using a cooperative game-theoretic approach based on Pareto optimality. A Lyapunov-based analysis demonstrates the closed-loop system’s Semi-Global Practical Finite-time Stability (SGPFS). Experimental validation confirms the effectiveness of the proposed control method. Note to Practitioners—In applications such as space operations, deep-sea exploration, and disaster rescue, RRMs must smoothly transition from free-space movement to tasks involving physical contact with external environments. Traditional force/position control methods often struggle to maintain stability and achieve rapid convergence under these conditions. This paper introduces a control strategy that combines ADP with backstepping to achieve finite-time optimal control. The proposed approach ensures stability, fast convergence, and robustness to uncertainties, addressing practical needs for energy efficiency and reliable force/position tracking. Moreover, the method optimizes control parameters through a cooperative game-theoretic framework based on Pareto optimality, balancing control effort and the convergence domain size. Experimental results confirm the practical effectiveness of the proposed strategy in complex environments.}, archive = {J_TASE}, author = {Yuexi Wang and Tianjiao An and Bo Dong and Mingchao Zhu and Yuanchun Li}, doi = {10.1109/TASE.2025.3527569}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10660--10671}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive dynamic programming-based finite-time optimal backstepping Force/Position control of reconfigurable robot manipulators via pareto optimal}, volume = {22}, year = {2025}, }
• (2025). Adjusting distributed cameras for robust moving object pose estimation. TASE, 22, 10650--10659. (www)

  @article{ , abstract = {Robust moving object pose estimation is crucial in fine manipulation tasks, such as surgical instrument tracking. This paper presents a distributed-camera system with robotic adjustments to maintain consistent tracking of moving objects, thus avoiding tracking failures. An integrated framework for camera adjustment and pose estimation is developed for this distributed-camera system. In each detection cycle, the camera exhibiting the largest deviation with the object is adjusted by a visual servoing technique. After adjustment, the camera extrinsics are re-calibrated in the following detection cycles. For the unadjusted cameras, an online extrinsic optimization method based on multi-frame detection results is proposed to refine the camera extrinsics. Based on the refined camera extrinsics and detection results from multiple cameras, the pose of moving objects relative to the principal camera can be robustly estimated. We test the performance of this system in both simulation environments and real-world scenarios. The results indicate that our system achieves higher pose estimation accuracy and exhibits strong resistance to limited field-of-view (FoV) compared to conventional equivalent fixed multi-camera systems. Note to Practitioners—The motivation of this work is to tackle the challenge of FoV limitation during moving object pose estimation. Most optical tracking systems address this issue through the fusion of multiple sensors. However, their optical tracking devices are typically fixedly installed, which can lead to failures in pose estimation when objects move out of the device’s FoV. To expand the tracking scopes while simultaneously ensuring the pose estimation accuracy, we proposed to apply narrow-FoV cameras of which the locations can be adjusted by gimbals to track moving objects, keeping them at the center of the FoV. In this way, these cameras can always detect points with a wider distribution at the center of the FoV, which is conducive to improving detection accuracy. Our proposed system achieves a significant resistance towards FoV limitation during the pose estimation process of objects in large-scale motion.}, archive = {J_TASE}, author = {Yaoqing Hu and Shaoan Wang and Dongyue Li and Xingyu Chen and Mingzhu Zhu and Zhanhua Xin and Junzhi Yu}, doi = {10.1109/TASE.2025.3527006}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10650--10659}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adjusting distributed cameras for robust moving object pose estimation}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered h ∞ filtering for fuzzy markov jump systems subject to mismatched quantization. TASE, 22, 10639--10649. (www)

  @article{ , abstract = {This paper is dedicated to a dynamic event-triggered $H_{\infty }$ filtering method of fuzzy Markov jump systems via a mismatched quantization scheme. The system outputs are triggered by a dynamic event-triggered mechanism and then quantized via a mismatched quantizer before being sent to the remote filter. The dynamic triggering scheme with a special diagonal matrix structure threshold is built to reduce the network burden. The quantizer is constructed in a multi-channel paradigm with a time-varying mismatch degree. Then, the remote reduce-order filter is designed to be both fuzzy-rule and mode-dependent. By adopting Finsler’s Lemma and the vertex separation method, sufficient conditions are derived in terms of form matrix inequalities. At last, the effectiveness of the proposed method is demonstrated by a tunnel diode circuit. Note to Practitioners—In practical networked systems, sampled analog signals must be quantized before being transmitted over a digital network. However, limited by imperfect hardware, the parameters of the encoder and decoder may not match. To address this challenge, this paper provides a mismatched quantizer design scheme. Additionally, frequent data transmission consumes limited energy and bandwidth resources. Conserving resources is essential for real industrial production, so a dynamic triggering scheme is proposed to reduce the data exchange frequency. A simulation example with practical background is presented to verify that the proposed scheme achieves satisfactory control performance.}, archive = {J_TASE}, author = {Yang Gu and Mouquan Shen and Ju H. Park and Qing-Guo Wang and Yang Yi and Yonghui Sun}, doi = {10.1109/TASE.2025.3527974}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10639--10649}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered h ∞ filtering for fuzzy markov jump systems subject to mismatched quantization}, volume = {22}, year = {2025}, }
• (2025). Discrete jacobian-pseudoinverse-free zhang neurodynamics algorithm handling path tracking of robot manipulator with unknown model. TASE, 22, 10625--10638. (www)

  @article{ , abstract = {Robot manipulator path tracking, recognized as a crucial aspect in robot manipulator control, has garnered significant attention from researchers. In this paper, to address the path tracking problem of robot manipulators with unknown models, a novel Jacobian pseudoinverse estimator is first proposed based on Zhang neurodynamics method. The estimator directly provides an efficient and accurate estimation of the Jacobian matrix pseudoinverse, avoiding the complicated operation of matrix pseudoinverse and preventing potential singularity phenomenon of the Jacobian matrix. By utilizing the Euler difference formulas, a discrete model-free and Jacobian-pseudoinverse-free Zhang neurodynamics algorithm is proposed. The proposed algorithm focuses on leveraging the available current and previous known information to predict the future unknown information. Detailed theoretical analyses and proofs ensure the convergence and stability of the proposed algorithm. Finally, comparative experiments with various effective model-free algorithms, and experimental validations on different types of robot manipulators (UR5, Franka Emika Panda, and Kinova Gen3 robot manipulators) using various experimental platforms (MATLAB, CoppeliaSim, and physical platforms) illustrate the effectiveness of the proposed algorithm. Note to Practitioners—This paper is motivated by addressing the prevalent challenge of unknown models in real-time path tracking for robot manipulators. In this paper, a novel discrete model-free and Jacobian-pseudoinverse-free Zhang neurodynamics algorithm is proposed. Different from the existing model-free algorithms, the proposed algorithm avoids the complicated operation of computing the pseudoinverse of matrix without compromising precision, significantly reducing the computational complexity and preventing potential singularity phenomenon of the Jacobian matrix. The average computation time per updating for the proposed algorithm is approximately $0.1\,{\mathrm { ms}}$ , which is significantly less than the sampling gap of the operation. This allows it to effectively meet the real-time requirements for robot manipulator path tracking. In addition, the error of the proposed algorithm is approximately $2\,\mu {\mathrm { m}}$ , which can meet the requirements of most practical application scenarios. Moreover, the accuracy of the algorithm is limited by the differential formula and sampling gap. Improving the accuracy and robustness of the algorithm by using more accurate difference formulas and filtering technique will be our future research direction.}, archive = {J_TASE}, author = {Jielong Chen and Yan Pan and Yunong Zhang and Ning Tan}, doi = {10.1109/TASE.2025.3526237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10625--10638}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discrete jacobian-pseudoinverse-free zhang neurodynamics algorithm handling path tracking of robot manipulator with unknown model}, volume = {22}, year = {2025}, }
• (2025). Qora: Neural-enhanced interference-aware resource provisioning for serverless computing. TASE, 22, 10609--10624. (www)

  @article{ , abstract = {Serverless is an emerging cloud paradigm that offers fine-grained resource sharing through serverless functions. However, this resource sharing can cause interference, leading to performance degradation and QoS violations. Existing white box-based approaches for serverless resource provision often demand extensive expert knowledge, which is challenging to obtain due to the complexity of interference sources. This paper proposes Qora, a neural-enhanced interference-aware resource provisioning system for serverless computing. We model the resource provisioning of serverless functions as a novel combinatorial optimization problem, wherein the constraints on the queries per second are derived from neural network performance model. By leveraging neural networks to model the nonlinear performance fluctuations under various interference sources, our approach better captures the real-world behavior of serverless functions. To solve the formulated problem efficiently, rather than adopting commercial optimizer solvers like Gurobi, we propose a two-stage-VNS algorithm that searches discrete variables more efficiently and supports Sigmoid activations, avoiding introducing redundant discrete variables. Unlike pure machine learning methods lacking theoretical optimal guarantees, our approach is rigorously proven globally optimal based on optimization theory. We implement Qora on Kubernetes as a serverless system automating resource provisioning. Experimental results demonstrate that Qora reduces the QoS violation rate by 98% while reducing up to 35% resource costs compared with the state-of-the-arts. Note to Practitioners—From the perspective of cloud service providers, this paper considers the automatic resource provisioning for serverless functions. To improve hardware utilization, cloud providers tend to co-locate serverless functions on the same server. However, co-located functions compete for shared resources (memory bandwidth, L3 cache, etc.), which causes interference and leads to performance degradation and QoS violations. We use neural networks to build the performance models of interference-prone serverless functions and form the resource allocation optimization problem with neural network performance models as constraints. Compared to white box modeling methods, our neural network modeling adapts to complex and variable interference. Compared to deep reinforcement learning methods, our combinatorial optimization methods have stronger interpretability. In order to solve this optimization problem efficiently, we design the two-stage-VNS solution algorithm. We implement Qora on Kubernetes as a serverless system, which can automatically allocate computing resources. Experiments with small-scale real clusters and large-scale simulations demonstrate the effectiveness of Qora.}, archive = {J_TASE}, author = {Ruifeng Ma and Yufeng Zhan and Chuge Wu and Zicong Hong and Yasir Ali and Yuanqing Xia}, doi = {10.1109/TASE.2025.3526197}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10609--10624}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Qora: Neural-enhanced interference-aware resource provisioning for serverless computing}, volume = {22}, year = {2025}, }
• (2025). Swarm robotic flocking with aggregation ability privacy. TASE, 22, 10596--10608. (www)

  @article{ , abstract = {We address the challenge of achieving flocking behavior in swarm robotic systems without compromising the privacy of individual robots’ aggregation capabilities. Traditional flocking algorithms are susceptible to privacy breaches, as adversaries can deduce the identity and aggregation abilities of robots by observing their movements. We introduce a novel control mechanism for privacy-preserving flocking, leveraging the Laplace mechanism within the framework of differential privacy. Our method mitigates privacy breaches by introducing a controlled level of noise, thus obscuring sensitive information. We explore the trade-off between privacy and utility by varying the differential privacy parameter $\epsilon $ . Our quantitative analysis reveals that $\epsilon \leq 0.13$ represents a lower threshold where private information is almost completely protected, whereas $\epsilon \geq 0.85$ marks an upper threshold where private information cannot be protected at all. Empirical results validate that our approach effectively maintains privacy of the robots’ aggregation abilities throughout the flocking process. Note to Practitioners—This paper was motivated by the problem of preserving privacy of individual robots in a swarm robotic system. Existing approaches to address this issue generally consider that accomplishing complex tasks requiring explicit information sharing between robots, while explicit communication in public channel carries the risk of information leakage. It is not always like this in real adversarial environments, and this assumption restricts the investigation of privacy in autonomous systems. This paper suggests that an individual robot can use its sensors onboard to perceive states of other neighbors in a distributed way without explicit communication. Despite avoiding information leakage during explicit information sharing between robots, the configuration of swarm can still reveal sensitive information about the ability of each robot. In this paper, we propose a privacy-preserving approach for flocking control using the Laplace mechanism based on the concept of differential privacy. The solution prevents an adversary with full knowledge of the swarm’s configuration from learning the sensitive information of individual robots, thus ensuring the security of swarm robots in terms of sensitive information during ongoing missions.}, archive = {J_TASE}, author = {Shuai Zhang and Yunke Huang and Weizi Li and Jia Pan}, doi = {10.1109/TASE.2025.3526141}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10596--10608}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Swarm robotic flocking with aggregation ability privacy}, volume = {22}, year = {2025}, }
• (2025). Safety-enhanced navigation planning for magnetic microrobots. TASE, 22, 10586--10595. (www)

  @article{ , abstract = {Magnetic microrobots demonstrate significant potential in medical applications by providing innovative solutions for precise treatment through targeted drug delivery, minimally invasive surgery, and vascular cleaning. However, within biological organisms, there are various complex obstacle environments that require a navigation technology prioritizing safety and emphasizing smoothness. This paper proposes a safety-enhanced navigation planning (SENP) algorithm to achieve multiple objectives such as safety, path smoothness, and short distance, enabling collision-free navigation in complex medical environments. Unlike traditional methods that require multiple heuristic cost functions to guide the navigation planning algorithm, our approach leverages the safety-enhancing features of the safe artificial potential field (SAPF) to promote collision-free navigation in complex medical environments. By adopting a goal-biased strategy to guide the generation of random sampling points, the number of iterations is reduced, and the convergence speed of the algorithm is improved. In addition, the traditional artificial potential field planning method often leads to the oscillation of the corridor, our method effectively addresses this issue along with the uncertainty in the quality of the initial path and the lengthy convergence time to the optimal path. Comparative analysis with various algorithms in different environments shows that our proposed method excels in terms of smoothness and path length under the premise of safety, making it suitable for magnetic microrobots in complex environments. Note to Practitioners—The motivation for this work lies in advancing safe, efficient, and rapid navigation strategies for magnetic microrobots in medical applications. While significant progress has been made in the development of magnetic microrobots, navigating through complex environments such as human blood vessels remains a substantial challenge. The ability to perform safe and effective motion planning within narrow and intricate channels is crucial for medical applications. In response to this need, we propose a path planning method specifically designed for magnetically actuated microrobots, based on SENP. Our approach emphasizes achieving a path that is not only short and smooth but also prioritizes safety throughout the navigation process. Compared to traditional sampling-based algorithms, our method effectively overcomes limitations such as initial path quality uncertainty and prolonged convergence to optimal paths. This allows for the rapid generation of a high-quality initial path with a focus on safety while significantly accelerating convergence. Moreover, our approach provides a robust solution for navigating narrow channels, making it highly suitable for challenging medical environments.}, archive = {J_TASE}, author = {Yueyue Liu and Linfeng Zhang and Xinyu Liu and Qigao Fan}, doi = {10.1109/TASE.2025.3525669}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10586--10595}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safety-enhanced navigation planning for magnetic microrobots}, volume = {22}, year = {2025}, }
• (2025). Fuzzy-based synchronization control for coupled neural networks under cyber attacks via intelligent impulsive algorithm. TASE, 22, 10574--10585. (www)

  @article{ , abstract = {This paper studies the fuzzy-based synchronization control problem of coupled neural networks under cyber attacks, where the considered attacks can block the communication links between fuzzy sub-neural networks. Firstly, an improved fuzzy network model is developed, which takes into account the inherent vulnerabilities of network. We design a fuzzy logic-based event-triggered delayed impulsive controller, where impulsive signals are generated by a dependent-Lyapunov intelligent impulsive selection algorithm. Particularly, it can mitigate attack effects, ensure the desired performance of fuzzy networks, and effectively exclude the Zeno behavior. Then, based on the proposed algorithm, some delay-dependent synchronization criteria are established for fuzzy networks on different scales delays, respectively. Finally, a practical example about resistance-capacitance circuit network is provided in different scenarios to show the validity of the theoretical results. Note to Practitioners—This paper was motivated by existing results on impulsive synchronization control of neural networks. Most existing results ignore the effects of external cyber attacks and delay during signal transmission, which is quite difficult to simulate the practical network model. This paper constructs a more general model to consider the inherent vulnerabilities of networks. Then, an intelligent impulsive selection algorithm is designed to resist the risks of attacks and obtain the expected performance. The obtained results are applied to resistance-capacitance circuit systems to verify the effectiveness, and it is expected that the proposed approach can be extended to more mechanical systems.}, archive = {J_TASE}, author = {Shiyu Dong and Kaibo Shi and Xiangpeng Xie and Mingyuan Yu and Huaicheng Yan and Xiao Cai}, doi = {10.1109/TASE.2025.3525658}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10574--10585}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fuzzy-based synchronization control for coupled neural networks under cyber attacks via intelligent impulsive algorithm}, volume = {22}, year = {2025}, }
• (2025). Small fault sample adversarial generation and diagnosis method for vehicular energy network. TASE, 22, 10560--10573. (www)

  @article{ , abstract = {This paper addresses the challenge of fault analysis in the Vehicular Energy Network (VEN) caused by small fault samples due to transient faults and complex disturbances. The proposed generation and diagnosis networks (GDNs) are developed without necessitating prior knowledge or manual intervention. The approach starts with an encoding and diagnosis network that converts multi-dimensional signals into images through supervised learning. A sample enhancement network, improved with module transfer and a relaxation objective function, is then proposed to increase the reliability of convergence and diversity of features for small fault samples. Additionally, a joint iterative training strategy between these two networks improves diagnostic accuracy and generalization through feature feedback. Performance validation on a semi-physical simulation platform demonstrates that the proposed GDNs achieve a 20% improvement in diagnostic accuracy with small datasets (200 samples) and maintain superior performance as sample volume grows. Thus, the proposed approach offers a potent solution for fault diagnosis in VENs with scarce samples, enhancing the analysis of complex systems. Note to Practitioners—This paper delves into fault diagnosis in the vehicular energy network (VEN) using small samples, employing a data-driven and deep learning model. The proposed method is versatile, suitable for analyzing complex systems with multiple and heterogeneous signals. An end-to-end model, named generation and diagnosis networks (GDNs), is introduced for generating small samples and conducting fault diagnosis without requiring prior knowledge or manual input. This method encodes multiple signals into signal images, which are then processed by a specially designed sample enhancement model, improved through the relaxation objective function and module transfer method. The enhanced samples are utilized for accurate analysis within the encoding and diagnosis networks’ diagnostic unit. The paper also provides a comparison of diagnosis results for reference. This approach enables researchers and engineers to efficiently augment and analyze small samples in practical applications, offering a practical framework for junior and inexperienced analysts. Preliminary experiments conducted using the RT-Lab semi-physical simulation platform suggest the method’s feasibility and effectiveness. Future research will explore the optimization of the model’s topology and parameters for lightweight design.}, archive = {J_TASE}, author = {Yongheng Pang and Dongsheng Yang and Mohammad Shojafar and Shuowei Jin and Mamoun Alazab and Shaohua Wan and Liang Zhao}, doi = {10.1109/TASE.2025.3525621}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10560--10573}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Small fault sample adversarial generation and diagnosis method for vehicular energy network}, volume = {22}, year = {2025}, }
• (2025). ROSec: Intra-process isolation for ROS composition with memory protection keys. TASE, 22, 10546--10559. (www)

  @article{ , abstract = {Robot Operating System (ROS) is a software framework for robotic systems that includes various packages for developing robotic applications. Composition is a package that combines multiple applications, namely, nodes, to be loaded and executed in a single process. However, permitting multiple nodes to share the address space could expand the attack surface such that vulnerabilities in a node are more likely to be exploited to subvert nodes running in the same space. We propose ROSec, an in-process isolation solution for ROS composition that utilizes Intel Memory Protection Keys. ROSec aims to enforce memory isolation between nodes within a process by preventing unauthorized access from one node to another. Unlike previous works that assume the number and sizes of nodes are statically defined and partitioned by developers, ROSec is designed to handle the dynamic nature of nodes that can be loaded and executed in a process at any time during execution. To achieve this, ROSec adopts a unique scheduling mechanism that utilizes the executor-centric execution model of ROS to perform two main operations for MPK-based isolation: protection key assignment and reassignment. Our evaluation shows that ROSec effectively enforces in-process isolation while incurring a 6.4% performance overhead on a real-world application. Note to Practitioners—Cyber-Physical Systems are the core of modern applications, particularly robotics, as they integrate computing and physical processes. ROS necessitates real-time and security guarantees, which, unfortunately, trade-off with each other. While traditional ROS architecture relies on process isolation to separate various nodes, ROS2 introduces a feature called composition, which allows multiple nodes to run inside a single process, thus exposing various nodes to potential malicious compromises from others. This paper proposes a technique that utilizes Intel memory protection keys (MPK) to provide intraprocess isolation for ROS composition. Given that ROS nodes are dynamic, ROSEC provides key assignment and reassignment techniques to configure MPK dynamically.}, archive = {J_TASE}, author = {Jiwon Seo and Martin Kayondo and Jeonghwan Kang and Kyeongryong Lee and Donghyun Kwon and Yunheung Paek}, doi = {10.1109/TASE.2024.3525050}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10546--10559}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ROSec: Intra-process isolation for ROS composition with memory protection keys}, volume = {22}, year = {2025}, }
• (2025). Distributed intelligent control method based on state self-learning and its application in cascade processes. TASE, 22, 10533--10545. (www)

  @article{ , abstract = {The multiple-reactor cascade operation is a distinctive characteristic in the process industry. However, it is difficult to establish an accurate and global model for multi-reactor cascading processes. Moreover, the intricate and dynamic operating state of the reactor, coupled with rear reactors, poses significant challenges to the fine control of the entire process. Therefore, this paper proposes a distributed intelligent control method based on state self-learning. Initially, the time-varying dynamic model of each reactor unit is established by learning the parameters of the regression model at each state point, achieving a nonlinear description of the reactor under complex conditions. Subsequently, leveraging the dynamic model and the material conservation principle between reactors, multi-step collaborative prediction is conducted along the reactor cascade direction. Thirdly, distributed model predictive control based on error self-correction is adopted to realize distributed intelligent control of the cascade reactor. This method is verified in a zinc smelting leaching process. The results indicate its superiority over common methods, offering higher prediction accuracy for the cascade process and enabling more effective control of individual reactors through distributed intelligence, which provides a novel and promising control paradigm for the cascade process.Note to Practitioners—The future heralds an era of the Internet of Everything, and this transformation extends to the production processes of the process industry. Presently, decentralized control methods are prevalent in the process industry. However, these methods lack communication between controllers and autonomy in learning. While decentralized control methods can effectively regulate most industrial processes, they struggle to achieve optimal control in scenarios with cascading reactors, where the reactors are interdependent. Distributed control methods offer promise to address these limitations. Regrettably, research and application of distributed control in process industry engineering remain limited, lacking suitable methods for controller autonomy and information exchange among different controllers. Consequently, this paper presents a novel control approach for stable and efficient regulation of multi-reactor cascades in the process industry, offering promising avenues for widespread application.}, archive = {J_TASE}, author = {Shulong Yin and Yonggang Li and Zhenxiang Feng and Bei Sun and Huiping Liang}, doi = {10.1109/TASE.2024.3524472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10533--10545}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed intelligent control method based on state self-learning and its application in cascade processes}, volume = {22}, year = {2025}, }
• (2025). PCDCT: Perception-complementarity-driven collaborative trajectory generation for vision-based aerial tracking. TASE, 22, 10520--10532. (www)

  @article{ , abstract = {This article proposes a perception-complementarity-driven trajectory generating method for multiple unmanned aerial vehicles (UAVs), which can effectively enhance the visibility of the target for UAVs in unknown environments. Traditional methods often rely on prior maps or additional sensors to assist with obstacle avoidance. Nevertheless, these methods are not only costly but also fail to effectively avoid occlusions caused by obstacles. Different from existing methods, the trajectory planned by the proposed method not only enables the vision-based UAVs to maintain the perception of obstacles and the target on the one hand, but also preserves topological equivalence with the predicted target trajectory on the other hand. Specifically, a vision-based mutual observation approach among UAVs is proposed to enhance the overall perception capability of the UAVs system. On this basis, a target-guided collaborative trajectory planning method is proposed to ensure the planned collision-free trajectory for other UAV in the formation maintains target visibility. In addition, a trajectory feasibility assessment method is proposed to obtain the collaborative trajectory planned by the UAV at the optimal observation location. Finally, comparative simulations are conducted with three state-of-the-art methods, demonstrating the advantages of the proposed method in maintaining target visibility and tracking efficiency during the vision-based aerial tracking. The real-world experiment demonstrates the feasibility of the proposed method. Note to Practitioners—Most existing vision-based multi-UAV target tracking methods require additional prior maps or laser sensors to assist in obstacle avoidance. In practical tracking scenarios, the limited perception range of cameras poses significant challenges for UAVs in synchronously observing moving target and environmental obstacles. The article proposes a trajectory generation method based on perceptual complementarity to ensure that the generated trajectory effectively perceives surrounding obstacles while enhancing the observation capability of UAVs towards moving target. The key insight of this work is to utilize mutual observation among multiple UAVs to assist in planning collision-free tracking trajectories, ensuring safety during the tracking process. Building upon this, by preserving topological equivalence with the predicted target trajectory, there is an improvement in the target visibility ratio during the tracking process. Furthermore, a trajectory feasibility assessment method is proposed to obtain the optimal collaborative trajectory from the UAV positioned at the optimal location in the formation. The effectiveness of the proposed method in improving flight safety and target visibility is validated through comparative experiments.}, archive = {J_TASE}, author = {Hanzhang Wang and Xuetao Zhang and Yisha Liu and Gang Sun and Xuebo Zhang and Yan Zhuang}, doi = {10.1109/TASE.2024.3524439}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10520--10532}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PCDCT: Perception-complementarity-driven collaborative trajectory generation for vision-based aerial tracking}, volume = {22}, year = {2025}, }
• (2025). Language-guided dexterous functional grasping by LLM generated grasp functionality and synergy for humanoid manipulation. TASE, 22, 10506--10519. (www)

  @article{ , abstract = {Dexterous Functional Grasping (DFG) is the crucial first step for humanoid robots to perform generalized manipulation tasks. However, enabling robots to learn language-guided DFG skills in real-world environments presents several challenges, including comprehending the complex relationship between task instructions and grasp functionality, generating feasible functional grasps of dexterous hands, and handling generalization for novel functional concepts. To tackle these challenges, we introduce SayFuncGrasp, a Large Language Model (LLM) based DFG framework that can synthesize versatile dexterous functional grasps from language instructions and achieve generalization on novel functional concepts. SayFuncGrasp first harnesses the open-ended manipulation knowledge from an LLM to infer grasp functionality based on language instructions. Subsequently, it employs the inferred grasp functionality to synthesize plausible DFG actions characterized by hand synergies. Simulation experiments show that SayFuncGrasp significantly outperforms the baseline method in open-set grasp functionality generalization. Real robot experiments demonstrate the effectiveness and generalizability of SayFuncGrasp for interactive humanoid manipulation tasks, achieving an overall grasp success rate of 64.66% and a manipulation success rate of 70.41%. Note to Practitioners—This research was motivated by the practical challenge of enabling humanoid robots with high-DoF dexterous hands to perform functional grasping based on verbal instructions. In industrial settings, such capabilities can significantly enhance the versatility and adaptability of humanoid assistants, allowing them to perform complex manipulations simply by being told what to do, thereby reducing programming complexity and increasing flexibility. Current dexterous functional grasping methods rely solely on visual input, without the ability to process language instructions. Furthermore, they are restricted to pre-defined functional concepts and cannot be generalized to novel object classes and manipulation tasks within natural language. Our newly proposed language-guided dexterous functional grasping system takes advantage of open-ended manipulation knowledge from LLMs to produce generalized functional grasps of dexterous robot hands according to verbal commands. Our experiment results demonstrate improved versatility and generalizability compared to the state-of-the-art.}, archive = {J_TASE}, author = {Zhuo Li and Junjia Liu and Zhihao Li and Zhipeng Dong and Tao Teng and Yongsheng Ou and Darwin Caldwell and Fei Chen}, doi = {10.1109/TASE.2024.3524426}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10506--10519}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Language-guided dexterous functional grasping by LLM generated grasp functionality and synergy for humanoid manipulation}, volume = {22}, year = {2025}, }
• (2025). Guaranteed 2D pose graph SLAM with bounded noises: An efficient interval approach. TASE, 22, 10494--10505. (www)

  @article{ , abstract = {This paper focuses on developing a performance guaranteed state estimation algorithm for 2D pose graph problems for mobile robots. Different from probabilistic methods, the measurement noises are only assumed to be bounded without any prior knowledge about their distributions. Based on the interval analysis, we first propose a vanilla sequential contractor that iteratively uses edge-wise noise bounds to contract pose intervals at the nodes, which can provide the guaranteed feasible domains that contain the ground-truth values. Then, to improve the efficiency in solving large-scale pose graphs, an efficient batch contractor is developed by improving the update order and exploiting a relaxation of the nonlinear measurement functions. The effectiveness and efficiency of our approaches are validated on simulated and real-world datasets. Note to Practitioners—Pose graph is one of the most popular formulations for the state estimations of mobile robots. There have been many probabilistic algorithms for pose graphs based on the Gaussian-like measurement noise assumption. However, the measurement noises in many practical situations may not follow Gaussian distributions but have hard bounds. Consequently, the existing pose graph algorithms are far away from achieving the expected high reliability in the practical safety-critical applications such as autonomous driving. To achieve guaranteed performance, an efficient interval based approach is proposed for the large-scale pose graph problems with hard bound measurement noises. It can provide the guaranteed hard error bounds for the robot poses, which has the potential in uncertainty quantification, reliability analysis and outlier detection of safety-critical systems.}, archive = {J_TASE}, author = {Yang Song and Heng Yang and Liang Zhao and Shoudong Huang}, doi = {10.1109/TASE.2024.3524373}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10494--10505}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guaranteed 2D pose graph SLAM with bounded noises: An efficient interval approach}, volume = {22}, year = {2025}, }
• (2025). Passivity-based connectivity maintenance of teleoperated multi-robots under DoS attacks. TASE, 22, 10484--10493. (www)

  @article{ , abstract = {Teleoperated multi-robots rely on communications, both between the human’s local robot and the multiple remote robots and among the remote robots themselves, to execute the remote tasks commanded by their human operator. If attacked, the communications may lead to task failure, loss of multi-robot connectivity and possibly unstable teleoperation. To render teleoperated multi-robots resilient to Denial of Service (DoS) attacks with arbitrary frequency and duration, this paper augments a passivity-based controller for normal teleoperation with: 1) a controller that stops the remote robots at safe distances from each other and from obstacles when a DoS attack starts; and 2) a controller that restores the multi-robot connectivity before resuming normal teleoperation when a DoS attack stops. The teleoperation of a simulated multi-robot system with one leader and two followers illustrates the effectiveness of the proposed distributed control strategy. Note to Practitioners—Research and industry increasingly seek to use robots in inaccessible unstructured environments. Teleoperated multi-robots are ideally suited for such environments because they fuse human cognition with remote multi-robotic execution. However, they can be hindered by cyber attacks on their communications. As cyber attacks become more prevalent, resilience to them becomes increasingly important for practical teleoperated multi-robots. This paper presents a first distributed control strategy for rendering teleoperated multi-robots resilient to DoS attacks. For industrial practitioners, the proposed strategy has two key advantages: 1) it is straightforward to implement; and 2) it is effective for DoS attacks with arbitrary frequency and duration. Future work will tackle teleoperated multi-robots under malicious attacks.}, archive = {J_TASE}, author = {Liu Yang and Daniela Constantinescu and Ligang Wu}, doi = {10.1109/TASE.2024.3524358}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10484--10493}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Passivity-based connectivity maintenance of teleoperated multi-robots under DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Distributed hybrid dynamic event-triggered consensus control for nonlinear multi-agent systems. TASE, 22, 10470--10483. (www)

  @article{ , abstract = {This article addresses both leaderless and leader-following consensus issues for Lipschitz nonlinear multi-agent systems. To begin with, the issues are discussed in view of leaderless scenarios. Firstly, a distributed dynamic event-triggered mechanism is introduced to mitigate continuous communication burdens among neighboring agents. This mechanism incorporates an open-loop estimation algorithm and an inner self-learning term into the triggering conditions. Secondly, to prevent Zeno behavior, a time/event hybrid mechanism is implemented. For each agent, based on the local state information at the current event-triggered instant and the recent event-triggered instant receiving from neighbors, open-loop state estimations are conducted. Then, by utilizing these open-loop state estimations, a distributed adaptive control protocol is developed within the framework of the hybrid dynamic event-triggered mechanism, including an updating mechanism for the coupling strength of each agent. The challenge posed by the Lipschitz nonlinearity is addressed by solving a Riccati equation. Additionally, the proposed method is improved to be suitable for leader-following multi-agent systems. Finally, simulation examples demonstrate the effectiveness of the proposed method. Note to Practitioners—The purpose of this paper is to introduce a hybrid dynamic event-triggered mechanism aimed at reducing the communication load in Lipschitz nonlinear multi-agent systems. By integrating a hybrid mechanism that combines both time-based and event-based triggers, the proposed scheme effectively minimizes unnecessary data exchanges among agents while naturally excluding the Zeno behavior. This is particularly advantageous in practical scenarios where bandwidth is limited or efficiency in data transmission is highly required in applications. The introduction of open-loop estimators ensures that continuous communication between two neighboring agents is not required. Furthermore, the introduction of an adaptive control protocol, which incorporates a projection algorithm for dynamically adjusting coupling gains, ensures that the interactions between agents remain within safe operational limits. The projection algorithm ensures that the adaptive gain remains within a prescribed range, thus preventing it from growing excessively and preserving system stability.}, archive = {J_TASE}, author = {Yuanqing Wang and Fanglai Zhu}, doi = {10.1109/TASE.2024.3524258}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10470--10483}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed hybrid dynamic event-triggered consensus control for nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). A priority-based multi-robot search algorithm for indoor source searching. TASE, 22, 10457--10469. (www)

  @article{ , abstract = {It is extremely important to quickly locate the source of a hazardous substance leak in order to reduce damage to life and property. Multi-robot source localization faces challenges in unknown indoor environments, such as navigating through dense environments, encountering large areas without airflow or concentration clues, experiencing frequent changes in robot measurements, and managing clusters of robots in confined spaces. This study proposes a priority-based multi-robot search algorithm to tackle these challenges. The algorithm consists of a priority-based search strategy, an exploration method based on frontier and Voronoi diagram, an airflow tracking method based on Rapidly-exploring Random Trees Star (RRT*), and a multi-robot collaborative method. The algorithm was compared with three other state-of-the-art algorithms in simulated environments, assessing varying team sizes, airflow speeds, and diverse scenarios. The algorithm was also evaluated in real-robots experiments. The evaluation results demonstrate that the algorithm exhibits outstanding performance in both simulated and real-robots experiments. Note to Practitioners—The aim of this study is to propose a multi-robot search algorithm designed to address the challenges encountered in source searching within unknown indoor environments. These challenges include navigating through dense environments, large areas without airflow and concentration clues, frequent changes in robot measurements, and the clusters of robots in confined spaces. This study proposes a priority-based multi-robot search algorithm. The core idea of the algorithm is to enable robots to adopt different search methods depending on their measurements. When lacking valid airflow and concentration measurements, robots engage in spatial exploration using an exploration method based on frontier and Voronoi diagram to increase the likelihood of encountering the plume. When robots are within a plume, they rely on an RRT*-based airflow tracking method to move towards the source. The RRT* also provides robots with reachable navigation goals in environments with dense obstacles. The multi-robot collaborative method operates based on the priority levels assigned to the robots. On one hand, it directs robots towards the global best to reduce search time. On the other hand, robots ignore global bests that are not of a higher priority than their own, preventing the clustering of multiple robots at the same location. The algorithm was compared with three other state-of-the-art algorithms in simulated environments. It was also evaluated in real-robots experiments. The evaluation results show that the algorithm exhibits outstanding performance in both simulated and real-robots experiments. To further aid research in this field, a dataset for multi-robot indoor source searching has been created and made available online to provide a benchmark for related research.}, archive = {J_TASE}, author = {Miao Wang and Bin Xin and Mengjie Jing and Yun Qu}, doi = {10.1109/TASE.2024.3524237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10457--10469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A priority-based multi-robot search algorithm for indoor source searching}, volume = {22}, year = {2025}, }
• (2025). Simulation optimization of spatiotemporal dynamics in 3D geometries. TASE, 22, 10442--10456. (www)

  @article{ , abstract = {Many engineering and healthcare systems are featured with spatiotemporal dynamic processes. The optimal control of such systems often involves sequential decision making. However, traditional sequential decision-making methods are not applicable to optimize dynamic systems that involves complex 3D geometries. Simulation modeling offers an unprecedented opportunity to evaluate alternative decision options and search for the optimal plan. In this paper, we develop a novel simulation optimization framework for sequential optimization of 3D dynamic systems. We first propose to measure the similarity between functional simulation outputs using coherence to assess the effectiveness of decision actions. Second, we develop a novel Gaussian Process (GP) model by constructing a valid kernel based on Hausdorff distance to estimate the coherence for different decision paths. Finally, we devise a new Monte Carlo Tree Search (MCTS) algorithm, i.e., Normal-Gamma GP MCTS (NG-GP-MCTS), to sequentially optimize the spatiotemporal dynamics. We implement the NG-GP-MCTS algorithm to design an optimal ablation path for restoring normal sinus rhythm (NSR) from atrial fibrillation (AF). We evaluate the performance of NG-GP-MCTS with spatiotemporal cardiac simulation in a 3D atrial geometry. Computer experiments show that our algorithm is highly promising for designing effective sequential procedures to optimize spatiotemporal dynamics in complex geometries. Note to Practitioners—This article proposes a novel simulation optimization framework for sequential decision making to optimize spatiotemporal dynamics in complex geometries. This framework incorporates the advantage of Bayesian modeling and Gaussian Process inference into Monte-Carlo tree search to effectively solve the sequential optimization problem. It has significant potential to contribute to the emerging discipline of computational engineering and medicine, and further realize precision control/treatment planning in various manufacturing and healthcare systems. This paper will be interesting to practitioners who are seeking effective computational and optimization tools for decision support to optimally control dynamic systems for restoring normal system functionality.}, archive = {J_TASE}, author = {Bing Yao and Fabio Leonelli and Hui Yang}, doi = {10.1109/TASE.2024.3524132}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10442--10456}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Simulation optimization of spatiotemporal dynamics in 3D geometries}, volume = {22}, year = {2025}, }
• (2025). STR: Spatial-temporal RetNet for distributed multi-robot navigation. TASE, 22, 10429--10441. (www)

  @article{ , abstract = {The core of multi-robot collision avoidance is to guide robots to avoid collisions with other robots and obstacles in a dynamic multi-robot environment, which has recently gained increasing interest among the main challenges of robotics. However, the current multi-robot navigation policy neural network exhibits weak position encoding capabilities for spatial environmental features in mapping environment states and robot actions, as well as an inability to recurrently infer information on dynamic environmental features in the temporal dimension, leading to insufficient safety and effectiveness in guiding robot motion. In this paper, we propose a novel spatial-temporal RetNet (STR) that encodes reciprocal collision avoidance states between robots in both spatial and temporal dimensions, aiming to enhance the safety and effectiveness of the policy neural network in guiding robots to accomplish specified tasks. The spatial state encoder module is developed based on parallel RetNet structure, which enhances the ability of the neural network in multi-robot navigation policies to extract reciprocal collision avoidance states between robots in spatial dimensions and overcomes the weak position encoding capability of advanced transformer-based multi-robot navigation policy neural networks. A temporal state encoder is designed by introducing the recurrent RetNet structure. This enhances the multi-robot navigation policy neural network’s ability to encode features in the temporal dimension of multi-robot movements and overcomes the transformer-based multi-robot navigation policy neural network’s inability to recurrently infer information in the time dimension. Simulation experiments were designed to demonstrate that the safety and effectiveness of our proposed method outperform the previous state-of-the-art approaches in guiding the robot to complete the task. Physical experiments illustrate that our policy can be effectively applied to real-world systems. Note to Practitioners—Multi-robot navigation has a wide range of real-world applications, such as multi-robot formation flying for search and rescue, autonomous warehouse operations, and robots navigating through human crowds. This paper introduces a novel Spatial-Temporal RetNet (STR) framework aimed at enhancing safety and effectiveness in multi-robot collision avoidance. STR addresses the limitations of existing methods by improving the neural network’s ability to extract reciprocal collision avoidance states in both spatial and temporal dimensions. The spatial state encoder strengthens the extraction of spatial features, while the temporal state encoder improves the handling of time-dependent information. Simulation and physical experiments demonstrate that STR enhances robot navigation in dynamic environments, making it suitable for real-world applications such as multi-robot coordination.}, archive = {J_TASE}, author = {Lin Chen and Yaonan Wang and Zhiqiang Miao and Mingtao Feng and Yuanzhe Wang and Yang Mo and Wei He and Hesheng Wang and Danwei Wang}, doi = {10.1109/TASE.2024.3524000}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10429--10441}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {STR: Spatial-temporal RetNet for distributed multi-robot navigation}, volume = {22}, year = {2025}, }
• (2025). Behavior scheduling for multi-robot path planning in unknown environment with communication constraints. TASE, 22, 10417--10428. (www)

  @article{ , abstract = {The multi-robot system that combines flexibility and autonomy has unique advantages in many scenarios. In unknown environment, the threat posed by the environmental uncertainty to the path planning process cannot be ignored. The robots not only need to avoid obstacles, but also need to maintain communication with other robots to ensure timely negotiation. Inspired by the division of labor in biological community, this paper proposes robot behavior scheduling algorithm for multi-robot path planning in unknown environment with communication constraints. First, to avoid the occurrence of oscillation caused by a single force determining the robot behavior, the constituent elements of robot behavior are described in the form of multiple forces. Then, various robot behaviors are instantiated through a diverse combination of these elements and parameters. Finally, two complementary robot behavior scheduling methods are designed. The experiential potential field jointly constructed by robots is proposed to guide the robots to juggle avoidance, communication and planning task for the first time. The effectiveness of these behavior scheduling methods is verified in theory. The simulation experiment results based on real sensor parameters show that a small amount of robot detours can make the multi-robot system fully connected for 70% of the entire path planning process. Note to Practitioners—The motivation of this article is to avoid communication interruptions among robots when planning paths in unknown environment. When the robots perform the complex tasks, communication interruptions will lead to the inability of sharing environmental maps and allocation plans. The lack of such information may delay the task progress and even cause greater loss. This article focuses on the path planning demands in unknown environment and considers the automatic maintenance and reconstruction of the communication links among robots. Based on the analyses of the motivation in reaching the target point, avoiding obstacles, and maintaining communication and etc., we first model the constituent elements of robot behavior. Based on the designed robot behavior, the robot behavior scheduling algorithm is proposed. The experiment results demonstrate that the effectiveness of the proposed algorithm is not affected by the changes in task scenarios, signal loss conditions and the number of robots. The experiments on Turtlebot3 burger robots further validates the practicality of the proposed algorithm. Additionally, the robot behavior scheduling algorithm has scalability and can enable the robots to complete more complex tasks by enriching the types of robot behaviors and scheduling mechanisms.}, archive = {J_TASE}, author = {Hui Lu and Meng Zhao and Ping Zhou and Kefei Mao}, doi = {10.1109/TASE.2024.3523875}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10417--10428}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Behavior scheduling for multi-robot path planning in unknown environment with communication constraints}, volume = {22}, year = {2025}, }
• (2025). Adaptive fault-tolerant tracking control for nonlinear systems with unknown actuator and multiple sensor faults. TASE, 22, 10406--10416. (www)

  @article{ , abstract = {This paper formulates fault-tolerant tracking control for high-order nonlinear systems with actuator faults, and each state signal is measured by a single sensor which possibly suffers partial failure. The unknown fault parameters of sensors and actuator are assumed to be time-varying and their upper and lower bounds are unknown. A parameter separating tactic is employed to tackle the coupling issues in view of real state signals and unknown fault parameters. The actuator fault with high-order power $u^{\kappa _{n}}(t)$ is transformed into a linear function of $v^{\kappa _{n} }(t)$ plus a bounded term $dv(t)$ to overcome the restriction of high-order power and separate the fault parameters, such that the control signal can be constructed directly. Neural networks (NNS) are utilized to identify the unknown coexisting nonlinear uncertainties. By formulating adaptive bound estimation scheme to construct the NNS-based fault-tolerant tracking controller, which can eliminate the effects of multiple sensor faults, actuator faults, external disturbances and identify errors. Finally, the simulations verify the feasibility of the proposed tactic. Note to Practitioners—HNSs has a more general system structure. In practice, many systems can be described or transformed as HNSs, such as space vehicles and mechanical systems. Owing to the influence of environment and the physical components of the practical engineering systems, sensor faults and actuator faults are inevitable in engineering systems. Due to the restrictions of high-order power, the actuator faults cannot be converted into a friendly form such that the controller cannot be designed directly. Thus, the conventional results often assume the powers equals to 1 or the fault parameters are known.}, archive = {J_TASE}, author = {Junchang Zhai and Yuping Qin and Hongxia Cui}, doi = {10.1109/TASE.2024.3523677}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10406--10416}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fault-tolerant tracking control for nonlinear systems with unknown actuator and multiple sensor faults}, volume = {22}, year = {2025}, }
• (2025). Hierarchical physics-informed neural network for rotor system health assessment. TASE, 22, 10392--10405. (www)

  @article{ , abstract = {Due to coupled nonlinearities and complex measurement noise, assess the condition of the rotor system remains a challenge, particularly in cases where historical run-to-failure data is lacking. To this end, we proposed a hierarchical physics-informed neural network (HPINN) to identify/discover the ordinary differential equations (ODEs) of a healthy/faulty rotor system from noise measurements and then assess the rotor condition based on the discovered ODEs. Specifically, the ODEs of a healthy rotor system are first stably identified from noisy measurement through HPINN guided by rotor dynamics. Based on the identified healthy ODEs, the extra fault terms in the ODEs of the faulty rotor system are then sparsely regressed from the predefined library embedded in HPINN, in which the phase compensation and alternating training strategy are developed to guarantee training convergence. Moreover, with the mathematical terms of discovered fault, the potential fault and the health indicator (HI) are diagnosed and constructed to assess the condition of the rotor system, respectively. Finally, the effectiveness of the proposed method is verified with simulation and test bench datasets, showing the potential for practical industrial applications. Note to Practitioners—This paper investigates the health assessment problem (condition monitoring and fault diagnosis) of the rotor system, a critical component in large rotating machinery. The proposed HPINN provides a hierarchical framework to firstly identify the ODEs of healthy rotor system and then discover the ODEs of faulty rotor system with limited monitoring data (3-5 seconds data collected from sensor commonly, depending on the rotating speeds). With the mathematical terms of discovered fault, the fault can be diagnosed and a health indicator (HI) can be constructed to assess the condition of rotor system in a fully interpretative way. This approach is applicable to large rotating machinery in safety-critical industries, such as circulating water pumps.}, archive = {J_TASE}, author = {Xue Liu and Wei Cheng and Ji Xing and Xuefeng Chen and Zhibin Zhao and Lin Gao and Rongyong Zhang and Qian Huang and Hongpeng Zhou and Wei Xing Zheng and Wei Pan}, doi = {10.1109/TASE.2024.3523417}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10392--10405}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical physics-informed neural network for rotor system health assessment}, volume = {22}, year = {2025}, }
• (2025). Nonlinear causal discovery via dynamic latent variables. TASE, 22, 10381--10391. (www)

  @article{ , abstract = {Distinguishing causality from mere correlation is a cornerstone in empirical research, as conflating the two can result in significant errors in decision-making, affecting policy formulation and the validity of scientific inferences. Traditional experimental designs, such as randomized trials, often fall short in complex systems where variables interact in a high-dimensional space with limited data. This paper aims to address these challenges by introducing an innovative causal discovery approach, extending beyond conventional methodologies by incorporating algorithmic advances in computational efficiency and design. We present a novel double Gaussian process state space causal model (GPSSCM) that contends with the multifaceted nature of causal inference, accounting for noisy observations and latent variables, which are commonly encountered in dynamic systems. Our methodological contribution includes the application of a Markov chain Monte Carlo technique for unraveling latent state dynamics and an expectation-maximization (EM) algorithm for robust parameter estimation. The acyclic nature of the causal graph is ensured through an integrated acyclic constraint within the EM framework, maintaining the integrity of the causal model. The efficacy of our proposed GPSSCM is evaluated through a series of tests on both synthetic data and empirical case studies from the industrial domain. The results highlight the model’s capacity to accurately infer complex nonlinear causal relationships, demonstrating its superiority over traditional structural equation modeling, especially when dealing with time series data and latent variables. This paper not only contributes a sophisticated tool for researchers and practitioners but also enriches the literature on causal discovery by offering a new perspective on the analysis of intricate systems, thereby facilitating more informed and ethical decision-making across various scientific fields. Note to Practitioners—Understanding the intricate web of causality is crucial for making informed decisions in various scientific and professional fields. Our study presents a Gaussian process state space causal model, which enhances the analysis of dynamic causal relationships in complex systems, particularly when dealing with noisy observations and latent variables. Leveraging a combination of Markov chain Monte Carlo and expectation maximization algorithms, the model ensures accurate estimation of parameters and causal structures. This paper is particularly relevant for those in fields such as economics, transportation, and biology, offering a sophisticated tool to support ethical decision-making and safeguard against errors in high-stakes environments. The practical implications of this research are underscored by its ability to inform targeted interventions and predict outcomes under new conditions, advancing the comprehension and application of causality in real-world scenarios.}, archive = {J_TASE}, author = {Xing Yang and Tian Lan and Hao Qiu and Chen Zhang}, doi = {10.1109/TASE.2024.3522917}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10381--10391}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nonlinear causal discovery via dynamic latent variables}, volume = {22}, year = {2025}, }
• (2025). Memorize my movement: Efficient sensorimotor navigation with self-motion-based spatial cognition. TASE, 22, 10368--10380. (www)

  @article{ , abstract = {Navigation is a fundamental capability for robots to operate in expansive spaces. Reliable navigation in unknown environments is crucial for deploying robots in areas such as disaster rescue and industrial inspection. In such scenarios, it is essential for robots to construct memories based on historical data to support long-term, optimized decision-making. However, many existing techniques focus on memorizing direct features from raw perceptions, often resulting in redundancy due to irrelevant textures and areas. This approach leads to inefficiencies in computation and storage, and produces a memory structure that lacks general applicability. We suggest that it may not be necessary to store specific scene features. Instead, recalling the robot’s episodic movements could provide sufficient cognitive cues for navigation. To address this, we introduce Memory Enhanced Navigation with Embedded Odometry (MENEO), a framework consisting of three steps: ego-motion estimation, memory aggregation, and adaptive policy generation. MENEO offers two main advantages: its streamlined architecture significantly boosts computational and storage efficiency, and its universal design supports various sensor modalities, adapts to multiple navigation tasks, and accommodates different scenarios. We test MENEO in two different environments: maze exploration using a lidar-IMU sensor, and image-goal visual navigation in photorealistic indoor scenes. In both cases, MENEO demonstrates competitive navigation performance, outperforming existing methods by reducing storage and computational requirements. Additionally, MENEO’s compact memory representation not only enhances adaptability across diverse environments but also shows flexibility in real-world applications. Note to Practitioners—In learning-based navigation, the memory mechanism is essential for long-term optimized policies. It allows intelligent robots to make informed decisions by utilizing a wide range of temporal and spatial cues derived from historical data. Traditional methods use various types of memory (such as internal, unstructured, or structured), but these often result in computational and storage inefficiencies due to the direct inclusion of complex and redundant raw scene features. Furthermore, because these memory systems are closely linked to specific scene features, they lack general applicability across different sensor configurations, scene types, and tasks. In this paper, we aim to eliminate the need to directly manage the redundant environmental features found in previous memory structures. Instead, we propose focusing solely on memorizing a robot’s self-movements. Since the pose sequence is streamlined and compact, our approach not only enhances computational and storage efficiency but also improves the interpretability and universality of the navigation system. This advantage enables MENEO to be seamlessly integrated into a wide variety of intelligent navigation systems. It is especially beneficial for small robots with limited computing power, such as those used in search and rescue operations, where enhanced memory can greatly enhance their autonomous navigation capabilities.}, archive = {J_TASE}, author = {Qiming Liu and Dingbang Huang and Zhe Liu and Hesheng Wang}, doi = {10.1109/TASE.2024.3522665}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10368--10380}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Memorize my movement: Efficient sensorimotor navigation with self-motion-based spatial cognition}, volume = {22}, year = {2025}, }
• (2025). RIGNet: Robot intention grasp for dense stacked targets with multi-task siamese schema through RoIs learning. TASE, 22, 10354--10367. (www)

  @article{ , abstract = {Autonomous grasping is a critical topic of robotic embodied intelligence. However, it remains challenging for robots to grasp an intended target, particularly in cluttered and densely stacked environments. This paper presents a novel solution by proposing a Robot Intention Grasp Network (RIGNet) with a multi-task siamese schema, which is based on an improved Region Proposal Network (RPN) and a Region of Interest (RoI) learning approach. More specifically, the RPN robustly outputs the RoIs to describe the candidates in stacked scenes, while the multi-task siamese network consisting of category comprehension and grasp perception modules serves to detect the intended object and generate the optimal grasp configuration. To improve the reasoning precision for grasp posture, a dynamic-oriented anchor matching strategy is further proposed to adapt to the grasp perception module. The proposed RIGNet is comparable to the state-of-the-art algorithms on single-object tasks and has better performance on stacked multi-object grasp detection. This study can provide new insights for autonomous robot reasoning in real time, facilitating an understanding of both the rationale and methodology for grasping an intended target in highly stacked environments. Note to Practitioners—This work is motivated by robotic embodied intelligence technique, which plays an irreplaceable role in robot autonomous grasping manipulation. Using this technology, the robot is capable of performing grasp tasks for human. The previous grasp detection methods focus on optimizing grasp posture for successful robot grasping manipulation as maximum as possible, but ignore the comprehension of candidate object’s attributes. Hence, this paper proposes a novel robot intention grasp network (RIGNet) for robot perception of how to grasp and comprehend why grasping. The improved RPN module robustly predict the RoIs to locate candidates in a bunch of overlap and occlusion objects, and then the multi-task siamese module finely detect the object category corresponding to the grasp configuration. What is more important, the proposed technology does not lose efficiency in single-object grasping, and even outperforms the state-of-the-art algorithms in multi-object stacked grasping. The novel RIGNet schema is suitable for human intends a real robot to execute desired grasp tasks.}, archive = {J_TASE}, author = {Xungao Zhong and Tao Gong and Junzhi Yu and Chengxian Zhou and Xunyu Zhong and Qiang Liu}, doi = {10.1109/TASE.2024.3522625}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10354--10367}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RIGNet: Robot intention grasp for dense stacked targets with multi-task siamese schema through RoIs learning}, volume = {22}, year = {2025}, }
• (2025). Power system transient stability assessment based on spatio-temporal broad learning system. TASE, 22, 10343--10353. (www)

  @article{ , abstract = {The safety, reliability and rapidity of the transient stability assessment (TSA) method are important for improving the reliability and security of power system operation. TSA methods based on deep learning have good evaluation performance, but the increasing number of network layers leads to the model training time being continuously extended. Therefore, a TSA method based on broad learning system (BLS) is proposed to improve the efficiency of power system TSA. Meanwhile, in order to overcome the deficiency of BLS in feature extraction, a TSA model based on spatio-temporal broad learning system (STBLS) is established, which combines BLS with graph convolutional network (GCN) and temporal convolutional network (TCN). First, the grid topology map and electrical measurement data are used as model inputs, and the graph convolution module and the temporal convolution module are used to extract the spatial features and temporal features of the system, respectively. The extracted spatio-temporal features are then used as inputs to the BLS classification module, which is utilized for the rapid assessment of transient stability. Simulation results demonstrate the effectiveness of the proposed model and highlights its superiority in terms of training time. Note to Practitioners—This paper is motivated by the realiza-tion of a fast and accurate assessment of power system transient stability (TSA). The main goal is to shorten the training time of the model as much as possible while ensuring that the assessment model has a high accuracy. In this paper, the broad learning system (BLS) is introduced into the field of TSA, and a TSA model based on spatio-temporal broad learning system (STBLS) is proposed by combining the BLS with graph convolutional net-work (GCN) and temporal convolutional network (TCN), which greatly shortens the training time of the model. Experimental results show that this method has higher efficiency and practical application value. In future research, we will conduct more com-prehensive validation of such methods and further optimize the model to improve its stability and generalization ability.}, archive = {J_TASE}, author = {Haiquan Zhao and Ruixue Ni}, doi = {10.1109/TASE.2024.3522369}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10343--10353}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Power system transient stability assessment based on spatio-temporal broad learning system}, volume = {22}, year = {2025}, }
• (2025). Design of stealthy joint attacks against cyber-physical systems: A reachable set approach. TASE, 22, 10333--10342. (www)

  @article{ , abstract = {This article studies the joint design of stealthy actuator and sensor attacks against cyber-physical systems with the aim of keeping the system’s state in an unsafe region. The Kullback-Leibler divergence is adopted as the metric of the joint attacks’ stealthiness. The attacker’s objective is realized by making the system’s ellipsoidal invariant reachable set under stealthy joint attacks belong to the unsafe set. Firstly, the relationship between the actuator attack and the shape of the ellipsoid is analyzed and it can be characterized by a non-convex optimization problem. Parameters of the actuator attack are obtained by solving another convex optimization problem constructed through applying a linear transformation to the original problem. Then, the sensor attack is analytically solved from a non-convex optimization problem to move the center of the ellipsoid to the desired target and increase the controller’s cost. Finally, an example of the flotation industrial process is illustrated to demonstrate effectiveness of the attack. Note to Practitioners—This paper aims to study security of cyber-physical systems from the perspective of attackers, which can help defenders fully understand the behavior of attackers. Existing works have not investigated which kind of stealthy attacks can move the state to the unsafe region. In this article, novel stealthy joint attacks are proposed such that the state of the attacked system is kept in the unsafe region. In detail, the actuator attack is to reshape the system’s ellipsoidal invariant reachable set and the sensor attack is to move the center of the ellipsoid to the desired target. In practical applications, the attacker need to obtain the system’s parameters and eavesdrop the input and output data, then solve the actuator attack from a convex optimization problem and compute the sensor attack by analytically solving a non-convex optimization problem. The attacks’ effectiveness is verified through the flotation industrial process. In the future, we will further investigate the design of stealthy attack strategies for nonlinear systems.}, archive = {J_TASE}, author = {Qirui Zhang and Wei Dai and Kun Liu and Lanhao Wang and Chunyu Yang}, doi = {10.1109/TASE.2024.3521970}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10333--10342}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design of stealthy joint attacks against cyber-physical systems: A reachable set approach}, volume = {22}, year = {2025}, }
• (2025). Fractional-order optimal control and FIOV-MASAC reinforcement learning for combating malware spread in internet of vehicles. TASE, 22, 10313--10332. (www)

  @article{ , abstract = {Internet of Vehicles (IoV) is gradually becoming popular, but it also brings more opportunities for malware intrusion. The intrusion of malware into IoV will cause a series of security issues and increase the incidence of road accidents. Therefore, the suppressing measures to combat the spread of malware in IoV will be fundamental and urgent. To address this critical issue, this paper proposes a fractional-order IoV (FIOV) to investigate malware propagation patterns in Road Side Unit (RSU) and Vehicles. To accurately reflect the actual spread of malware, the traffic density, the channel fading and the actual connectivity are considered in mathematical model. Then, the model-based optimal treatment and quarantine control strategy is derived by optimal control theory. Additionally, a novel model-free FIOV multi-agent soft actor-critic (FIOV-MASAC) approach is first proposed to suppress the malware propagation in IoV. Simulation experiments demonstrate that the proposed FIOV-MASAC approach exhibits better learning ability compared to other reinforcement learning (RL) algorithms.Note to Practitioners—Frequent attacks by malware on IoV are recognized as being challenging to prevent, with these attacks posing threats to data security and potentially resulting in traffic accidents and vehicle malfunctions. In response, a novel mathematical model has been introduced within this study to better predict the propagation trends of malware in IoV, effectively managing its spread within the vehicular network systems. While RL methods have been extensively utilized in the domain of control systems, it is noted that current RL methods depend on rich experience pools, rendering them inapplicable to more complex systems without adaptation. To address this, an effective and pragmatic RL algorithm has been devised in this study. This algorithm, devoid of the requirement for complex model establishment, is capable of intelligently learning and adjusting to the sophisticated environment of IoV, thereby effectively countering the propagation of malware. It should be highlighted that the RL method proposed herein is applicable to the majority of epidemic systems, enabling the achievement of stable control while substantially minimizing control expenditures. The integration of this method is anticipated to augment the security and robustness of IoV in the face of malware attacks.}, archive = {J_TASE}, author = {Guiyun Liu and Hao Li and Lihao Xiong and Zhihao Tan and Zhongwei Liang and Xiaojing Zhong}, doi = {10.1109/TASE.2024.3521614}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10313--10332}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fractional-order optimal control and FIOV-MASAC reinforcement learning for combating malware spread in internet of vehicles}, volume = {22}, year = {2025}, }
• (2025). Extension of the observability rank condition to time-varying nonlinear autonomous systems. TASE, 22, 10303--10312. (www)

  @article{ , abstract = {The study presents an innovative approach for observability analysis in time-varying, nonlinear autonomous systems. Traditional criteria assert that state observability applies solely to nonlinear systems where the system input is not explicitly state-dependent. In contrast, this study presents a groundbreaking observability criterion that incorporates the impact of system input without increasing the dimension of the analysis. This new criterion is applied to the relative navigation system of a Mars spacecraft, resulting in an observability conclusion for monitoring a non-cooperative target using angles-only measurements. Note to Practitioners—Observability is a critical property of observation systems that guides designs and operations. This paper introduces a new method for conducting observability analysis in time-varying, nonlinear autonomous systems. Traditional criteria typically limit state observability to nonlinear systems where the system input is not explicitly state-dependent. In contrast, this study proposes an innovative observability criterion that incorporates the influence of system input without increasing the dimensionality of the analysis. This new criterion has been successfully applied to the relative navigation system of a Mars spacecraft, resulting in an observability assessment for monitoring a non-cooperative target using angles-only measurements. This work has the potential to advance the understanding and application of observability in complex autonomous systems, especially in scenarios involving dynamic changes and nonlinearities.}, archive = {J_TASE}, author = {Tianshu Dong and Dayi Wang and Fangzhou Fu}, doi = {10.1109/TASE.2024.3521610}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10303--10312}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Extension of the observability rank condition to time-varying nonlinear autonomous systems}, volume = {22}, year = {2025}, }
• (2025). Inferring cable-suspended end-effector oscillations from hydraulic actuators’ responses in diaphragm wall hydraulic grabs. TASE, 22, 10287--10302. (www)

  @article{ , abstract = {A Diaphragm Wall Hydraulic Grab (DWHG), used in civil engineering for bulk excavation, is featured with a cable-suspended end-effector (also named attachment tool). The end-effector begins to oscillate during DWHG operation. Essential insights into the DWHG’s performance or productivity in operation could be derived from end-effector oscillations. However, due to limited robustness, the end-effector oscillation curves from a DWHG in operation can not be tracked directly by motion sensors attached to the end-effector. This article presents and evaluates a novel approach for DWHGs to infer cable-suspended end-effector oscillations from their hydraulic actuators’ responses. Hydraulic actuators’ responses depict the changes of hydraulic parameters in hydraulic actuators due to end-effector oscillations. The main contribution of this article is an investigation of how far oscillations of a cable-suspended end-effector in a DWHG can be inferred from their hydraulic actuators’ responses. For this purpose, a workflow is conducted: firstly, end-effector oscillations and their hydraulic actuators’ responses are collected for a typical DWHG steering sequence. Secondly, the collected end-effector oscillation and hydraulic actuators’ response curves are analyzed and processed. Thirdly, a model to infer end-effector oscillations from hydraulic actuators’ responses is built. Fourthly, an evaluation shows that the inferred oscillation curves (model output) are similar to the true end-effector oscillations. Note to Practitioners—This article is motivated by the demand to determine the oscillation curves of the cable-suspended end-effector (also named attachment tool) at a Diaphragm Wall Hydraulic Grab (DWHG) used in civil engineering for bulk excavation. Existing approaches for construction machines commonly install motion sensors at the end-effector to track its oscillations. However, mounting any sensors at the end-effector in DWHGs is not practicable because the end-effector is exposed to mechanical shocks and moisture. Thus, mounted sensors would be damaged or demolished. This article suggests and investigates a novel approach for DWHGs to infer end-effector oscillations from pressure data in hydraulic actuators that actuate the end-effector. Therefore, the pressure curves at the hydraulic actuators are tracked while the end-effector oscillates and are utilized afterward to infer these oscillations. A practical guide for a DWHG is presented and evaluated to collect, analyze, process, and utilize pressure data from hydraulic actuators to infer end-effector oscillations.}, archive = {J_TASE}, author = {Marius Krüger and Birgit Vogel-Heuser and Daniel Waterman and Suhyun Cha and Dominik Hujo and Theresa Prinz and Daniel Pohl and Matthias Semel}, doi = {10.1109/TASE.2024.3521599}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10287--10302}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Inferring cable-suspended end-effector oscillations from hydraulic actuators’ responses in diaphragm wall hydraulic grabs}, volume = {22}, year = {2025}, }
• (2025). Data-driven iterative learning temperature control for rubber mixing processes. TASE, 22, 10274--10286. (www)

  @article{ , abstract = {Considering the four challenges of non-identical initial states, non-repetitive uncertainties, different batch lengths, and unavailable mathematical model of a rubber mixing process (RMP), this article proposes a data-driven iterative learning temperature control (DDILTC) for the RMP. Specifically, an iterative linear data model (iLDM) is developed to formulate the iterative dynamics of RMP and is further used as a one-step iterative linear predictive model to estimate the RMP’s temperature that is unavailable when the current batch length is shorter than the desired one. The unknown parameters of the iLDM are estimated iteratively by designing an iterative adaption law. Further, an iterative learning based observer is designed to estimate the non-repetitive uncertainties and non-identical initial states as an extended state. The proposed DDILTC is a data-driven method and the iLDM is only used to formulate the iterative relationship of the input-output between two batches instead of a mathematical model of the RMP with physical meanings. Simulation study verifies the results. Note to Practitioners—The mixing temperature of a rubber mixing process (RMP) is a critical variable, ensuring the desired plasticity and viscosity of the rubber compounds. Indeed, RMP is a typical batch process performing repetitively over the finite time interval. However, no ILC results about the RMP temperature control have been reported even though ILC can learn the control experience from the past batches to improve control performance. The main reason lies in that the practical environments of RMP make it impossible to satisfy the strictly repetitive conditions, i.e., the initial states, disturbances, and batch lengths are all iteration-varying. Furthermore, it is difficult to establish a mathematical model of the RMP due to its large production scale and complex dynamics along both time and iteration directions. Therefore, the main motivation of this paper is to study the iterative learning temperature control problem of RMP by considering the nonrepetitive uncertainties of initial states, disturbances, and batch lengths, bypassing the use of any model information. An iterative linear data model (iLDM) is established to equivalently reformulate the unavailable two-dimensional dynamic behavior of RMP and to facilitate the controller design and analysis. The gradient uncertainty of RMP is reformulated as the unknown parameters in the iLDM and can be iteratively estimated by designing an iterative adaptation algorithm. The non-repetitive initial states and disturbances can be estimated by designing an iterative observer. Moreover, the unavailable mixing temperatures at the unreachable operation points are estimated by using the iLDM as the iterative predictive model. To summarize, the proposed method is simple in computation and easy in implementation since only the I/O data is used, and thus it is of great practical significance.}, archive = {J_TASE}, author = {Ronghu Chi and Zhihao Zhou and Huimin Zhang and Na Lin and Biao Huang}, doi = {10.1109/TASE.2024.3521332}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10274--10286}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven iterative learning temperature control for rubber mixing processes}, volume = {22}, year = {2025}, }
• (2025). Robustness-guaranteed observer-based control strategy with modularity for cleantech EMLA-driven heavy-duty robotic manipulator. TASE, 22, 10248--10273. (www)

  @article{ , abstract = {This paper introduces an innovative observer-based modular control strategy in a class of $n_{a}$ -degree-of-freedom (DoF) fully electrified heavy-duty robotic manipulators (HDRMs) to 1) guarantee robustness in the presence of uncertainties and disturbances, 2) address the complexities arising from several interacting mechanisms, 3) ensure uniformly exponential stability, and 4) enhance overall control performance. To begin, the dynamic model of HDRM actuation systems, which exploits the synergy between cleantech electromechanical linear actuators (EMLAs) and permanent magnet synchronous motors (PMSMs), is investigated. In addition, the reference trajectories of each joint are computed based on direct collocation with B-spline curves to extract the key kinematic and dynamic quantities of HDRMs. To guarantee robust tracking of the computed trajectories by the actual motion states, a novel control methodology, called robust subsystem-based adaptive (RSBA) control, is enhanced through an adaptive state observer. The RSBA control addresses inaccuracies inherent in motion, including modeling errors, non-triangular uncertainties, and both torque and voltage disturbances, to which the EMLA-driven HDRM is susceptible. Furthermore, this approach is presented in a unified generic equation format for all subsystems to mitigate the complexities of the overall control system. By applying the RSBA architecture, the uniformly exponential stability of the EMLA-driven HDRM is proven based on the Lyapunov stability theory. The proposed RSBA control performance is validated through simulations and experiments of the scrutinized PMSM-powered EMLA-actuated mechanisms. Note to Practitioners—Following strict global regulations, such as the 2015 Paris Agreement, there has been significant attention paid to the electrification trend. In this regard, the advancement of zero-emission electromechanical linear actuator technology has played a substantial role in developing fully electrified HDRMs. However, these systems are highly nonlinear and complex, comprising several interacting components, such as electric motors, reduction gearboxes, screw mechanisms, and load-bearing structures. Each of these components is prone to adverse effects arising from inaccuracies in modeling equations, sensor readings, and torque or voltage disturbances. As a result, achieving high-performance control presents significant challenges for engineers and necessitates computationally intensive approaches in practice. This paper presents a subsystem-based approach, enhanced by a robust state observer, to 1) mitigate the impact of uncertainties and disturbances substantially, 2) alleviate the computational burden and complexity of the targeted system, 3) prove mathematical stability, and 4) offer highly accurate and fast tracking performance. The proposed approach employs the dynamic motion of the studied EMLA-actuated HDRM, decomposing it into distinct subsystems and introducing a unified generic equation control for all subsystems. This modularity feature paves the way for researchers to extend the proposed approach to address other intricate applications.}, archive = {J_TASE}, author = {Mehdi Heydari Shahna and Mohammad Bahari and Jouni Mattila}, doi = {10.1109/TASE.2024.3520638}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10248--10273}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robustness-guaranteed observer-based control strategy with modularity for cleantech EMLA-driven heavy-duty robotic manipulator}, volume = {22}, year = {2025}, }
• (2025). Generalized modeling of overactuated aerial manipulators: Theory and application. TASE, 22, 10232--10247. (www)

  @article{ , abstract = {This paper proposes a unified dynamic model for aerial robots which encompasses all known actuation principles including tilting propellers and centroid variation methods such as moving masses or robotic manipulators. Of course, one can envision a wide variety of vehicles, all with different combinations of actuation principles. Therefore, a generalized modeling methodology for such vehicles is developed and presented in this paper. The modeling approach is verified through a comparative analysis of MATLAB simulations and laboratory experiments with a tilting propeller aerial manipulator vehicle, named Toucan. Finally, in order to fully explore and exploit the capabilities of the designed aerial manipulator, contact-based experiments with force tracking are performed using the proposed adaptive impedance controller. To demonstrate the advantages of such a vehicle, the adaptive impedance control method is used to generate position and orientation commands and achieve end-effector force tracking on a flat surface while the aerial manipulator maintains a neutral attitude. This greatly increases the stability and safety of contact-based operations, as the vehicle does not require changes in attitude to meet the force requirements. Note to Practitioners—This paper was motivated by a surge of different configurations of aerial vehicles emerging. A modeling method for generalized aerial vehicles is presented to determine which vehicle configurations are best suited for a specific application. This approach considers vehicles with additional actuators such as tilting propellers and robotic manipulators. The effectiveness of the proposed method is validated on a custom-designed quadrotor with four moving masses emulating centroid variation of a generic robotic manipulator. This original design is augmented with tilting propellers, yielding a truly generalized experimental platform. The generalized modeling algorithm and the experimental platform are tested both in free flight and in a scenario that demonstrates applying the desired force to a flat surface. Overall, we showcase the proposed generalized modeling is a useful tool for the rapid prototyping and design of aerial manipulators, paving the way for more advanced and capable systems in the future. With this approach, practitioners can gain a better understanding of the potential and limitations of their aerial vehicles, and ultimately improve their performance and effectiveness in specific applications.}, archive = {J_TASE}, author = {Lovro Markovic and Marko Car and Matko Orsag and Stjepan Bogdan}, doi = {10.1109/TASE.2024.3520489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10232--10247}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Generalized modeling of overactuated aerial manipulators: Theory and application}, volume = {22}, year = {2025}, }
• (2025). A test case generation scheduling model based on running event interval characteristics. TASE, 22, 10216--10231. (www)

  @article{ , abstract = {System testing mainly focuses on finding bugs before online deployment. However, the system’s complexity increases due to a surge in the number of users, which makes scheduling the test case generation time more difficult, impeding testing fidelity and automatic testing. In this paper, we propose a data-driven framework for scheduling the test cases and improving testing fidelity. The developed framework involves two stages: 1) mining the running event interval characteristics from logs obtained from the actual environment in the first stage, and 2) utilizing those characteristics to develop a model for the test case generation time schedule in the second stage. First, we select the running event interval, which is the interval between two adjacent running events processed by the system, as the feature for analysis. Then, we analyze the interval’s distribution characteristics, observing that it adheres to the power-law distribution from multiple aspects. Second, based on the distribution characteristics obtained, we develop a composite event chain with associated relationships (CECAR) model to schedule test case generation time. Finally, we conduct a series of experiments using logs obtained from eight regions. The results show that, compared to baseline methods, the CECAR model reduces average errors in the generated quantity, power exponent, and specific granularity intervals by approximately 1% to 9% across multiple layers, demonstrating the effectiveness of the proposed method. We also made our dataset and code publicly available at https://github.com/yizhenli-xjtu/CECAR. Note to Practitioners—System testing is a promising way to eliminate bugs and improve system stability before online deployment. If the testing method does not align with the system’s running characteristics in the actual environment, the testing fidelity and efficiency may be greatly reduced. Scheduling the test case generation time is one of the fundamental ways to address this challenge. The CECAR model developed in this paper can reproduce temporal characteristics and incorporate the concurrent relationships between different events, in turn, achieving the goal of fidelity improvement. Furthermore, based on the CECAR model, we can simulate arbitrary scenarios by combining different types of devices, which can greatly improve the testing’s flexibility and practicality.}, archive = {J_TASE}, author = {Yizhen Li and Tao Qin and Jinzi Zou and Huan Li and Chao-Bo Yan and Xiaohong Guan}, doi = {10.1109/TASE.2024.3520311}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10216--10231}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A test case generation scheduling model based on running event interval characteristics}, volume = {22}, year = {2025}, }
• (2025). Effective prediction of gait phase for assisted walking by means of gait-based adaptive oscillators. TASE, 22, 10205--10215. (www)

  @article{ , abstract = {How to optimally synchronize exoskeleton powered assistance remains a problem that limits the broad application of such devices. Kinematic change during frequent switching between go and stop, a common and representative activity of daily living (ADL), makes it challenging to predict the gait phase and deliver assistance due to unpredictable movements. Conventional adaptive oscillators (AO) have been verified to be effective in gait phase prediction in steady-state walking. When walking cadences are changed, it usually requires multiple walking strides to synchronize the assistance, causing inaccurate or even unwanted force disturbances that can be dangerous in some cases. To solve this problem, a gait-based AO is proposed in this paper. It has two-level AO systems. The high-level AO learns from last stride and updates the low-level AO, which is designed to estimate the gait phase of the current stride in real time. This approach significantly decreases the time required to learn walking kinematics, while simultaneously improving the accuracy of predictions. Experiments were conducted on seven participants, and the results showed that the proposed gait-based AO can predict the gait phase faster, more accurate, and more stable than the conventional one during non-steady-state walking. Note to Practitioners—This article developed a new gait phase prediction method called gait-based adaptive oscillator. It aims to provide fast and reliable gait phase prediction in situations characterized by frequent transitioning between stop and go (non-steady-state walking). This method has the potential to be an alternative to existing phase prediction methods for robotic exoskeleton assisted walking in daily life, as it requires no model training before use while can synchronize exoskeleton assistance within two strides.}, archive = {J_TASE}, author = {Xiangyang Wang and Yue Ma and Chunjie Chen and Zhuo Wang and Sheng Guo and Kin Huat Low and Xinyu Wu}, doi = {10.1109/TASE.2024.3520148}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10205--10215}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Effective prediction of gait phase for assisted walking by means of gait-based adaptive oscillators}, volume = {22}, year = {2025}, }
• (2025). Corrections to “Secure distributed adaptive control of nonlinear multi-agent systems”. TASE, 22, 10204. (www)

  @article{ , abstract = {This correction addresses a citation oversight in the paper [1]. While [1] has cited recent publication about the model-free adaptive control (MFAC) theory, it omits foundational references that are critical to understanding the core principles underlying MFAC, such as dynamic linearization, generalized Lipschitz condition, and the time-varying pseudopartial derivative. To rectify this, the correction supplements [1] with citations to pioneering contributions in MFAC theory. This ensures proper attribution and provides readers with a deeper and more comprehensive understanding of the theoretical foundations of MFAC.}, archive = {J_TASE}, author = {Yongxia Shi and Ehsan Nekouei}, doi = {10.1109/TASE.2024.3520022}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10204}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Corrections to “Secure distributed adaptive control of nonlinear multi-agent systems”}, volume = {22}, year = {2025}, }
• (2025). Adaptive iterative learning control with termination condition for MASs performing multiple tasks. TASE, 22, 10193--10203. (www)

  @article{ , abstract = {This paper investigates an adaptive iterative learning control (AILC) method for multiagent systems (MASs) performing multiple tasks. Different from traditional results for the single task, a multiple tasks case is considered in this work, which can complete various cooperative control. It should be pointed out that only one of the multiple tasks is performed in each iteration. For multiple tasks, a neural network (NN) is employed to create a mapping relationship between the input and output of nonlinearity, which is integrated into AILC to improve the control input. Additionally, an auxiliary signal is developed to compensate for the residual error caused by NN approximation and differentiator estimation. Then, a termination condition including mean square error and desired performance is established for the AILC method. By utilizing Lyapunov stability theory, it is proven that the tracking error converges to zero without termination condition and satisfies the desired accuracy with termination condition. In the simulation, multiple single-link manipulators are used to perform three different cooperative control tasks to validate the effectiveness of the proposed approach. Note to Practitioners—Most of the existing AILC methods are investigated to perform a single task for MASs. However, practical applications of MASs, such as manipulators, unmanned vehicles, and aerospace, often require the ability to perform multiple tasks. One motivation is to effectively learn control experience from different tasks. To achieve this, an NN is designed to learn experiences, which are further incorporated into AILC method to perform multiple tasks. In addition, due to limitations in computing resources, controllers cannot be learned indefinitely in practical applications. To address this challenge, a termination condition is established to stop learning when the desired performance is met. Based on the above considerations, the proposed AILC method with termination condition can reduce computational resources and flexibly perform multiple tasks in practical applications.}, archive = {J_TASE}, author = {Zhiqiang Li and Hui Ma and Hongru Ren and Hongyi Li}, doi = {10.1109/TASE.2024.3519636}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10193--10203}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive iterative learning control with termination condition for MASs performing multiple tasks}, volume = {22}, year = {2025}, }
• (2025). A restricted-learning network with observation credibility inference for few-shot degradation modeling. TASE, 22, 10177--10192. (www)

  @article{ , abstract = {Multiple sensors are widely used in the monitoring of the degradation process and prediction of the remaining useful lifetime (RUL) of units in complex engineering systems. However, ensuring the prognostic performance with only a few units available remains difficult. Under a few-shot scenario, the discordant observations that exist in sensor data introduce considerable uncertainty into the degradation model, which leads to an empirical loss far from the expected loss. On the other hand, the learned degradation model tends to be overfitted on the limited available units and results in a biased model parameter distribution, which limits the model generalization capability on unseen units. To address these issues, this paper proposes a restricted-learning network with observation credibility inference (OCI) for few-shot degradation modeling. We initially introduce the OCI to figure out discordant observations from sensor data. Then, OCI is incorporated into restrictive learning through the deletion of discordant observations from sensor data, which enforces a prior distribution constraint on degradation model parameters to prevent overfitting. Finally, a posterior augmented classifier is learned to estimate health status based on the posterior sensor paths, and the RUL can be predicted subsequently. A case study that uses the degradation dataset of aircraft engines demonstrates the superiority of the proposed method over benchmark methods under few-shot scenarios.Note to Practitioners—This paper aims to develop a few-shot degradation modeling method for conducting status monitoring and RUL prediction. Specifically, the developed method addresses two challenging issues in practice: 1) How to figure out discordant observations exist in sensor data; 2) How to prevent overfitting issues under few-shot scenarios. To implement this method, four steps are included as follows: First, collect multiple sensor data and failure time of historical units. Second, construct the degradation model network, and train the network with restricted parameter distribution after deleting discordant observations. Third, construct and learn the classifier for predicting the probability of failure. Fourth, estimate the degradation status of in-service units, and predict the RUL based on the classifier. The proposed method is expected to be able to characterize various degradation processes and be applied to the degradation modeling of engineering systems with limited data available.}, archive = {J_TASE}, author = {Ying Wang and Fangyu Li and Di Wang and Wei Qin}, doi = {10.1109/TASE.2024.3519317}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10177--10192}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A restricted-learning network with observation credibility inference for few-shot degradation modeling}, volume = {22}, year = {2025}, }
• (2025). Balanced safety-critical high-gain control for uncertain nonlinear systems with input saturation. TASE, 22, 10162--10176. (www)

  @article{ , abstract = {In this paper, we propose the dynamic high-gain scaling technique and solutions to input saturation for uncertain strict-feedback nonlinear systems. Although high gain affords fast response and high accuracy for the improvement of tracking performance, there are two inescapable potential risks: (i) excessive high gain would amplify the negative effects from non-vanishing mismatched uncertainties; (ii) high gain may conflict with the limited regulation capacity. Herein, two strategies based on invariance property are applied to high-gain control, aiming to address these two risks separately. On the one hand, by defining the function of performance robustness evaluation (PRE), the scaling gain grows to speed up the convergence rate and then maintains at an acceptable high level while guaranteeing the robustness to uncertainties in an invariant set. On the other hand, for handling the input saturation, the control barrier function (CBF)-based quadratic program (QP) describes the function of performance safety evaluation (PSE) that helps assess the system safety and decide whether the compensation for saturation is necessary, as such, the balance between performance and saturation gets achieved. Numerical simulations and a semi-physical experiment are performed to investigate the performance of our proposed methodology. Note to Practitioners—The motivation of this article is that stringent time response constraints are necessary for security or to increase productivity in practical applications, while control constraints exist widely in control systems. The lack of constraint satisfaction may inevitably result in safety defects, performance degradation. Based on these observations, a balanced safety-critical high-gain control scheme is proposed for input-constrained nonlinear uncertain strict-feedback systems. The contribution focuses on finding a balanced relationship between rapid response ability, robustness to mismatched uncertainties, and finite regulation capability in the high-gain control framework, since excessive control gain would potentially amplify the effect of non-vanishing mismatched uncertainty or lead to unexpected control saturation. Moreover, the theoretical derivation demonstrates that: only the system has the tolerance of matched uncertainties when the high gain grows to infinity, otherwise there must exit a maximum value for high gain to maintain the robustness and stability; the PSE function assesses the system safety under saturation conditions and decides whether to continue improving control performance or compensate for saturation for the maintenance of stability.}, archive = {J_TASE}, author = {Peng Wang and Xiuhui Peng and Xiaoling Liang and Shuzhi Sam Ge}, doi = {10.1109/TASE.2024.3519188}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10162--10176}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Balanced safety-critical high-gain control for uncertain nonlinear systems with input saturation}, volume = {22}, year = {2025}, }
• (2025). A novel incremental defect detection method via elastic heterogeneous distillation network. TASE, 22, 10149--10161. (www)

  @article{ , abstract = {In industrial processes, the defect data is continuously accumulated over time, and new classes of defects can arise at any time. When a well-trained model is adapted to new classes, its performance in old classes will sharply decline. To solve the above issue, a novel incremental defect detection method named elastic heterogeneous distillation network (EHD-Net) is proposed. First, an elastic knowledge transfer method is proposed to selectively transfer the core features of the old classes, so that the old knowledge is retained and new knowledge is effectively learned at the feature level. Second, warping heterogeneous distillation learning is proposed for the first time. In warping heterogeneous distillation learning, a gradient warping layer is proposed to balance the optimization direction of gradients for both old and new classes, and the proposed heterogeneous distillation learning strategy can clarify the association and difference between new and old classes, so that the rapid learning of new knowledge and comprehensive transfer of old knowledge are guaranteed at the decision level. Finally, a well-trained evolutionary model is employed to achieve the detection of both old and new classes. The proposed method can effectively overcome the catastrophic forgetting of old classes and guarantee the independent learning of new classes. Experimental results on two industrial datasets show that EHD-Net outperforms existing advanced methods. Note to Practitioners—The motivation for this paper is an important and highly practical issue called class-incremental defect detection. Specifically, new classes of defects always appear in industrial scenarios over time and when a well-trained detection model adapts to the new classes, its performance in the old classes decreases dramatically. To solve the above issues, an elastic heterogeneous distillation network (EHD-Net) via elastic knowledge transfer and warping heterogeneous distillation learning is proposed. The proposed method not only effectively suppresses the catastrophic forgetting of old classes while realizing the fast learning of new classes, so that the model can simultaneously achieve the high-accuracy detection of both the new and old classes without the participation of old class samples. The validity of the proposed method is verified under two industrial datasets, which fully guarantees its practical application value.}, archive = {J_TASE}, author = {Xiangkai Shen and Jinhai Liu and Huaguang Zhang and Lin Jiang and He Zhao and Hongyu Yang}, doi = {10.1109/TASE.2024.3519164}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10149--10161}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel incremental defect detection method via elastic heterogeneous distillation network}, volume = {22}, year = {2025}, }
• (2025). Socially acceptable bipedal robot navigation via social zonotope network model predictive control. TASE, 22, 10130--10148. (www)

  @article{ , abstract = {This study addresses the challenge of social bipedal navigation in a dynamic, human-crowded environment, a research area largely underexplored in legged robot navigation. We present a zonotope-based framework that couples prediction and motion planning for a bipedal ego-agent to account for bidirectional influence with the surrounding pedestrians. This framework incorporates a Social Zonotope Network (SZN), a neural network that predicts future pedestrian reachable sets and plans future socially acceptable reachable set for the ego-agent. SZN generates the reachable sets as zonotopes for efficient reachability-based planning, collision checking, and online uncertainty parameterization. Locomotion-specific losses are added to the SZN training process to adhere to the dynamic limits of the bipedal robot that are not explicitly present in the human crowds data set. These loss functions enable the SZN to generate locomotion paths that are more dynamically feasible for improved tracking. SZN is integrated with a Model Predictive Controller (SZN-MPC) for footstep planning for our bipedal robot Digit. SZN-MPC solves for collision-free trajectory by optimizing through SZN’s gradients. Our results demonstrate the framework’s effectiveness in producing a socially acceptable path, with consistent locomotion velocity, and optimality. The SZN-MPC framework is validated with extensive simulations and hardware experiments. Note to Practitioners—This paper is motivated by the challenge of navigating bipedal robots through dynamic, human-crowded environments in a socially acceptable manner. Existing methods for social navigation often only address obstacle avoidance and are demonstrated on a robot with simple dynamics. This paper proposes the Social Zonotope Network (SZN), a novel neural network that couples pedestrian future trajectory prediction and robot motion planning to facilitate socially aware navigation for bipedal robots such as Digit, designed by Agility Robotics. The social behaviors are learned from real open-sourced pedestrian data using the SZN, which outputs the future predictions as reachable sets for each agent in the environment. The SZN is then integrated into a trajectory optimization problem that takes into account personal space preferences and bipedal robot capabilities to design trajectories that are both collision-free and socially acceptable. This work also highlights the computational efficiency of the SZN design that makes it suitable for real-time integration with motion planners. The framework is validated through extensive simulations and hardware experiments. From a practical standpoint, this research provides a framework that can be applied to bipedal robots to improve automation in human-populated environments such as hospitals, shopping centers, and airports. The framework’s ability to automatically adapt to surrounding human movement helps minimize disruptions and ensures that the robot’s presence is not a hindrance to the flow of human traffic. Future work will focus on outdoor deployment, which will require onboard perception capabilities to detect surrounding pedestrians.}, archive = {J_TASE}, author = {Abdulaziz Shamsah and Krishanu Agarwal and Nigam Katta and Abirath Raju and Shreyas Kousik and Ye Zhao}, doi = {10.1109/TASE.2024.3519012}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10130--10148}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Socially acceptable bipedal robot navigation via social zonotope network model predictive control}, volume = {22}, year = {2025}, }
• (2025). MISOCP model for reactive power optimization with nonuniform voltage regulators in unbalanced three-phase active distribution networks. TASE, 22, 10118--10129. (www)

  @article{ , abstract = {Reactive power optimization (RPO) is a key task in the operation and control of active distribution networks (ADNs). The nonlinear power flow constraints and the integers introduced by voltage regulator (VR) constraints make the nonconvex RPO model difficult to solve. In this paper, a RPO model is proposed considering the nonuniform tap ratios of VRs. The three phases in VRs are independently controlled and the phase coupling effects are strictly considered. Moreover, to deal with the bilinear relationship among continuous complex voltage phasors and discrete tap ratios, the voltage phasor matrix is decoupled into real and imaginary parts, and a status variable method is proposed to exactly linearize the bilinear terms. Further, the nonlinear power flow constraints are relaxed to a set of second-order cone constraints without neglecting the coupling effect among the three phases and are proved to be equivalent to the semi-definite constraints, which does not require the rank-1 verification and is more efficient and scalable. Thus, the original nonconvex RPO for three-phase unbalanced ADNs is simplified to a mixed-integer second-order cone programming (MISOCP) model. Case studies validate the model’s effectiveness, and the computational efficiency meets practical requirements. Note to Practitioners—This paper simplifies the complex problem of reactive power optimization for active distribution networks with unbalanced three-phase power flows. We offer a novel RPO model improving the typically nonconvex and nonlinear optimal power flow problem, and addressing the particular challenges imposed by voltage regulators optimization with nonuniform tap ratios. Independent control of the VR phases and accurate incorporation of phase coupling effects have been realized, greatly simplifying the optimization process. The transformation of nonlinear power flow constraints into a computationally efficient set of second-order cone constraints eliminates the need for cumbersome rank-1 verification. Our mixed-integer second-order cone programming model integrates smoothly with standard optimization procedures, improving voltage regulation and network stability. The effectiveness and efficiency of the approach have been validated in case studies, offering a viable tool for industry application.}, archive = {J_TASE}, author = {Chenggang Mu and Tao Ding and Shunqi Wang and Wenhao Jia and Mohammad Shahidehpour and Zhao Luo}, doi = {10.1109/TASE.2024.3518590}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10118--10129}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MISOCP model for reactive power optimization with nonuniform voltage regulators in unbalanced three-phase active distribution networks}, volume = {22}, year = {2025}, }
• (2025). Enhanced automated cell micromanipulation via programmable magnetic microgripper design. TASE, 22, 10107--10117. (www)

  @article{ , abstract = {Achieving efficient and robust grasping manipulation in microscopic scenarios is crucial for robotic applications. As a typical end-effector, the microgripper that can perform gripping-holding-releasing operations, plays an essential role in the area of micromanipulation. To facilitate the development of industrial microassembly and cell micromanipulation technologies, we propose a robotic microgripper (RM) with a two-finger structure based on magnetic programmable soft materials. First, we synthesize and characterize the magnetic programmable soft material, and fabricate the microfinger based on the soft material. The analytical model of the microgripper was developed using beam deformation theory and magnetic field modeling. By combining the deformation mechanism of flexible microfingers with visual feedback, we modeled the gripping force of the microgripper and proposed a feedback adaptive grasping strategy (FAGS). The cooperative control of two flexible microfingers via remote magnetic field actuation achieves the grasping of microobjects with various complex shapes and in different media. A micromanipulation robotic system was constructed by combining RM with a micromanipulation arm, enabling targeted grasping, transportation, and posture control of microobjects in complex scenarios. Note to Practitioners—The microgripper plays a pivotal role in advancing automated robotic cell micromanipulation and microassembly. This article is motivated to develop a microgripper based on soft programmable magnetic materials. The proposed microgripper features a straightforward structural design, which simplifies the manufacturing process. The actuation is implemented by visual feedback and controlled variations in magnetic field intensity. Compared to existing rigid microgrippers, soft microgrippers provide higher stability of contact with the object being manipulated. For cell manipulation, a robotic dexterous micromanipulation system is devised by integrating the RM onto a multi-degree-of-freedom micromanipulator arm, enabling precise operations such as grasping, releasing, transporting, and orientation control of zebrafish embryonic cells. In microassembly tasks, the microgripper demonstrates the ability to handle micro-parts of diverse shapes. The proposed microgripper holds substantial promise in advancing the automation of robotic cell micromanipulation and microassembly processes.}, archive = {J_TASE}, author = {Youchao Zhang and Fanghao Wang and Yuqian Ye and Xiangyu Guo and Xiao Wang and Alois Knoll and Yixian Wang and Changsheng Dai and Yibin Ying and Mingchuan Zhou}, doi = {10.1109/TASE.2024.3517671}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10107--10117}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhanced automated cell micromanipulation via programmable magnetic microgripper design}, volume = {22}, year = {2025}, }
• (2025). SmallNet: A small defects detection network for magnetic chips based on context-weighted aggregation and feature multiscale loop fusion. TASE, 22, 10095--10106. (www)

  @article{ , abstract = {Accurate detection of surface defects on magnetic chips is a necessary and difficult task, especially for small defects, which lack discriminative and robust features due to their small size and weak characteristics. At present, the low accuracy of small defect detection seriously restricts the development of automated visual inspection and needs to be solved urgently. Thus, we propose a novel small defects detection network based on context-weighted aggregation and feature multiscale loop fusion. First, a context-weighted aggregation module (CAM) that enriches feature representations by combining context and attention mechanisms is proposed. We believe that any further improvements will be futile if robust feature representations cannot be obtained. Second, inspired by the mechanism of looking and thinking twice, we aggregate left-right feedback connections into feature pyramids and creatively propose a loop-shaped feature pyramid network (Loop-FPN), enabling multiscale features to be fused up and down and connected left and right. This loop-shaped structure makes the connection of each layer more direct and allows the features at each scale to be fully integrated, which improves the utilization of multiscale features and facilitates the detection of small defects. Finally, we apply the proposed network to practical detection and the results show that our network achieves 97.57% precision, 91.91% recall, and 98.39% AP, which are 0.13%, 1.43%, and 0.99% higher than the current best-performing comparative methods, respectively. Note to Practitioners—Current vision inspection technology has low accuracy and poor stability in small defect detection, which cannot meet the industrial inspection requirements. Small defect detection has become the biggest challenge in the field of visual inspection and has seriously restricted its development. Our proposed network can solve this problem well by mining small defect context, fusing multiscale features and utilizing attention mechanisms, and has been successfully applied in magnetic chip production line. Furthermore, we developed an image acquisition system that can capture defects on all surfaces with high accuracy and without dead space, allowing our network to not only detect small defects on magnetic chips of varying sizes and irregular shapes, but also to adapt to changes in lighting conditions, backgrounds, and viewpoints. Our work provides an effective, reliable, and convenient quality control solution for magnetic chip production.}, archive = {J_TASE}, author = {Weijian Liang and Yaoru Sun and Siyu Zhang and Lizhi Bai and Jun Yang}, doi = {10.1109/TASE.2024.3517537}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10095--10106}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SmallNet: A small defects detection network for magnetic chips based on context-weighted aggregation and feature multiscale loop fusion}, volume = {22}, year = {2025}, }
• (2025). Neural network-based nonlinear stabilizing control for 3-D offshore crane with double-pendulum effect. TASE, 22, 10084--10094. (www)

  @article{ , abstract = {Wave-induced ship motions pose great challenges to the design of control systems for offshore cranes, especially with double-pendulum effect and unknown system dynamics. In this paper, a neural network-based nonlinear stabilizing controller is proposed for 3D offshore cranes to accomplish boom positioning and payload swing-elimination under wave-induced ship roll and pitch motions. Critical and practical-oriented issues including double-pendulum effect, boom position limitations, actuator input dead zones, and unknown system dynamics are considered simultaneously. First, the dynamic model of 3D offshore cranes with double-pendulum effect is established by using the Lagrange’s modeling method. By combining the state variables with perturbation terms, new auxiliary variables are introduced for model transformation to simplify model analysis and controller design. Then, by analyzing the transformed model, the neural network is rationally designed to estimate the unknown system dynamics and input dead zones. In addition, barrier Lyapunov functions (BLFs) are employed to the controller ensuring that the boom operates within a safe range. The Lyapunov-based theory is utilized to rigorously prove the stability and convergence of the designed control system. Hardware experiments are elaborately designed to verify the performance of the proposed method in terms of effectiveness, robustness, and anti-disturbance. Note to Practitioners—This paper studies the stabilizing control problem of offshore crane systems. In practice, the double-pendulum effect between the payload and hook is evident, and it is necessary to simultaneously lift and rotate the boom to position and stabilize the payload smoothly and accurately under wave-induced ship motions. However, existing studies typically oversimplify the model of offshore cranes, neglecting crucial factors such as their 3D characteristics, the double-pendulum effect, and the intricate wave-induced ship motions. This limits the application of the corresponding control methods in practice. Moreover, most of the existing control methods ignore unknown model dynamics, input dead zones, etc., which is not favorable for practical applications. To address the above problems, this paper proposes a neural network-based nonlinear stabilizing controller, which can accomplish the stabilization control for 3D offshore cranes with double pendulum effect in the presence of unknown dynamics and input dead zones. The effectiveness of the proposed method is validated on the self-built hardware platform.}, archive = {J_TASE}, author = {Ling Yang and Gang Li and Xin Ma}, doi = {10.1109/TASE.2024.3516873}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10084--10094}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Neural network-based nonlinear stabilizing control for 3-D offshore crane with double-pendulum effect}, volume = {22}, year = {2025}, }
• (2025). PCA-based sensor drift fault detection with distribution adaptation in wastewater treatment process. TASE, 22, 10071--10083. (www)

  @article{ , abstract = {Accurate detection of sensor drift fault in wastewater treatment process (WWTP) is essential for maintaining normal system operation and making correct decisions. However, since the WWTP is influenced by numerous internal and external factors, the data acquired from the actual WWTP is always multi-distributed, thus bringing difficulties to the accurate detection of sensor drift fault for the slow gradual change. To address this problem, a PCA-based sensor drift fault detection method with distribution adaptation (DAPCA) is proposed in this study. It presents a novel PCA-based fault detection method including a temporal WaveCluster for adaptive clustering for multi-distributed data, and a robust PCA-based fault detection with a smoothing mechanism using a combined index. Firstly, an improved WaveCluster algorithm is designed to cluster the multi-distributed data adaptively by considering both the spatial and temporal characteristics. Secondly, a robust PCA algorithm is presented that incorporates a smoothing mechanism to increase its robustness to noise interference. Thirdly, to strike a balance between traditional statistical indexes, a combined index is introduced with adaptive thresholds for multi-distributed data, thus enhancing the overall detection accuracy. To assess the performance of DAPCA, it is tested on both benchmark and real datasets. The results show that it attains the superior detection accuracy with higher F1-scores and lower false alarm rates than comparative methods. Furthermore, DAPCA is demonstrated to be more robust to various types of noises, significantly reducing the false alarms caused by the noise. Note to Practitioners—In the context of wastewater treatment process (WWTP), the inherent exposure of sensors to harsh environmental conditions renders them prone to drift fault. Furthermore, the complex operational dynamics of WWTP contribute to the emergence of a multi-distribution of data, thereby exacerbating the challenges associated with accurate detection of drift fault. Motivated by this, the present paper proposes a PCA-based sensor drift fault detection method with distribution adaptation (DAPCA) in WWTP, which prevents the degradation of detection accuracy caused by changes in data distribution. It presents a novel PCA-based fault detection method including a temporal WaveCluster for adaptive clustering for multi-distributed data, and a robust PCA-based fault detection with a smoothing mechanism using a combined index. Consequently, the effectiveness of the proposed DAPCA is validated via comparisons to other models, which performs a superior detection accuracy with higher F1-scores and lower false alarm rates. Furthermore, DAPCA is demonstrated to be more robust to many types of noises, significantly reducing the false alarms caused by the noise. In conclusion, for multi-distributed data, DAPCA is able to accurately detect sensor drift fault in WWTP, and can be further extended for sensor drift fault detection in other industrial processes.}, archive = {J_TASE}, author = {Junfei Qiao and Jianing Zhang and Wenjing Li}, doi = {10.1109/TASE.2024.3516710}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10071--10083}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PCA-based sensor drift fault detection with distribution adaptation in wastewater treatment process}, volume = {22}, year = {2025}, }
• (2025). A new structural relation extraction framework for SAR occluded target recognition. TASE, 22, 10056--10070. (www)

  @article{ , abstract = {Partially occluded target recognition is a pressing issue in synthetic aperture radar (SAR) target recognition. Occlusion causes the loss of crucial information, like target structure details. This paper proposes a new structural relation extraction framework to address partial occlusion. It is achieved through the tailored design of counterfactual samples synthesizing and jigsaw mutual learning (CSS-JML). The ASC model parameters have clear physical meanings, aiding in understanding local structural changes. By integrating SAR and ASC images, effective structural relationship representations are extracted, mitigating occlusion effects. The CSS module is designed to generate occluded counterfactual SAR and ASC images using pairs of target data. There is no longer a requirement for further annotation information because this new generation process is limited by recognition tasks. The JML module employs mutual learning to complete jigsaw puzzle tasks in both modalities. And in this process, we design two types of similarity constraints to facilitate the extraction of unified structural information across different modalities. The FA module interacts with recognition features, facilitating the classification and identification of partially obscured targets. Experimental results on MSTAR-based and FUSARShip-based test datasets with three occlusion patterns demonstrate the method’s superiority in most occluded conditions, confirming its effectiveness in SAR occluded target recognition. Note to Practitioners—The motivation for this paper stems from the need to design a robust and efficient SAR target recognition method under partial occlusion. The structural relationships within samples can provide valuable information for recognizing occluded targets. We designed a counterfactual generation module to generate occluded target samples using paired targets in an unsupervised manner. This module is jointly optimized with subsequent recognition tasks, without the need for additional information. Recognizing the interaction bottleneck in mutual learning tasks, we designed two similarity constraints to promote the extraction of unified representations. Finally, we designed a feature affinity module to introduce structural representation information into the recognition branch. The proposed method achieves excellent classification performance in various occlusion environments.}, archive = {J_TASE}, author = {Jiaxiang Liu and Zhunga Liu and Longfei Wang and Zuowei Zhang}, doi = {10.1109/TASE.2024.3516391}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10056--10070}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A new structural relation extraction framework for SAR occluded target recognition}, volume = {22}, year = {2025}, }
• (2025). TactCLNet: Tactile continual learning network based on generative replay for object hardness recognition. TASE, 22, 10045--10055. (www)

  @article{ , abstract = {Currently, deep neural networks can be extremely effective in robotic tactile perception. However, a major challenge is to solve the problem of continual learning of robotic tactile perception in an open and dynamic environment. In this paper, we propose a novel continual learning method for the domian incremental learning task in the field of tactile perception. To be specific, we introduce a morphology-specific variational autoencoders which can mitigate catastrophic forgetting by generating pseudo-samples for training in the continual learning process. We integrate the generative model and the discriminative model into one model, which reduces the size of model and improves the continual learning ability. In addition, considering the ordinal information between the hardness levels, we propose to add conditional information to the model and introduce a modified loss function to combine the latent value with the hardness information, which improves the continual learning performance by controlling the distribution and quality of pseudo-sample generation. Following this, we designed a tactile robot experiment, collected hardness data, and tested our model on this object hardness recognition task. We show experimentally that, after training, the model can still maintain the accuracy of more than 94% after learning three tasks in terms. Note to Practitioners—In the field of robotics tactile perception, the issue of continual learning in robots is a crucial problem that urgently requires resolution. We hope robots to effectively engage in continual learning across multiple tasks, ensuring the acquisition of new knowledge while mitigating the risk of forgetting previously acquired knowledge. In this paper, we propose a novel continual learning method for the domian incremental learning task. we introduce a morphology-specific variational autoencoders based on replaying pseudo-samples during continual learning process which reduces the size of model and improves the continual learning ability. We enhance model performance by integrating generative and discriminative models, incorporating conditional information to control the distribution of replayed sample types, and leveraging sequential relationships among samples. It is proved that the proposed method is able to effectively improve the accuracy in a tactile domian incremental learning task.}, archive = {J_TASE}, author = {Yiwen Liu and Zhengkun Yi and Senlin Fang and Yupo Zhang and Feng Wan and Zhi-Xin Yang and Xu Lu and Xinyu Wu}, doi = {10.1109/TASE.2024.3516378}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10045--10055}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TactCLNet: Tactile continual learning network based on generative replay for object hardness recognition}, volume = {22}, year = {2025}, }
• (2025). Fleet service reliability analysis of self-service systems subject to failure-induced demand switching and a two-dimensional inspection and maintenance policy. TASE, 22, 10029--10044. (www)

  @article{ , abstract = {A fleet of self-service systems, such as electric vehicle charging piles (EVCPs), is usually installed in a specific location. During operation, these systems are subject to random failures. However, they are usually operated without on-site staff. It is quite common that a customer may switch to other unoccupied systems for service when the initially selected system is found to have failed or fails during service. This is called failure-induced demand switching (FDS). With continuous customer arrivals and system failures, such FDS events occur repeatedly and interact dynamically, making modeling and enhancing service levels quite difficult. The challenge becomes even greater when a unique two-dimensional inspection and maintenance (IM) policy is adopted to handle the maintenance needs of self-service systems in hopes of retaining their service level with respect to long-run demand satisfaction. In this paper, we investigate the long-term service reliability of a fleet of self-service systems subject to FDS and a two-dimensional IM policy. First, we model the fleet state transition process and characterize its analytical properties. Next, we measure the fleet’s long-term service reliability and obtain the analytical expressions for crucial service level metrics, such as the expected number of failed systems, the expected length of an operation cycle, and service reliability loss due to imperfect monitoring. The managerial implications regarding the selections of EVCPs and IM policy are proposed based on a numerical study of two fleets of EVCPs in Hong Kong. These implications are expected to assist the operators in ensuring fleet service levels in the long run at a minimal operation and maintenance cost. Note to Practitioners—This paper models the service reliability of a fleet of self-service systems (e.g., EVCPs) over time. Service reliability, reflecting the fleet’s service level, is defined as the proportion of demands being fulfilled and is of the utmost concern of system operators. However, under repeated FDS due to continuous customer arrivals and system failures, it is difficult to assess the fleet’s service reliability using existing methodologies. The task becomes more challenging when failures of such systems are not perfectly detected in practice. This paper develops mathematical models and a novel two-dimensional inspection and maintenance policy to overcome the technical barriers. The models enable assessing the service reliability of various self-service systems experiencing non-constant service rates. Two case studies of fleets of AC and DC EVCPs in Hong Kong are provided to demonstrate the practical applicability of the proposed models. Indeed, the core findings of this work assist practitioners in: (i) finding the optimal maintenance policy that maximizes the fleet’s service reliability, (ii) assessing the service reliability loss due to imperfect failure detection, and (iii) evaluating the sensitivity of the optimal service reliability with respect to the customer arrival rate, system failure rate, and customer behavior in reporting system failures. We also show that the proposed models provide the exact solutions to the above-mentioned metrics when the maintenance duration is longer than the service duration.}, archive = {J_TASE}, author = {Yian Wei and Yao Cheng and Haitao Liao}, doi = {10.1109/TASE.2024.3516049}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10029--10044}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fleet service reliability analysis of self-service systems subject to failure-induced demand switching and a two-dimensional inspection and maintenance policy}, volume = {22}, year = {2025}, }
• (2025). Knowledge-informed machine learning for cancer diagnosis and prognosis: A review. TASE, 22, 10008--10028. (www)

  @article{ , abstract = {Cancer remains one of the most challenging diseases to treat in the medical field. Machine learning (ML) has enabled in-depth analysis of complex patterns from large, diverse datasets, greatly facilitating “healthcare automation” in cancer diagnosis and prognosis. Despite these advancements, ML models face challenges stemming from limited labeled sample sizes, the intricate interplay of high-dimensionality data types, the inherent heterogeneity observed among patients and within tumors, and concerns about interpretability and consistency with existing biomedical knowledge. One approach to address these challenges is to integrate biomedical knowledge into data-driven models, which has proven potential to improve the accuracy, robustness, and interpretability of model results. Here, we review the state-of-the-art ML studies that leverage the fusion of biomedical knowledge and data, termed knowledge-informed machine learning (KIML), to advance cancer diagnosis and prognosis. We provide an overview of diverse forms of knowledge representation and current strategies of knowledge integration into machine learning pipelines with concrete examples. We conclude the review article by discussing future directions aimed at leveraging KIML to advance cancer research and healthcare automation. A live summary of the review is hosted at https://lingchm.github.io/kinformed-machine-learning-cancer/ offering an evolving resource to support research in this field.Note to Practitioners—Recognizing the challenges posed by inter-patient and intratumoral heterogeneity, constrained sample size, and interpretability requirements in cancer applications, practitioners should consider integration of existing biomedical knowledge into their modeling frameworks. This strategy holds promise for enhancing model performance, robustness, and interpretability. We review classic machine learning and deep learning models that incorporated domain knowledge in their cancer diagnosis and prognosis models spanning models that use clinical, imaging, molecular, and treatment data. Pros and cons of each integration approach are discussed. Key design questions such as which knowledge to leverage, how to represent it effectively, and how to seamlessly integrate it into their models need be examined for each case. Collaboration between modeling scientists and medical experts is essential in this endeavor.}, archive = {J_TASE}, author = {Lingchao Mao and Hairong Wang and Leland S. Hu and Nhan L. Tran and Peter D. Canoll and Kristin R. Swanson and Jing Li}, doi = {10.1109/TASE.2024.3515839}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {10008--10028}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Knowledge-informed machine learning for cancer diagnosis and prognosis: A review}, volume = {22}, year = {2025}, }
• (2025). Partial identification of coupled matrices for stochastic hybrid multi-layer delayed networks. TASE, 22, 9994--10007. (www)

  @article{ , abstract = {This paper focuses on partial identification problems of coupled matrices for stochastic hybrid multi-layer delayed networks with inter-layer and intra-layer couplings and Markovian switching. The main approach is to combine Lyapunov method, Kirchhoff’s matrix-tree theorem and stochastic analysis. By pinning control technique, partial identification criterion is got. As a special case, the whole identification criterion is also derived. At last, three numerical examples of classical Lü and Lorenz systems are provided to demonstrate the validity of the presented results. Especially the impact of Markovian switching on partial topology identification is discussed. Note to Practitioners—This paper was motivated by existing results on synchronization of multi-layer complex network with stochastic perturbation. The existing results mainly focus on exponential synchronization or asymptotic synchronization based on known coupled matrices of networks. In this paper, the coupled matrices are unknown, the drive-response systems can achieve the partial synchronization with probability one via pinning controller. Furthermore, combining with stochastic version of the LaSalle theorem, the partial identification of coupled matrices criterion has been obtained. It should be noted that the Markovian switching and time delay in this work play an important role in the identification.}, archive = {J_TASE}, author = {Chunmei Zhang and Hui Yang and Jiamin Zhou and Hui Zhou}, doi = {10.1109/TASE.2024.3515213}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9994--10007}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Partial identification of coupled matrices for stochastic hybrid multi-layer delayed networks}, volume = {22}, year = {2025}, }
• (2025). Robust model predictive control for nonlinear systems with incremental control input constraints. TASE, 22, 9983--9993. (www)

  @article{ , abstract = {This paper presents a robust model predictive control (RMPC) algorithm for nonlinear discrete-time systems subject to bounded disturbances and incremental control input constraints. To guarantee recursive feasibility, a terminal inequality constraint is integrated into the proposed RMPC algorithm. By employing constraint tightening techniques, we derive an upper bound on admissible disturbances that ensures the input-to-state stability (ISS) for the closed-loop system. The effectiveness of the proposed algorithm is validated through numerical simulations and practical experiments involving the control of a four-wheel mobile robot. The results demonstrate the capability of the proposed method to maintain system stability and optimize control performance in the presence of external disturbances. Note to Practitioners—In practical engineering, the prevalence of external perturbations and the necessity for incremental control input constraints significantly complicate the control system design process. Compared with traditional control methodologies, model predictive control (MPC) is better equipped to address disturbances and constraints, achieving enhanced control accuracy and safety. This paper introduces an enhanced RMPC method specifically designed to control a broad class of nonlinear systems in the presence of disturbances and input constraints. Additionally, we provide insights into the relationship between specific design parameters of the RMPC algorithm and the upper bounds of permissible disturbances, offering practical guidelines for implementation. The proposed method is validated through simulations and practical experiments with a four-wheeled mobile robot. The results confirm that the approach reliably maintains system stability while efficiently optimizing control inputs. Future work will focus on extending the algorithm to potential robotic systems and exploring alternative disturbance-handling methods, such as observer-based and set-membership approaches.}, archive = {J_TASE}, author = {Fang-Jiao Zhao and Yong-Feng Gao and Xue-Fang Wang and Hao-Yuan Gu and Xi-Ming Sun}, doi = {10.1109/TASE.2024.3515172}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9983--9993}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust model predictive control for nonlinear systems with incremental control input constraints}, volume = {22}, year = {2025}, }
• (2025). Optimal placement of programmable tooling machines considering hierarchical structure via sparse learning for multistage assembly processes. TASE, 22, 9970--9982. (www)

  @article{ , abstract = {End-of-line product dimensional quality assurance is crucial in multistage assembly processes (MAPs). Active control strategies involve the deployment of controllable, programmable tooling machines (PTs) to adjust part positions for in-process dimensional error compensation. In MAPs, multiple PTs may be deployed within a single stage or across various stages. To enable the normal operation of deployed PTs within a specific stage, an associated supporting platform (SP) is necessary. Consequently, the lower PT-level and the upper platform-level define a two-level hierarchical structure in MAPs. This paper aims to propose an optimal placement strategy of PTs, focusing on the final dimensional quality of the product and the total cost related to the number of accommodated PTs and required SPs. Based on the stream-of-variation (SOV) model of MAPs, we develop a novel sparse learning framework along with the corresponding parameter estimation algorithm to achieve the optimal placement. The case study demonstrates the effectiveness of our proposed method for the optimal placement of PTs in reducing dimensional variation in MAPs. Note to Practitioners—This paper was motivated by the widespread use of programmable tooling machines (PTs) in multistage assembly processes (MAPs), such as controllable fixtures and computer numerical control (CNC) machines. These PTs automatically compensate for in-process dimensional errors and enhance the final assembly quality. Multiple PTs can be allocated within the same stage or across various stages, but an associated supporting platform (SP) is needed to install, maintain, and manage the PTs at a specific stage. Given the high costs of PTs and SPs, it is crucial to find an optimal placement of PTs that minimizes the number of PTs and required SPs while ensuring assembly products acceptable by industrial quality standards. To address this issue, we propose a novel methodology that formulates the problem into a sparse learning framework, which can be efficiently solved using our developed algorithm. This research provides valuable insights for practitioners working on the dimensional quality control of final products in MAPs and similar manufacturing processes. By implementing our proposed optimal placement strategy of PTs, practitioners can assure product quality while minimizing the associated total cost.}, archive = {J_TASE}, author = {Chengyu Tao and Juan Du and Jian Liu}, doi = {10.1109/TASE.2024.3515158}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9970--9982}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal placement of programmable tooling machines considering hierarchical structure via sparse learning for multistage assembly processes}, volume = {22}, year = {2025}, }
• (2025). Distributed observer-based dynamic-memory event-triggered security control for interconnected linear systems. TASE, 22, 9958--9969. (www)

  @article{ , abstract = {This paper explores the observer-based event-triggered security control problem for linear systems. Initially, a distributed dynamic-memory event-triggered scheme (DMETS) is proposed based on observed system states. This scheme is designed using information from past transmitted data packages, with the flexibility to dynamically adjust the quantity of transmitted data packages. Subsequently, a comprehensive mathematical model of hybrid cyber-attacks is established, and an observer-based distributed feedback control approach is proposed by utilizing the memory states. Furthermore, the asymptotic stability criterion of system is established by using Lyapunov direct method, and the design approaches of controller and observer are provided. Finally, contrast illustrative examples are presented to validate the advantage of DMETS and the effectiveness of the distributed control method. Note to Practitioners—This paper is prompted by the challenges of excessive consumption of network resources and the need for enhanced security control in linear systems. The prevailing approach to addressing this issue involves designing appropriate data transmission schemes, specifically ETS. However, a common limitation among these existing ETSs is the inability to efficiently utilize past transmitted data or dynamically adjust the quantity of data utilized in response to changes in system states. To overcome this limitation, we propose a DMETS, which aims to integrate memory characteristics and facilitate dynamic adjustments in memory data quantities.}, archive = {J_TASE}, author = {Jin Yang and Qishui Zhong and Hongjing Liang and Kaibo Shi and Zhao Yang Dong}, doi = {10.1109/TASE.2024.3515148}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9958--9969}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed observer-based dynamic-memory event-triggered security control for interconnected linear systems}, volume = {22}, year = {2025}, }
• (2025). Advanced two-stage islanding detection framework utilizing kinetic energy rate of change. TASE, 22, 9948--9957. (www)

  @article{ , abstract = {Inverter-based generations (IBGs) in distribution networks require robust anti-islanding protection systems (AIPS) to ensure operational integrity. However, conventional AIPS often struggle to detect accidental islanding events in scenarios where loads and generation are perfectly matched. Therefore, to get over this challenge, a novel hybrid anti-islanding protection scheme using Sandia frequency shift (SFS) phase angle transformation and RMS value of absolute frequency deviation at the point of common coupling (PCC) has been proposed in this paper. The proposed approach triggers the Sandia phase angle transformation solely upon suspicion of an accidental islanding occurrence, utilizing specific parameters monitored at the Point of Common Coupling (PCC). The approach outlined in this methodology is expected to enhance the power quality and stability of IBGs substantially. The performance analysis of the proposed hybrid method for the time domain has been carried out in MATLAB/Simulink. Experimental validation was conducted using a power hardware-in-the-loop (PHIL) testbed setup. The effectiveness of the proposed anti-islanding protection system (AIPS) was tested across more than 200 diverse islanding and non-islanding events. The results demonstrate its robustness and cost-effectiveness, with an islanding detection time as low as 180 milliseconds. Note to Practitioners—This paper introduces a new method for detecting unintentional islanding, a potentially hazardous scenario if left undetected. The proposed methodology offers improved robustness, ease of implementation, and higher accuracy than existing techniques. It aims to address power quality issues associated with active islanding detection methods and achieve zero non-detection zones. With a rapid detection time of only 180 milliseconds, the proposed method ensures swift and reliable identification of islanding events. Experimentally tested across more than 200 islanding and non-islanding scenarios, it proves to be highly accurate, making it a promising solution for real-time islanding detection applications.}, archive = {J_TASE}, author = {Rakesh Shamrao Patekar and Bijaya Ketan Panigrahi}, doi = {10.1109/TASE.2024.3514929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9948--9957}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Advanced two-stage islanding detection framework utilizing kinetic energy rate of change}, volume = {22}, year = {2025}, }
• (2025). A learning-based hybrid artificial bee colony algorithm for energy-efficient distributed heterogeneous type-2 fuzzy welding shop scheduling problem with factory eligibility. TASE, 22, 9934--9947. (www)

  @article{ , abstract = {In the era of economic globalization, the distributed heterogeneous welding shop scheduling problem (DHWSP) has been considered. Meanwhile, in some actual production scenarios, some jobs can only be processed in certain designated factories (i.e., factory eligibility), and the inevitable uncontrollable system disturbances (e.g., machine maintenance and human factors) lead to uncertain processing time for jobs. However, the research considering uncertain processing time in DHWSP with factory eligibility remains unexplored. Considering the advantages of interval type-2 fuzzy number (IT2FN) in representing the high level of uncertainty, the concept of IT2FN is introduced to tackle uncertain processing time. Then, under the context of green manufacturing, this paper investigates an energy-efficient distributed heterogeneous type-2 fuzzy welding shop scheduling problem with factory eligibility (EDHFWSP-FE). A learning-based hybrid artificial bee colony algorithm (LHABC) is designed to minimize both total energy consumption and makespan in EDHFWSP-FE. Within LHABC, a cooperative initialization is presented to create excellent initial solutions. In employed bee phase, a Q-learning based method is developed to help solutions select a superior neighborhood structure. In onlooker bee phase, a variable neighborhood search (VNS) is proposed to excavate promising neighborhood solutions. In scout bee phase, an estimation of distribution algorithm (EDA) based method is devised to generate new excellent solutions. Finally, experimental results on 27 test instances demonstrate that LHABC outperforms five other multi-objective optimization algorithms. Note to Practitioners—This paper aims to extend DHWSP to EDHFWSP-FE according to practical situations. In the actual production procedure, DHWSP poses more significant challenges. Not only operation sequencing, factory allocation and welding machine allocation matters, but also uncertain processing time and production constraint (e.g., factory eligibility) in the environment are important. Thus, this paper formulates an EDHFWSP-FE with the goals of minimizing total energy consumption and makespan. This problem model is applicable across many welding manufacturing enterprises with complex production environments. To solve this problem, we design a LHABC that hybridizes ABC with other effective methods to enhance its performance. Each component of LHABC is meticulously designed based on the specific characteristics of the problem domain, offering efficient schedules that are both energy-efficient and high-performing for practical implementation. Experimental results verify that all improved components of LHABC contribute to its performance, and LHABC outperforms other optimization algorithms in solving the problem.}, archive = {J_TASE}, author = {Fei Yu and Liang Gao and Chao Lu and Lvjiang Yin}, doi = {10.1109/TASE.2024.3514863}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9934--9947}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A learning-based hybrid artificial bee colony algorithm for energy-efficient distributed heterogeneous type-2 fuzzy welding shop scheduling problem with factory eligibility}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time safety control for unknown systems and its application to robotic manipulator. TASE, 22, 9923--9933. (www)

  @article{ , abstract = {In this paper, the concept of prescribed-time safety is introduced in the case that initial states are not in a safe set, which requires that the trajectories of systems visit a safe set within the prescribed time and then remain in the safe set. In contrast to safety-critical applications that require trajectories to always be in the safe set, prescribed-time safety control presents a greater challenge due to the introduction of a convergence time constraint. This paper presents a method to ensure prescribed-time safety for robotic systems. First, the prescribed-time control barrier function (PTCBF) is proposed for systems with relative degree of one, and it is extended to prescribed-time high order control barrier function (PTHOCBF) for systems with arbitrary relative degrees. Then, quadratic programming (QP) subject to PTCBF condition is constructed to solve control input. However, the proposed method cannot guarantee safety for systems with uncertain models. To address this limitation, a prescribed-time sliding mode disturbance observer (PTSMDO) is proposed to estimate the uncertainty. The estimated error is close to 0 before the time that states stay in the safe set. Based on the observed value of the uncertainty, the control input is solved by QP. Finally, the effectiveness of the proposed method is verified by a simulation and a physical experiment on Franka Emika robot. Note to Practitioners—This paper is motivated by the challenge encountered in robotic systems, which necessitate the control of the system’s state to reach a predefined safe set and subsequently carry out tasks within this safe set in various practical applications, particularly in the realm of robot motion planning. In comparison to existing methods such as finite-time CBF and fixed-time CBF, the proposed prescribed-time CBF approach in this paper offers the advantage of allowing users to preselect the arrival time. Additionally, given that robot systems are typically subject to uncertainties stemming from parameter errors and external disturbances, this paper also introduces a method to the issue of prescribed-time safety for robot systems featuring uncertain models. The uncertain term within the model can be accurately estimated prior to the moment when the state arrives at the safe set, employing the proposed disturbance observer. Subsequently, the prescribed-time CBF is integrated with the estimation results to construct QP for control input. The simulation and experiment conducted on the Franka Emika robot verify that the proposed method is feasible.}, archive = {J_TASE}, author = {Sihua Zhang and Di-Hua Zhai and Yuhan Xiong and Yuanqing Xia}, doi = {10.1109/TASE.2024.3514680}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9923--9933}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time safety control for unknown systems and its application to robotic manipulator}, volume = {22}, year = {2025}, }
• (2025). GPD: Learning geometric primitive deformation for unseen object pose estimation. TASE, 22, 9903--9922. (www)

  @article{ , abstract = {Witnessing the rapid progress and development in instance-level object pose estimation, increasing attention has shifted to the more challenging problem for unseen objects, which is in great demand for various robotic applications. In this paper, we propose the GPD, a novel framework for unseen object pose estimation, including both category-level and cross-category objects. The key innovation of the GPD model is the effective utilization of geometric primitives in target reconstruction and pose estimation, as it can generalize the learned primitive deformation across intra-class and inter-class instances. Additionally, we also design an advanced scheme for representative object feature extraction, including attention-aware excitation, multi-scale fusion, and semantic feature encoding. Extensive evaluations validate the effectiveness of individual innovation modules and the overall superior performance of the GPD. It not only achieves the SOTA results on category-level benchmarks CAMERA25 and REAL275, but also demonstrates impressive generalization ability across novel objects on the GraspNet-1Billion dataset. Furthermore, we deploy the trained GPD model for vision-guided robotic grasping experiments in simulation and real-world settings, again exhibiting its outstanding robustness and practicability in robotic manipulations. Note to Practitioners—This paper is motivated by the problem of unseen object pose estimation and robotic manipulation in unstructured environments. For intelligent robots expected to interact with their surroundings, rather than just passively perceiving them like surveillance cameras, 6Dof pose estimation is a critical capability. However, existing approaches generally face two key challenges. On the one hand, robots are likely to encounter unseen objects in real-world applications. Without the availability of prior models or specific training data for these unseen objects, instance-level and category-level methods may become ineffective or even fail to work. On the other hand, the error tolerance of precise tabletop robotic manipulation is very tight, and the varying lighting conditions and background noise impose higher robustness requirements on pose estimation algorithms. To address these difficulties, we propose a novel network that learns geometric primitive deformation for pose estimation. This model is less dependent on object prior information, thereby enhancing the generalization ability. Additionally, by incorporating cross-modal excitation and multi-scale fusion during feature extraction, our model can capture representative appearance and geometric information of objects for accurate pose estimation. Extensive experimental results on benchmark datasets quantitatively validate the superior performance of our approach. We also demonstrate its effectiveness in robotic applications through unseen object grasping experiments on Kinova and Franka Emika robot platforms. In the future, we plan to explore primitive combination schemes for compound object representation, enabling pose estimation for more complex-shaped objects.}, archive = {J_TASE}, author = {Qiwei Meng and Jason Gu and Yun-Hui Liu}, doi = {10.1109/TASE.2024.3514143}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9903--9922}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GPD: Learning geometric primitive deformation for unseen object pose estimation}, volume = {22}, year = {2025}, }
• (2025). Statistical feasibility robust optimization with polyhedron uncertainty set for hydrogen-data center microgrid operations. TASE, 22, 9884--9902. (www)

  @article{ , abstract = {Robust optimization (RO) has been widely used in the hydrogen-data-center microgrid (H2-DCMG) optimal operations. However, the operation results based on RO are too conservative. Statistical feasibility can be introduced into RO to reduce conservatism. Therefore, statistical feasibility-based RO is adopted in the H2-DCMG operations optimization. In this study, a data-driven statistical-feasibility-based robust rolling optimization framework is constructed for the optimal operations of the c-DCMG. In this framework, DC temperature is influenced by the uncertain outside temperature, and an uncertainty set is needed to express the uncertain outside temperature. Unfortunately, the existing studies that construct the uncertainty set satisfying statistical feasibility only constructs the ellipsoid uncertainty sets. The ellipsoid uncertainty sets will be converted into the second-order cone constraints, which will increase the complexity when solving. In this study, the Statistical-Guarantee-based Vertex Link (SGVL) algorithm is proposed to construct the polyhedron uncertainty sets, which are used to describe the uncertainty of the outside temperature. Moreover, statistical feasibility-based DC temperature bound is guaranteed by the optimal operations obtained based on these uncertainty sets. Compared with the traditional ellipsoid uncertainty set, the polyhedron uncertainty set can reduce the complexity of the optimization problem and improve the efficiency of the solution process. Case studies based on the real-world temperature dataset are processed. The results show that the introduction of statistical feasibility can reduce the total operation cost by 0.29%~0.64%. The average solving time of the optimization problems based on the polyhedron uncertainty sets constructed using the SGVL algorithm also reduces by 7%~13%. The cases also verify that other uncertainty parameters and different kinds of forecasters do not influence the performance and effectiveness of the framework. Note to Practitioners—This paper focuses on optimizing the operations of a hydrogen-data center microgrid (H2-DCMG) to minimize its operation cost in the way of statistical feasibility-based robust optimization. The statistical feasibility is introduced into the framework for the relaxation of the data center (DC) temperature bound. The Statistical-Guarantee-based Vertex Link (SGVL) algorithm is proposed to construct the uncertainty set of outside temperature that guarantees the dispatch results satisfying statistical feasibility. In practice, the settings of the framework and the data used in the SGVL are significant. Firstly, in the rolling dispatch, the length of the forecasting period needs also to be set after the trade-off between the solving efficiency, the dispatch economy and the DC temperature stability. Secondly, the pre-set DC temperature bound is advisable to be set below the maximum temperature that does not harm its normal operation. When constructing the uncertainty sets, the two parameters of the statistical feasibility need also to be set after the trade-off between the economics of dispatch results and the acceptability of constraint violations. The dataset and the actual forecasting error need to remain equally distributed to guarantee the performance of the dispatch of the proposed framework. The proposed framework can be readily implemented and integrated into the actual dispatch of H2-DCMG.}, archive = {J_TASE}, author = {Juntao Duan and Yuanzheng Li and Yang Li and Pu Yang and Zhigang Zeng}, doi = {10.1109/TASE.2024.3514101}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9884--9902}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Statistical feasibility robust optimization with polyhedron uncertainty set for hydrogen-data center microgrid operations}, volume = {22}, year = {2025}, }
• (2025). Supply chain design optimization with heterogeneous risk-aware agents. TASE, 22, 9872--9883. (www)

  @article{ , abstract = {Modern supply chain networks (SCN) are becoming increasingly complex, with vulnerable entities exposed to uncertain disruptions that affect local or global supply chain attributes. We model a stochastic mixed-integer program to minimize the overall cost of SCN design and operations, in response to lead-time and demand uncertainties following given probability distributions. We formulate a heterogeneous risk-aware model to trade off between cost and delay/shortage by considering different risk-attitudes amongst supply chain agents. In particular, we employ the Conditional Value-at-Risk (CVaR) as a coherent risk measure for quantifying risk while attaining solution tractability. We derive managerial insights from our numerical studies, finding the most benefit from diversifying agents in the root tier, since their disruptions affect all other tiers in the SCN. We find that as agents become more risk averse, the optimal solutions for key agents (such as assemblers), seek more backup suppliers and allocate extra capacities to achieve resiliency and reliability. Practitioners can use the outcomes of our framework and studies to guide SCN design considering heterogeneous risk attitudes between agents. Note to Practitioners—With growing uncertainties in global supply chains, inefficient responses to disruptions can lead to large penalties and long-term impacts such as customer dissatisfaction. This research is motivated by the challenges arising during the operations of supply chains under both lead-time and demand uncertainties. We employ optimization and centralized control approaches to optimize supply-chain network design as well as response strategies to disruptions, and our framework can handle heterogeneous risk preferences as it models the risk attitude of each individual entity or agent in supply chains. Our model can be utilized to completely or partially re-design resilient supply chains, to better prepare for unknown features and uncertainties. Our case study provides insights about risk-averse supply-chain designs that can reduce response cost, but increase initial investments on backups and redundancies.}, archive = {J_TASE}, author = {Juan-Alberto Estrada-Garcia and Dawn M. Tilbury and Kira Barton and Siqian Shen}, doi = {10.1109/TASE.2024.3513815}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9872--9883}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Supply chain design optimization with heterogeneous risk-aware agents}, volume = {22}, year = {2025}, }
• (2025). Modeling, positioning, and deep reinforcement learning path following control of scaled robotic vehicles: Design and experimental validation. TASE, 22, 9856--9871. (www)

  @article{ , abstract = {Mobile robotic systems serve as versatile platforms for diverse indoor applications, ranging from warehousing and manufacturing to test benches dedicated to evaluating automated driving (AD) functions. In AD systems, the path following (PF) layer is responsible for defining steering commands to follow the reference path. Recently explored approaches involve artificial intelligence-based methods, such as Deep Reinforcement Learning (DRL). Despite their promising performance, these controllers still suffer from time-consuming training phases and may experience performance degradation when deviating from training conditions. To address these challenges, this paper proposes novel DRL controllers addressing the simulation-to-reality gap in unknown scenarios by: (i) training via an expert demonstrator which also speed up the learning phase; and (ii) a weight adaptation strategy for the resulting neural network (NN) to strengthen controller robustness and enhance PF performance. In addition, an experimentally validated vehicle model is used for training the proposed DRL algorithm and as a model for a federated extended Kalman filter (FEKF) system employed for sensor fusion in vehicle localisation. The proposed DRL-based PF controllers are experimentally evaluated through key performance indicators across multiple maneuvers not considered during training, and it is shown that they outperform benchmarking model-based controllers from the literature. Note to Practitioners—This paper presents a comprehensive toolchain for controlling mobile robots, which includes: (i) a simple yet effective two-stage least-square approach for parameter identification of the longitudinal and lateral dynamics of scaled robotic vehicles; (ii) the utilisation of a no-reset FEKF to enhance positioning leveraging all sensors commonly available on scaled robotic vehicles; (iii) the inclusion of an expert demonstrator to expedite the training phase and address the simulation-to-reality gap resulting from discrepancies between simulation and experimental environments; and (iv) an adaptation strategy for dynamically adjusting the weights of the resulting NN to further improve robustness for scenarios not considered during the traning.}, archive = {J_TASE}, author = {Carmine Caponio and Pietro Stano and Raffaele Carli and Ignazio Olivieri and Daniele Ragone and Aldo Sorniotti and Patrick Gruber and Umberto Montanaro}, doi = {10.1109/TASE.2024.3513701}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9856--9871}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling, positioning, and deep reinforcement learning path following control of scaled robotic vehicles: Design and experimental validation}, volume = {22}, year = {2025}, }
• (2025). Output-feedback adaptive periodic disturbances attenuation for linear MIMO systems subject to input delay. TASE, 22, 9843--9855. (www)

  @article{ , abstract = {This paper presents a novel output-feedback direct adaptive controller with a decoupling design to completely attenuate periodic disturbances in the linear multi-input multi-output (MIMO) system subject to input delay. First, the family of strictly causal stabilizing controllers for this system is derived by using the Youla parameterization method. The stabilizing controller consists of a fixed base controller J and an adjustable block Q. Periodic disturbances can be completely attenuated by adjusting the block Q. Then, to improve the attenuation rate of the periodic disturbance, a diagonal decoupling strategy is proposed. Based on the internal model principle (IMP), the interpolation condition for the complete attenuation of periodic disturbance is established. Next, to attenuate unknown time-varying periodic disturbances, a parameter adaptive algorithm is designed to update online the Q parameters. Finally, the effectiveness of the proposed controller is validated by simulation results. Note to Practitioners—The simultaneous presence of input delays and periodic disturbances in real systems not only limits control effectiveness but also leads to instability in the closed-loop system. Therefore, it is crucial to attenuate the periodic disturbances in input delay systems. However, existing adaptive controllers cannot completely attenuate unknown time-varying periodic disturbances if the system state is unavailable. Moreover, the multi-channel cross-coupling inherent in MIMO systems further constrains the controller performance. Motivated by this challenge, this paper proposes an output-feedback direct adaptive controller with decoupling design to completely attenuate periodic disturbances in linear MIMO input delay systems. It is hoped that the proposed method can provide valuable theoretical and technical guidance for the design of controllers aimed at compensating for periodic disturbances in linear MIMO systems subject to input delay.}, archive = {J_TASE}, author = {Jiahao Zhu and Shujin Yuan and Lisheng Mou and Jun Luo and Huayan Pu}, doi = {10.1109/TASE.2024.3513396}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9843--9855}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Output-feedback adaptive periodic disturbances attenuation for linear MIMO systems subject to input delay}, volume = {22}, year = {2025}, }
• (2025). Trajectory progress-based prioritizing and intrinsic reward mechanism for robust training of robotic manipulations. TASE, 22, 9829--9842. (www)

  @article{ , abstract = {Training robots by model-free deep reinforcement learning (DRL) to carry out robotic manipulation tasks without sufficient successful experiences is challenging. Hindsight experience replay (HER) is introduced to enable DRL agents to learn from failure experiences. However, the HER-enabled model-free DRL still suffers from limited training performance due to its uniform sampling strategy and scarcity of reward information in the task environment. Inspired by the progress incentive mechanism in human psychology, we propose Progress Intrinsic Motivation-based HER (P-HER) in this work to overcome these difficulties. First, the Trajectory Progress-based Prioritized Experience Replay (TPPER) module is developed to prioritize sampling valuable trajectory data thereby achieving more efficient training. Second, the Progress Intrinsic Reward (PIR) module is introduced in agent training to add extra intrinsic rewards for encouraging the agents throughout the exploration of task space. Experiments in challenging robotic manipulation tasks demonstrate that our P-HER method outperforms original HER and state-of-the-art HER-based methods in training performance. Our code of P-HER and its experimental videos in both virtual and real environments are available at https://github.com/weixiang-smart/P-HER. Note to Practitioners—This work is motivated to develop a fast and effective learning method for intelligent robotic manipulation of typical industrial tasks, including pushing, picking, and placing workpieces, which are essential and fundamental processing plan activities for accomplishing robotic machining and assembly applications towards smart manufacturing. The introduction of reinforcement learning enables robots to learn manipulation tasks autonomously, which can save the effort for engineers to teach or hard program the robot and also reduce labor costs. However, the existing HER-based reinforcement learning algorithms are with low training efficiency and performance due to the uniform sampling and scant task reward. Inspired by human learning, this work introduces a progress incentive mechanism to identify valuable trajectory data for effective training. In addition, a novel rewarding method, that applies additional intrinsic rewards for agents learning valuable trajectory space, results in fast and robust learning. The setting of important weight parameters in the rewarding method is given in the paper, which provides a practical reference for applying the proposed algorithm. The average success rate of two actual manipulation tasks in simulation and real robotic manipulation environments are 96% and 92.5%, respectively, which demonstrates that the method is effective for both environments and there is 3.5% average gap of successful rate dropping from simulation scenarios to real ones due to the inherent mismatches between simulation and reality. The high success rate demonstrated in the real Workpieces-sorting task exemplifies the potential of the trained policies for application in industrial scenarios.}, archive = {J_TASE}, author = {Weixiang Liang and Yinlong Liu and Jikun Wang and Zhi-Xin Yang}, doi = {10.1109/TASE.2024.3513354}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9829--9842}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trajectory progress-based prioritizing and intrinsic reward mechanism for robust training of robotic manipulations}, volume = {22}, year = {2025}, }
• (2025). Adaptive neural sliding mode control for singularly perturbed systems with randomly perturbed sampling periods. TASE, 22, 9821--9828. (www)

  @article{ , abstract = {This paper addresses the problem of adaptive neural network sliding mode control for sampled-data singularly perturbed systems subjected to randomly perturbed sampling periods and unrestricted deception attacks. In contrast to the conventional transition probabilities,the utilization of sojourn probabilities are essential to capture the inherent randomness associated with the sampling periods. Furthermore, in the presence of unrestricted deception attacks, a neural network-based approach is utilized to estimate and mitigate their impact on system performance. Additionally, a sliding mode controller is developed based on both the sampling period mode and the singularly perturbed parameter. This controller ensures the exponential ultimate boundedness in the mean square sense and the reachability of the specified sliding surface for the closed-loop system. Finally, the proposed theory is validated through a practical example, demonstrating its effectiveness and soundness. Note to Practitioners — Due to the advent of the digital era, periodically sampled-data systems have received much attention. However, practical engineering conditions often introduce non-ideal factors, such as network congestion, noise interference, and voltage instability, which can cause randomly perturbed sampling. As considered in this paper, a sojourn probability, which is more easily obtained by statistical methods, is applied to model the variation of sampling periods. On the other hand, in real network environments, the boundedness of an attacker’s injection attack cannot be guaranteed. Therefore, in this paper, we consider using a neural network-based approach to estimate and mitigate their impact on system performance in the presence of unrestricted deception attacks. Meanwhile, sliding mode control, which has drawn widespread attention for its cost-saving and reliability, is utilized to achieve exponential ultimate boundedness in the mean square sense and reachability of the specified sliding surface of the controlled system.}, archive = {J_TASE}, author = {Jun Cheng and Shan Liu and Huaicheng Yan and Ju H. Park}, doi = {10.1109/TASE.2024.3513314}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9821--9828}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neural sliding mode control for singularly perturbed systems with randomly perturbed sampling periods}, volume = {22}, year = {2025}, }
• (2025). Active disturbance rejection predefined-time consensus tracking for nonlinear multiagent systems via self-triggered communication. TASE, 22, 9810--9820. (www)

  @article{ , abstract = {This paper addresses the active disturbance rejection predefined-time consensus tracking problem for non-affine nonlinear multiagent systems (MASs) with self-triggered communication. With the active disturbance rejection strategy, the uncertainty of the system can be estimated and the issue of explosion of complexity under the backstepping framework is handled. To improve the utilization of network resources, a self-triggered mechanism that the next trigger moment is calculated from the current information is adopted, for communication of each follower. The advantage of this triggered mechanism is that the triggering conditions are not required to be continuously monitored. Meanwhile, a time-varying tuning function-based predefined-time method is designed to achieve convergence within the predetermined time. Compared with existing finite/fixed-time convergence consensus schemes, its biggest advantage is that the upper bound of convergence time can be a positive constant arbitrarily selected by users. Additionally, the problem of input saturation is considered in the controller design to improve the generality of our scheme. Finally, the effectiveness of the proposed scheme is verified via simulation example. Note to Practitioners—In this paper, the active disturbance rejection predefined-time consensus tracking control problem for MASs with non-affine properties is discussed, which is embodied in numerous applications, such as sensor networks, space satellites and robot formations. The convergence time required for all agents to conclude consensus is used as an important indicator to evaluate the control performance in the consensus control of MASs. The existing control methods including finite and fixed time methods can only achieve the system to converge in a preset time. Based on this, it is a challenging topic to achieve predefined-time convergence by setting the accurate convergence time t. Meanwhile, to improve the utilisation of network resources, a self-triggered mechanism is designed. The advantage of this triggered mechanism is that the triggering conditions are not required to be continuously monitored, which greatly reduces the control cost. Additionally, the problem of input saturation is considered in the controller design to improve the generality of our scheme.}, archive = {J_TASE}, author = {Zhenhuan Wang and Zhongwen Cao and Liang Zhang and Huanqing Wang and Ning Xu}, doi = {10.1109/TASE.2024.3513297}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9810--9820}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active disturbance rejection predefined-time consensus tracking for nonlinear multiagent systems via self-triggered communication}, volume = {22}, year = {2025}, }
• (2025). Efficient constrained motion planning using direct sampling of screw-constraint manifolds. TASE, 22, 9793--9809. (www)

  @article{ , abstract = {Manipulating articulated objects is especially difficult if the robot is operating autonomously or far from any human operator. Object articulation imposes strict constraints on robot motion, making it a challenge to generate valid trajectories to complete the task. Problems compound when the robot is mobile and operates in an uncontrolled environment, where the location or articulation model is unknown a priori. In this work, we leverage screw theory to model constraints imposed on a generic manipulator by simple articulated objects and present two novel, fast, and robust methods–Sequential Path Stepping (SPS) and Direct Screw Sampling (DSS)–for planning trajectories by directly sampling these constraints. We show that these methods are hardware-agnostic and work in cluttered environments using long, complex paths modeled by multiple screw-axis constraints. We demonstrate that modeling constraints using multiple screw axes handles objects with multiple DoF, or multi-step tasks (e.g., turning a knob before opening the door). In addition, the direct sampling component of the proposed approaches is implemented as a module that used with existing well-known probabilistic planning methods, allowing customization across different hardware, domains, or planning problems. We validate our methods across many planning and inverse kinematic elements, with three different mobile and stationary manipulators, and on a set of challenging planning problems that include single- and multiple-screw constraints. Results demonstrate a 97.6% success rate planning in cluttered environments, in less than 0.2 seconds.Note to Practitioners—This paper was motivated by the articulated manipulation problem for mobile manipulators. The solutions also apply to non-articulated object manipulation and task planning. Existing approaches do not include helical constraints (e.g., turning a threaded bolt), require operator oversight, or lack integration with MoveIt: the de facto kinematic manipulation standard. We address these issues with two methods that utilize screw theory to enable helical (inclusive of revolute and prismatic) constraints, require little input from operators, and align with standard plan-execute, task definition, and robot configuration capabilities offered by MoveIt and ROS. Through experimentation, we show that our methods easily plan manipulations of arbitrary articulated objects-including those with multiple DoF-are relatively quick, successful, and hardware agnostic.}, archive = {J_TASE}, author = {Adam Pettinger and Janak Panthi and Farshid Alambeigi and Mitch Pryor}, doi = {10.1109/TASE.2024.3513160}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9793--9809}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient constrained motion planning using direct sampling of screw-constraint manifolds}, volume = {22}, year = {2025}, }
• (2025). Rule-based predictive control for battery scheduling in microgrids under power generation and load uncertainties. TASE, 22, 9782--9792. (www)

  @article{ , abstract = {This paper addresses the control of the state of charge (SoC) of a Battery Energy Storage System (BESS) in a microgrid, considering uncertainties in load and Renewable Energy Sources (RES) generated power estimations. To achieve this objective, we propose RubPC, a novel rule-based Model Predictive Control (MPC). We partition the feasible operation space of the microgrid into two subzones, referred to as the white and yellow zones. The yellow zone represents the boundary space between the feasible and unfeasible operation spaces. In RubPC, we initially implement MPC on a predefined optimization window to determine the optimal SoC of the BESS, aiming to keep the microgrid within the white zone. Noting that mismatches between estimated and actual load and generated power may lead to constraint violations, we introduce a rule-based controller as a supervisory control. This controller monitors the microgrid’s state, and if the microgrid enters the yellow zone, it adjusts the control to maintain the microgrid within the white zone. We validate our proposed method by simulating it using data from an electrified quarry site in Sweden. Note to Practitioners—Optimizing the charge and discharge schedule of BESSs in microgrids offers a promising avenue for substantial economic and technical benefits. However, the successful realization of these benefits hinges on accurately estimating and aligning the values of load and RES-generated power. In industries, the consequences of mismatches between these estimates and actual values can translate into unexpected costs that may outweigh the anticipated economic benefits of BESS’s optimal charge and discharge schedule. This paper underscores the critical importance of addressing this concern to ensure the viability of BESS applications in various industries. To tackle this challenge, we present RubPC, an innovative Rule-Based MPC framework. Unlike conventional approaches, RubPC is specifically designed to effectively handle discrepancies between estimated and actual values, thereby preventing potential constraint violations. Our aim is to offer practitioners a robust solution that not only brings economic benefits but also ensures their safe and reliable operation.}, archive = {J_TASE}, author = {Mojtaba Kaheni and Jiali Fu and Alessandro V. Papadopoulos}, doi = {10.1109/TASE.2024.3512882}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9782--9792}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rule-based predictive control for battery scheduling in microgrids under power generation and load uncertainties}, volume = {22}, year = {2025}, }
• (2025). Enhancing antipodal grasping modalities in complex environments through learning and analytical fusion. TASE, 22, 9767--9781. (www)

  @article{ , abstract = {The robotic pick-and-place system is applied widely in many fields such as assembly, packaging, bin-picking, and sorting. In this paper, we present a deep learning and analytical-based method for generating antipodal grasping multi-modality in highly cluttered scenes. Our method takes advantage of three types of grasp poses to deal with the complexity of environment and achieves efficient computation time for real applications. A new synthetic training datasets are generated in Isaac Sim including approximately 35000 RGB-D images and an automatic labeling algorithm is developed. We utilize convolutional neural networks (CNNs) for predicting antipodal grasping parameters on objects and a filtering algorithm to avoid collisions and calculate grasp’s depth simultaneously. Our approach processes the entire task in approximately 0.2 seconds, achieving a success rate of over 96% and more than 98% collision-free grasps in cluttered scenes. The method was verified by experiments with RB10 robot arm, two-fingers grippers, depth camera L515, and several objects in different scenes. The article shows a simple, effective, and highly applicable approach in real environments. The real experimental video of our method is shown at https://www.youtube.com/watch?v=GvJZxUyQr3w. Note to Practitioners—The robotic pick-and-place task is fundamental to the progress of automation. This article presents a novel method for efficiently picking objects from complex environments characterized by obstacles, overlaps, and a diversity of objects. These scenarios pose challenges in approaching objects and avoiding collisions. Our method combines learning and analytical approaches, demonstrating high performance in terms of accuracy, generalization across various scenes and grippers, and rapid computation, with the entire process taking approximately 0.2 seconds. These advancements are validated through a series of real robotic grasping experiments and comparisons with state-of-the-art methods. We believe that our method represents a significant contribution to the field of automation science and holds promise for real-world applications.}, archive = {J_TASE}, author = {Tat Hieu Bui and Yeong Gwang Son and Juyong Hong and Yong Hyeon Kim and Hyouk Ryeol Choi}, doi = {10.1109/TASE.2024.3512005}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9767--9781}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing antipodal grasping modalities in complex environments through learning and analytical fusion}, volume = {22}, year = {2025}, }
• (2025). A discrete-time neural network control method based on deterministic learning for upper-limb rehabilitation robot. TASE, 22, 9753--9766. (www)

  @article{ , abstract = {Accurate trajectory training is a challenging issue of upper-limb rehabilitation robots. This paper presents a novel discrete-time neural network control method to address the problems of system uncertainties and tracking accuracy in repetitive trajectory training. This control method consists of both an adaptive neural network controller and a learning controller. The adaptive neural network controller satisfying persistent excitation condition enables not only stable tracking control, but also accurate learning for closed-loop system dynamics. The learning controller utilizes the learned knowledge to provide high-performance control. In order to examine the effectiveness of the proposed control method, a series of simulation and real-world experiments with system uncertainties were conducted, in comparison of proportion integration differentiation control, sliding mode control and event-triggered adaptive neural control. Results substantiate that the proposed control method can precisely learn the unknown dynamics of human-robot system along the subject-specific reference trajectories, and control the robot to assist the arm for accurate and fast trajectory tracking with small control gains by reutilizing the learned knowledge. This control method may play a role in accurate trajectory training for upper-limb rehabilitation robots. Note to Practitioners—This work is motivated by the practical requirements of rehabilitation robots in repetitive motor training. Trajectory tracking is a fundamental but efficient training mode of rehabilitation robots. However, uncertainty and nonlinearity of the human-robot system dynamics may increase the difficulty of controlling the robots for accurate, efficient and reliable trajectory tracking training. To this end, this paper proposes a learning-based control method, which could learn the uncertain and nonlinear system dynamics by utilizing an elaborately designed neural network controller and thus achieve superior control performance using the learned knowledge. This control method can be potentially applied in variety of rehabilitation robots, showing advantages for repetitive trajectory training. First, it can accurately mode the uncertain dynamics of human-robot system and achieve personalized rehabilitation. Second, it does not need any parameter adaptation in the similar repeated motions, and can be more easily designed with digital implementations, thereby achieving better performance in the aspects of time saving. Third, it can ensure the tracking accuracy of the rehabilitation robot for rehabilitation efficiency and avoid secondary injury.}, archive = {J_TASE}, author = {Na Zhang and Fukai Zhang and Yibin Li and Cong Wang and Ke Li}, doi = {10.1109/TASE.2024.3511996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9753--9766}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A discrete-time neural network control method based on deterministic learning for upper-limb rehabilitation robot}, volume = {22}, year = {2025}, }
• (2025). Reverse engineering for unbalance excited vibration control in rotating system combining force-vibration transmission model and parameter identification. TASE, 22, 9741--9752. (www)

  @article{ , abstract = {In advanced manufacturing field, reverse engineering based on development requirements plays an important role to ensure the performance of the mechatronic equipment. As a main vibration source, control of unbalanced excitation from rotating system can realize overall vibration performance improvement by avoiding it been amplified through transmission paths. This paper proposes the control method for the steering wheel vibration excited by the condenser-radiator-fan module (CRFM) in automobile design and manufacturing based on reverse engineering, involving the force-vibration transformation model and parameter identification. The force-vibration transmission mechanism is analyzed by formulating the established vibration model from the CRFM to the steering wheel. Then, the design requirements for the CRFM is reversely derived based on the development requirements of the steering wheel vibration level and the formulated force-vibration transfer function (FVTF) following the reverse engineering principle. Meanwhile, determination of design requirements, involving the FVTF and physical natural modal frequency, depends on parameter identification according to the field testing during the development process. To further figure out the crucial design parameters for reverse engineering, detailed analyses about the influences of FVTF parameters on the vibration isolation effect are discussed. The feasibility of the proposed vibration control development process is evaluated by applying it in solving an abnormal steering wheel vibration problem of a car. The design guidance for the steering wheel vibration control, involving the hardness and geometry of the CRFM vibration isolation rubber, are analyzed in detail, following the proposed reverse engineering process. The field-testing results show that the vibration level of vehicle steering wheel can be controlled effectively by the proposed method, and provides guiding significance for analyze of system vibration problems caused by rotating system excitation. Note to Practitioners—The unbalance excited vibration from rotating system has been the main source in mechatronic equipment, which can be most likely amplified through transmission paths. This paper proposes a control method for the steering wheel vibration excited by the condenser-radiator-fan module in automobile design and manufacturing based on reverse engineering, involving the force-vibration transformation model and parameter identification. The practical value of the proposed vibration control development process is evaluated by applying it to solving an abnormal steering wheel vibration problem in a car. The natural frequency design requirements of the cooling module, involving the hardness and geometry of the rubber, are analyzed and given in detail, including design guidance. The online vehicle steering wheel vibration test results show that the vibration level can be controlled effectively by the proposed method, which provides guiding significance for analyzing system vibration problems caused by rotating system excitation.}, archive = {J_TASE}, author = {Huimin Shen and Jintao Ding and Lihong Yang}, doi = {10.1109/TASE.2024.3511576}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9741--9752}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reverse engineering for unbalance excited vibration control in rotating system combining force-vibration transmission model and parameter identification}, volume = {22}, year = {2025}, }
• (2025). Listen, perceive, grasp: CLIP-driven attribute-aware network for language-conditioned visual segmentation and grasping. TASE, 22, 9729--9740. (www)

  @article{ , abstract = {Endowing robots with the ability to understand natural language and execute grasping is a challenging task in a human-centric environment. Existing works on language-conditioned grasping achieve end-to-end grasping detection based on language. However, these works lack fine-grained visual grounding, resulting in cognitive deficits for robots. Moreover, they ignore the correlation between visual attributes of objects and grasping, leading to coarse grasp poses. To this end, we propose a CLIP-driven aTtribute-aware network (CTNet) for language-conditioned visual segmentation and grasping, enabling the robots to listen, perceive, and grasp the referred object in real-world applications. Specifically, we first employ Listen stage to understand basic linguistic and visual concepts. Subsequently, we introduce Perceive stage to mine multi-modal features and visual attribute cues (e.g., boundary and spatial location), then yield a language-conditioned segmentation mask. Further, we design Grasp stage to aggregate the perceived attribute information and refine the spatial location and grasping rectangle, generating a high-quality grasp pose. Lastly, we provide an extended large dataset Ref-OCID-Grasp to train and test our method, achieving a grasping accuracy of 97.76% and segmentation OIoU of 91.82%. The real-world robotic applications demonstrate the effectiveness of our proposed approach. The project, video, and dataset can be found at https://ctnetgrasp.github.io. Note to Practitioners—Most of the existing grasping methods focus on clearing all objects in the workspace. However, as robots integrate into human society, robots should learn to grasp the desired object by understanding human language. Therefore, language-conditioned grasping is a significant skill for human-robot collaboration. The prior works directly complete the grasp detection through the language-grasp paradigm, but they ignore the discussion on whether the robot understands the concept of vision and language expression of the object. Therefore, this paper proposed the Listen-Perceive-Grasp paradigm, in which the Listen-Perceive stage is responsible for the conception alignment of the object in language expression and visual pixels, and the Perceive-Grasp stage achieves the constraining and refining the grasp detection by the perceived visual attributes such as boundary and shape. Experiments show that this method can obtain a refiner grasp pose in cluttered environments and perform language-conditioned grasping well in the real world. In future research, we will work on 6-DoF grasping and multi-object disambiguation conditioned on language.}, archive = {J_TASE}, author = {Jialong Xie and Jin Liu and Saike Huang and Chaoqun Wang and Fengyu Zhou}, doi = {10.1109/TASE.2024.3510777}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9729--9740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Listen, perceive, grasp: CLIP-driven attribute-aware network for language-conditioned visual segmentation and grasping}, volume = {22}, year = {2025}, }
• (2025). Kinematic calibration of a novel two-axis solar tracker. TASE, 22, 9718--9728. (www)

  @article{ , abstract = {This paper deals with the kinematic calibration of a two-axis over-constrained solar tracker. In an over-constrained mechanism, it is difficult to derive the constraint among geometric errors and then the accurate error model can not be achieved. This paper presents a new error modeling method with considering constraints among geometric errors to derive a complete, continuous and minimal model. The constraint among geometric errors are derived from an algebraic approach with strict theoretical proof. Furthermore, an iterative Tikhonov regularization method is presented for parameter identification and the iteration termination condition is related to the measurement noise. The traditional Tikhonov regularization method is a posterior method that depends on the measurement data in calibration, while the proposed regularization method is a prior method with clear physical meaning. The kinematic calibration of the two-axis solar tracker is used to validate the proposed methods. The orientation error of the solar tracker is reduced by more than 95% after kinematic calibration. Note to Practitioners—The geometric error constraints are proven in the article to exist in an over-constrained two-axis solar tracker, which leads to the unsuitability of previous calibration methods. With our proposed method, the accuracy of the solar tracker has been improved by 95% after calibration. In our calibration, the error modeling method can be extended to other over-constrained robots, and the Tikhonov regularization method has good stability and accuracy in error identification. In future research, we will address the problem of deformation and joint clearance in kinematic calibration.}, archive = {J_TASE}, author = {Jun Wu and Hao Ye and Yanling Tian}, doi = {10.1109/TASE.2024.3510745}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9718--9728}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Kinematic calibration of a novel two-axis solar tracker}, volume = {22}, year = {2025}, }
• (2025). Adaptive dynamic surface funnel control of hydraulic systems with finite-time prescribed tracking performance. TASE, 22, 9708--9717. (www)

  @article{ , abstract = {Hydraulic servo systems are generally affected by disturbances, parametric uncertainties, and unmodeled dynamics. In order to improve the tracking performance of hydraulic servo systems, a new adaptive output feedback funnel control method with uncertainty compensation is proposed based on dynamic surface technology, which avoids the differential explosion problem generated in backstepping control and keeps the tracking performance always within a predetermined range. The observers and adaptive control are fused together to compensate for system model uncertainties and estimate unknown states; Meanwhile, a new control error variable with finite-time performance function is used to constrain system performance. The closed-loop system using this controller has high accuracy and strong robustness. It has been proven that the estimation errors are bounded and the control error of the closed-loop system is kept within the bound of funnel function. Experiments were conducted to test the high-performance properties of the proposed algorithm. Note to Practitioners—Hydraulic servo systems have received increasing research attention in recent years because of their widespread range of applications. In this paper, the high transient performance and precise steady-state tracking control of hydraulic servo system are studied. We propose a new adaptive output feedback funnel control method with uncertainty compensation based on dynamic surface technology. the associated stability proof is then presented. Experiments are conducted to demonstrate the effectiveness of the approach. In future research, we will address the problem of high precision control of hydraulic manipulator.}, archive = {J_TASE}, author = {Zhangbao Xu and Zhenle Dong and Hao Shen and Jianyong Yao}, doi = {10.1109/TASE.2024.3510744}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9708--9717}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive dynamic surface funnel control of hydraulic systems with finite-time prescribed tracking performance}, volume = {22}, year = {2025}, }
• (2025). ILC-based tracking control for linear systems with external disturbances via an SMC scheme. TASE, 22, 9698--9707. (www)

  @article{ , abstract = {Iterative Learning Control (ILC) is renowned for its capability to achieve precise tracking control for systems with repetitive actions at a fixed time interval. However, pursuing the dual objective of high-precision tracking and rapid convergence is a persistent challenge in the field of learning control. To address this problem, a novel ILC method is designed for a class of discrete-time linear systems subject to non-repetitive disturbances in this paper. Particularly, the updating term in ILC is constructed inspired by the principle of sliding mode control (SMC), which results in the learning process being divided into two distinct stages: a rapid reaching stage and a slow sliding stage. As a result, a balance between convergence speed and tracking performance can be ensured via the proposed ILC method. In addition, to attenuate the effects of non-repetitive disturbances, the disturbance compensation mechanism is integrated into the proposed ILC method. Moreover, the optimal value of the learning gain can be determined using the predicted root mean square (RMS) errors of subsequent iterations, eliminating the need for additional tuning actions. Finally, simulation examples are provided to validate the effectiveness and superiority of the proposed new ILC method. Note to Practitioners—For many mechanical components in mechatronic systems and robotics, the motions are repeatable. Iterative learning control (ILC) is a well-established technique ideally suited for enhancing the performance of such repetitive tasks without excessive requirements on sensor-feedback quality or control-loop bandwidth. However, most existing ILC approaches in the literature primarily focus on improving convergence accuracy, while little attention is paid to convergence speed in the iteration domain, especially in the presence of disturbances. This paper addresses the limitations of classical ILC schemes, and draws inspiration from the sliding mode control (SMC) technique. To be specific, a novel SMC-based ILC algorithm is proposed that allows to achieve a good balance between the fast convergence and precise tracking performance, especially in case of iteration variant disturbances. Also, it will be shown how the optimal learning gains can be determined. Base on the examples of multi-axis gantry robot and injection molding process, simulations support the theoretical results, and meanwhile show the effectiveness and advantage of the proposed ILC strategy.}, archive = {J_TASE}, author = {Rongni Yang and Yingjie Gong and Wojciech Paszke}, doi = {10.1109/TASE.2024.3510738}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9698--9707}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ILC-based tracking control for linear systems with external disturbances via an SMC scheme}, volume = {22}, year = {2025}, }
• (2025). Robotic grasps of cylindrical and cubic objects via real-time learning-based shape detection. TASE, 22, 9681--9697. (www)

  @article{ , abstract = {Robots grasping objects are critical capabilities in warehouse environments and industrial settings. A robotic grasp generally occurs in a scenario where it is unfeasible for a worker to efficiently complete a tedious task, such as picking food and drink cans (cylinder-shaped) and packaging boxes (cube-shaped). It is worth noting that the tops of cylinders and cubes can be represented by ellipses and rectangles in the two-dimensional (2D) space, respectively. Therefore, a robot can grasp cylinder-shaped and cube-shaped objects by ellipse and rectangle detection. However, it faces the challenge of how to accurately detect cylindrical and cubic objects in real-time for robot grasping. To tackle the above research problem, we propose a grasping system that enables a robot to grasp cylinder-shaped and cube-shaped objects in static and dynamic environments by the proposed ellipse and rectangle detector. An end-to-end learning model is constructed to first incorporate a one-stage detection backbone and then, accommodate the proposed adaptive multi-branch multi-scale net with a designed iterative feature pyramid network, local inception net, and multi-receptive-field feature fusion net to generate object detection recommendations. Employing depth information, the coordinates of detected objects are converted to the 3D space via sampling a series of registered depths and pixels on objects from the live video stream. Comparisons with recent detection methods on the same dataset indicate that the proposed ellipse and rectangle detectors present better performance. Abundant grasping experiments are conducted to illustrate that a robot, empowered by the proposed detector, has the capability of grasping cylindrical and cubic objects in dynamic scenarios. (Video on YouTube, https://youtu.be/KK1OtW6GvL0). Note to Practitioners—This paper is motivated by the problem of how to enable a robot to grasp objects with the basic geometric primitives-ellipses and rectangles in static and dynamic scenarios. Our target is to provide a potential solution for flexible industrial settings in operating moving cylinder-shaped and cube-shaped objects (food and drink cans and packaging boxes) in dynamic scenarios such as conveyors of production lines and logistics lines. We constructed a supervised learning model that can accurately and quickly detect ellipses and rectangles. Through the verification of the comparisons with recent methods and robotic grasping experiments, the behavior of the proposed method can be used in practical applications. In the future, we will deploy this robotic grasping system based on the proposed perception method to grasp food and drink cans and packaging boxes from moving conveyors on production lines and logistics lines.}, archive = {J_TASE}, author = {Huixu Dong and Jiadong Zhou and Haoyong Yu}, doi = {10.1109/TASE.2024.3510592}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9681--9697}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic grasps of cylindrical and cubic objects via real-time learning-based shape detection}, volume = {22}, year = {2025}, }
• (2025). A biological structure-inspired infrastructure health assessment method. TASE, 22, 9667--9680. (www)

  @article{ , abstract = {Structural health monitoring with wireless sensor networks(WSNs) plays an increasingly critical role in modern municipal. While there still remains a gap in getting a comprehensive understanding of complex structural data. Skin diseases can be well diagnosed with modern medical technology, from which similar methods can be learned to improve the intuitiveness and accuracy of structural health monitoring. This work proposes a multi-layered skin-like architecture based method(MSHA), which each layer has its own functions and on the whole presents the disaster situation. This biological structure-inspired architecture has three layers: 1) data substrate layer; 2) connectivity structure layer; 3) pathological manifestation layer, to simulate the three-layer structure of skin. First, a temporal feature extraction method is proposed, which can provide temporal correlations of each node. Second, a dimensional independent spacial feature extraction method is proposed, these first two methods form the connectivity structure layer, which is mainly composed of a spatio-temporal correlation model. Third, a structural health evaluation method for the pathological manifestation layer is proposed to fuse heterogeneous data and calculate multi-granularity structural risk with the features extracted from the connectivity structure layer. The experimental results show that MSHA can achieve not only accurate prediction and intuitive disaster situation, but also low energy consumption and longer lifetime. Note to Practitioners—This paper was motivated by the problem of effectively monitoring the structural health of infrastructure with wireless sensor networks. Existing methods for infrastructure structural health monitoring have space for improvement in maximizing the utilization of correlations between sensor nodes, temporal sequences, and between different sensor types, and fail to produce intuitive results to characterize structural health. In this paper, inspired by the multilayer structure of skin, we propose a new method for analyzing and presenting the structural changes of infrastructure under wireless sensor network data, which fully analyzes the relationship between sensor data in time, space, and type, and is able to predict the structural data in the future period and visually express the current safety situation with size and color information. In this paper, we mathematically describe the model construction method and feature transfer of the constructed intelligent monitoring method. We validate the proposed method using actual tunnel data. The experimental results show that the method is feasible and can accurately indicate the current safety situation in each area while achieving high accuracy in predicting future data, as well as good performance in terms of energy consumption and life cycle. In the future, we will explore optimal strategies for the placement of sensor nodes and improve the convenience and adaptability of the models across multiple scenarios.}, archive = {J_TASE}, author = {Meng Zhang and Gang Li and Bin He and Dongming Zhang and Yu Tian and Bin Cheng}, doi = {10.1109/TASE.2024.3510466}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9667--9680}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A biological structure-inspired infrastructure health assessment method}, volume = {22}, year = {2025}, }
• (2025). Path planning strategy based on principal component federation for multi-agent in connected vehicles. TASE, 22, 9651--9666. (www)

  @article{ , abstract = {Path planning is an important mean to alleviate traffic congestion and reduce travel cost in the internet of vehicles. Existing path planning strategies primarily rely on shortest path or single-agent algorithms. However, they encounter challenges related to global dynamic coordination safety and resource constraints. Therefore, we propose a Multi-Agent dynamic Path Planning strategy based on the Principal component Federation (MA3PF) to address the these challenges. In this strategy, we introduce the Heuristic Multi-Agent Deep Policy Gradient algorithm (H-MADPG), which incorporates future traffic states and leverages shared agent experiences through migration learning. This approach ensures fast convergence and addresses resource limitations through a combination of centralized evaluation and distributed decision-making. Next, we propose the Adaptive Client-based Principal component Federation learning algorithm (ACPFed) for real-time prediction of future traffic flow. This algorithm utilizes principal component analysis to select model parameters and incorporates Bayesian optimization to dynamically determine client weights. These enhancements improve global coordination security and reduce communication overhead. Experimental results on real and simulated datasets demonstrate that our proposed MA3PF method outperforms existing agent algorithms, such as DARP and MAPF. It achieves superior route planning in complex environments, resulting in a reduction of travel distance by 21.90% and time loss by 39.41%. Additionally, MA3PF improves the real-time prediction accuracy of future traffic states while achieving a significant 70% reduction in communication cost. Note to Practitioners—This paper was motivated by the problem of holding multi-vehicle path planning. Existing path planning approaches only consider path planning for a single vehicle, which 1) fail to consider the challenges posed by resource constraints and data security in multi-vehicle planning 2) neglect the influence of future traffic states on path planning. This paper suggests a new dynamic path planning strategy, which incorporates real-time information about future traffic states, ensures data security, and reduces communication overhead using a specially designed ACPFed algorithm, simultaneously leveraging transfer learning and distributed decision-making to address resource constraints. In this paper, we present a mathematical characterization of the multi-vehicle path planning problem, incorporating future traffic states, and analyze the problem to derive an effective heuristic scheme MA3PF. The experimental results on real and simulated datasets demonstrate that the proposed MA3PF achieves significant improvements in terms of route distance and time loss compared to the traditional agent algorithm. In future research, we will explore more complex road environments and incorporate the effects of safety attacks on dynamic agent path planning.}, archive = {J_TASE}, author = {Ruonan Li and Yang Qin and Jie Liu and Lu Zang and Jinlong Li}, doi = {10.1109/TASE.2024.3510432}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9651--9666}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Path planning strategy based on principal component federation for multi-agent in connected vehicles}, volume = {22}, year = {2025}, }
• (2025). A novel hybrid 2.5D map representation method enabling 3D reconstruction of semantic objects in expansive indoor environments. TASE, 22, 9636--9650. (www)

  @article{ , abstract = {This paper presents a novel method for creating a hybrid 2.5D semi-semantic map, merging a 2D geometric map with a sparse 3D object map, specifically designed for expansive indoor environments. The primary motivation behind this method is to tackle the issue of temporal and viewpoint discontinuity in RGB-D observations of individual objects within such environments. Notably, current RGB-D SLAM research mainly focuses on small-scale scenarios, often overlooking this specific issue. To address this challenge, our approach proposes to represent objects with a set of object-specific keyframes, and optimizes the spatial relationships between these keyframes in a deferred offline mode. Another objective is to alleviate the accumulation of trajectory drift, which can adversely impact object association/reidentification in expansive environments. To achieve this, we leverage a 2D pose graph SLAM module that indirectly provides initial poses of the RGB-D sensor, thereby facilitating the construction of the sparse 3D object map. Addressing the challenge arising from the dimensional disparity between the two sub-maps (2D vs 3D), we implement a joint optimization strategy to refine the hybrid map, ensuring accuracy compatibility between them. The effectiveness of our proposed method is validated through experiments conducted in real-world environments, and a time efficiency analysis demonstrates the potential of our algorithm to operate in real-time. Note to Practitioners—The motivation of this paper is to create 3D object maps for expansive indoor environments. The challenge arises from the fact that most existing RGB-D SLAM algorithms are designed for small-scale indoor scenarios, lacking suitability for expansive environments with empty areas. In real expansive environments, the short-range capacity of RGB-D sensors increases the risk of failures in continuously observing objects in empty areas, potentially causing the robot to lose its way. This results in two consequences: 1) discontinuity in the viewpoints of individual objects, challenging the online reconstruction of 3D objects; 2) accumulations of trajectory drift during the robot’s loss complicate the reidentification of an object upon its reappearance. Our approach addresses these issues by concurrently generating a 2D geometric map, effectively mitigating trajectory drift. Additionally, we employ a set of object-specific keyframes to represent object reconstruction, efficiently handling viewpoint discontinuity in object observations. Experimental results demonstrate the applicability of our algorithm to real-world environments beyond room-scale scenarios, closely aligning with the challenges encountered in practical applications. The analysis also indicates the potential for real-time operation, rendering our SLAM algorithm advantageous for practical deployment in robotic systems. An additional benefit is that, apart from the 3D object map, our method creates a 2D geometric map, supporting most basic navigation tasks for industrial robots. In future research, we will further improve the time-consuming process and enhance the efficiency of the algorithm.}, archive = {J_TASE}, author = {Shuai Guo and Yazhou Hu and Mingliang Xu and Jinguo Liu and Zhaojie Ju}, doi = {10.1109/TASE.2024.3510420}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9636--9650}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel hybrid 2.5D map representation method enabling 3D reconstruction of semantic objects in expansive indoor environments}, volume = {22}, year = {2025}, }
• (2025). NEST: Network-energy-stress threat against thermal energy equipment. TASE, 22, 9622--9635. (www)

  @article{ , abstract = {Thermal energy equipment, a critical component for transferring and utilizing thermal energy derived from primary energy sources, is indispensable in Industrial Control Systems (ICSs). With the deep integration of information technologies, ICSs are threatened by new cyber-physical security risks, where the physical systems could be influenced by attacks from the cyber network. While cyber-physical security risks in ICSs have been well studied in various domains, such as power systems and smart structural systems, little attention has been paid to the cyber-physical security of thermal energy equipment. In this paper, we propose a novel cyber-physical threat against thermal energy equipment, namely the Network-Energy-Stress Threat (NEST), which reveals attacks from the cyber network could induce an inhomogeneous distribution of thermal energy, thus causing remarkable thermal stress that can induce physical damage to the thermal energy equipment. From the attacker’s perspective, we propose an inherent vulnerability-based algorithm to explore the threat space of potential attack strategies and find an approximately optimal attack strategy utilizing the nonlinear NEST model. Then, we propose a cyber-physical defense method to detect anomalous states stemming from the NEST. Experimental results on a simulated Solar Power Tower (SPT) plant have validated the existence of the NEST against thermal energy equipment and demonstrated the effectiveness of the proposed algorithm and the proposed detector. Note to Practitioners— This paper is motivated by the practical threat of physical damage to Industrial Control Systems (ICSs) caused by cyber-physical attacks. While cyber-physical security risks in ICSs have been well studied across plenty of types of ICS and physical infrastructures, there is a lack of a framework to describe cyber-physical threats to thermal energy equipment. To address this issue, we propose the Network-Energy-Stress Threat (NEST) to reveal how cyber attacks can induce abnormal thermal stress, causing physical damage to thermal energy equipment—a critical component widely deployed in ICSs. The existence of the NEST has been demonstrated through experimental results on a simulated Solar Power Tower (SPT) plant.}, archive = {J_TASE}, author = {Hanqi Zhou and Yaling He and Ting Liu and Yang Liu and Jinao Shang and Xiangming Wang}, doi = {10.1109/TASE.2024.3510336}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9622--9635}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {NEST: Network-energy-stress threat against thermal energy equipment}, volume = {22}, year = {2025}, }
• (2025). Analytical approximations for the expression of absorption current by ferromagnetic boundary inside the MINI-magnetically shielded cabin. TASE, 22, 9611--9621. (www)

  @article{ , abstract = {The MINI-magnetically shielded cabin (MINI-MSC) offers a near-zero magnetic environment for optically pumped atomic magnetometers (OPMs) to measure the magnetocardiography (MCG) and magnetoencephalography (MEG). However, the MINI-MSC’s confined interior and restricted uniform area make it challenging to capture high-quality bodily signals. Once the external interference surpasses the passive magnetic shielding capacity of the MINI-MSC, the volume and the homogeneity of the uniform region are reduced significantly, leading to the poor quality of the signal collected by OPMs. Therefore, an uniform active magnetic compensation system (UAMCS) composed of bi-planar coils (BCs) is introduced to suppress interference for enlarging the uniform region. The performance of the UAMCS depends on the BCs, but under finite-size ferromagnetic boundaries, approximate magnetic field values that deviate from reality are usually provided by the image method with the inaccurate calculation. To overcome these limitations, an analytical expression of the absorption current is derived to accurately characterize the coupling effect, thereby maintaining the high uniformity performance effectively. A forward analytical model is proposed to characterize magnetic fields, and the equivalent current could be analytically determined, which are put into the target field method (TFM) to express the superimposed the magnetic field. The inhomogeneity errors of BCs using the forward analytical model are 0.39 times of the image method. After the magnetic compensation, the signal-to-noise ratio at 0.1 Hz, 1 Hz and 10 Hz were improved by 39.5 dB, 34.3 dB, and 6.7 dB, respectively, verifying the magnetic noise can be well compensated in the whole target region. Note to Practitioners—The challenge of ensuring high signal quality in MINI-MSC for OPMs is critical in biomedical imaging applications. An forward analytical method is introduced to character the absorbed magnetic field of shielding layers, and an advanced UAMCS utilizing BCs is used for the precise magnetic field compensation to significantly enhance the signal-to-noise ratio. This approach addresses the limitations of traditional image methods, offering improved accuracy in magnetic compensation and enhance the reliability of MEG and MCG tests.}, archive = {J_TASE}, author = {Kangqi Tian and Xu Zhang and Minxia Shi and Jianzhi Yang and Ziyang Shi and Yuqing Hu and Leran Zhang and Haifeng Zhang and Shiqiang Zheng and Gang Liu}, doi = {10.1109/TASE.2024.3509871}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9611--9621}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Analytical approximations for the expression of absorption current by ferromagnetic boundary inside the MINI-magnetically shielded cabin}, volume = {22}, year = {2025}, }
• (2025). Robust neural dynamics for distributed time-varying optimization with application to multi-robot systems. TASE, 22, 9598--9610. (www)

  @article{ , abstract = {This paper develops a robust neural dynamics method for the distributed time-varying optimization problem with time-varying constraints. First, instead of assuming the objective functions and constraints to be static like a majority of the existing research, we consider distributed optimization from a time-varying perspective. Second, by employing the Lagrangian framework, we transform constraints and the concerned objective function that takes the summation of all local functions into a dynamic error function. Third, the robust neural dynamics method is capable of utilizing the time-varying information while solving constrained distributed optimization problems, and meanwhile handling disturbances purely based on its structure, thus lightening the communication and privacy burden. We provide proof of the convergence of the proposed method with activation functions under different disturbances. The comparative results on both illustrative examples and applications validate the efficiency. Note to Practitioners—The highlight of this paper lies in the co-design of time-varying computation and noise-tolerant ability for the proposed distributed optimization method, which could be beneficial to real-world scenarios. Different from existing methods focusing on static objective functions and constraints, we investigate how to find the trajectory formed by the time-varying optimal solutions even with the perturbation of noises. Besides, saturated or even nonconvex activation functions mimicking the synapses of the brain are incorporated into the proposed optimization methods to help accelerate the convergence and counteract the instability induced by noises, without heavily adding to the system’s burden. We also provide theoretical proof to not only guarantee its capability but also guide how to set the involved parameters. This paper proposes a noise-tolerant method for accurately solving distributed time-varying optimization problems and validates its performance on a multi-robot system, which could be suitable for plenty of real-world distributed optimization problems.}, archive = {J_TASE}, author = {Lin Wei and Yongji Guan and Long Jin}, doi = {10.1109/TASE.2024.3508878}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9598--9610}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust neural dynamics for distributed time-varying optimization with application to multi-robot systems}, volume = {22}, year = {2025}, }
• (2025). A novel industrial defect recognition method under label noise based on parallel hybrid penalty network. TASE, 22, 9584--9597. (www)

  @article{ , abstract = {Defect recognition optimizes industrial manufacturing processes by monitoring the health of equipment and structures. The success of current deep learning (DL)-based defect recognition methods relies on the assumption that the samples are label-noise free, yet the low-quality of industrial data inevitably introduces label noise, resulting in a substantial drop in recognition performance. To tackle this problem, this paper develops a parallel hybrid penalty network (PHL-Net). It solves two essential issues in defect recognition under label noise: low sample utilization and difficulty in recognizing the ambiguous sample in the decision boundary region, so that the industrial defect recognition performance can be greatly enhanced under label noise. PHL-Net consists of three stages: data selection (DS) for explicitly selecting clean and noisy data; parallel feature mining (PFM) for improving the sample utilization by label-driven mining and distance-driven mining; hybrid penalty learning (HPL) to improve decision reliability to label noise, thereby the ambiguous samples in the decision boundary region can be accurately recognized. In PHL-Net, PFM and HPL are co-optimized to facilitate data selection. In turn, reliable data selection prevents the model from overfitting label noise. Three groups of experimental results on real-world industrial datasets MPV-W and NEU-CLS have demonstrated the effectiveness of our PHL-Net over state-of-the-art alternatives. Note to Practitioners—Defect recognition plays a vital role in industrial production, ensuring product quality and efficiency. When training deep learning-based defect recognition models, having datasets with precise annotations is essential. Unfortunately, label errors, known as label noise, are common when annotating industrial datasets due to the need for domain expertise, subjective judgments, and confusing samples. This makes it challenging to create a large-scale and accurate industrial dataset. To address this issue, our goal is to develop a robust model that can achieve optimal recognition performance even on datasets with noisy labels. On this basis, this paper proposes a parallel hybrid penalty network (PHL-Net). It improves recognition performance under label noise by improving sample utilization and recognizing ambiguous samples near the decision boundary. In experiments, we utilize the noise transition matrix to introduce label noise for MPV-W and NEU-CLS. The results of the experiments demonstrate that our PHL-Net outperforms the comparison methods on industrial datasets, highlighting the practical significance of our approach. This provides inspiration and reference for industrial defect recognition under label noise.}, archive = {J_TASE}, author = {Xiaoyuan Liu and Jinhai Liu and Mingrui Fu and Huaguang Zhang and Fengyuan Zuo and Hang Xu}, doi = {10.1109/TASE.2024.3508774}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9584--9597}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel industrial defect recognition method under label noise based on parallel hybrid penalty network}, volume = {22}, year = {2025}, }
• (2025). Model-based predictive impedance variation for obstacle avoidance in safe Human–Robot collaboration. TASE, 22, 9571--9583. (www)

  @article{ , abstract = {Human-robot collaboration (HRC) in manufacturing environments requires that physical safety can be guaranteed. Control methods that implicitly regulate the interaction forces between a controlled robot and its environment, such as impedance control, are often used for safety in HRC. However, these methods could be complemented by restricting the robot operational space for additional safety guarantees. In this context, obstacle avoidance might benefit from considering a prediction of the controlled-robot motion and/or the behavior of the human collaborator. To this end, we proposed to include linearized Safety Control Barrier Functions (SCBFs) in a linear Model Predictive Control (MPC) strategy for robot impedance variation online. The convex optimization problem that was obtained from our proposal presented two advantages compared to nonlinear MPC alternatives. First, optimality was ensured in our method under linearity assumptions on human guidance and linearized robot dynamics, whereas a controller synthesized by nonlinear MPC strategies would depend on the fundamental characteristics of the problem. Second, our method enabled implementation at a faster control frequency, thus allowing a rapid adaptation to changes occurring in the robot environment. Finally, experimental validation was performed using a Franka Emika Panda robot in a human-robot collaborative scenario, and the stability of the method was shown using Lyapunov theory. Note to Practitioners—Modern-day industrial manufacturing environments are characterized by collaboration between human operators and robot manipulators. In this scenario, where humans and robots share workspace, physical safety is required. This research aims to improve safety in human-robot collaboration by proposing a novel robot control strategy. In our approach, the interaction forces between the controlled robot and its environment were regulated implicitly using impedance control, to allow, among other interactions, that an operator could manually guide the robot. Then, obstacle avoidance was included to modify the robot impedance behavior for restricting undesired collisions with, for example, the operator head, while ensuring stability of the method. Our main contribution is that the proposed formulation allows to consider a prediction of the robot motion and/or the operator behavior for robot obstacle avoidance. It was shown in experiments with a real robot that adding prediction capabilities reduced the risk of undesired collisions, while also decreasing the robot trajectory error. Moreover, the method could be implemented at a fast rate so that the robot could react rapidly to changes in its environment. Also, this implementation allowed to achieve a minimal variation with respect to the nominal impedance behavior of the robot. To conclude, this method is intended for scenarios where a robot is required to interact with its, possible restricted, environment: for example, a robot with a drill attached to its end-effector that is being guided to modify its trajectory, but where the operator should not be allowed to accidentally be harmed; or a robot performing a polishing task where a section of the polished object should remain unpolished. Therefore, using the proposed robot control strategy, a possible extension of this research would be to provide an improved prediction of the human operator intention depending on the desired robotic task and the role of the operator.}, archive = {J_TASE}, author = {Julian M. Salt Ducaju and Björn Olofsson and Rolf Johansson}, doi = {10.1109/TASE.2024.3508718}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9571--9583}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model-based predictive impedance variation for obstacle avoidance in safe Human–Robot collaboration}, volume = {22}, year = {2025}, }
• (2025). Synchronization for coupled stochastic nonlinear systems with time-varying and non-strongly connected structure. TASE, 22, 9559--9570. (www)

  @article{ , abstract = {This article studies the synchronization problem in a class of coupled stochastic nonlinear systems featuring time-varying and non-strongly connected coupling structures. In order to solve the problem posed by non-strong connectivity, an innovative hierarchical method is proposed, which enables us to study the synchronization in a layered manner. For achieving synchronization, actual controllers are constructed through the application of new coordinate transformations and the design of virtual controllers. Furthermore, by leveraging graph theory techniques and Lyapunov method, the mean square exponential synchronization criterion is derived. Finally, the theoretical results are applied to coupled Chua’s circuits, and their effectiveness is confirmed via numerical simulations. Note to Practitioners—This article is motivated by the synchronization results of stochastic strict-feedback nonlinear systems. At present, most studies concentrate on simple single-input single-output stochastic nonlinear systems, yet interconnected multi-input multi-output stochastic nonlinear systems hold greater significance in real-world scenarios. This research expands its focus to the coupling structures between interconnected stochastic nonlinear systems, which are not only time-varying but also allow non-strongly connected topological forms. This extension significantly enhances the practical application of traditional single-input single-output nonlinear systems. Additionally, considering the non-strongly connected nature, an innovative hierarchical approach is proposed. This method allows for layer-by-layer synchronization analysis, concurrently broadening the application of graph theory in this field. In the process of ensuring the synchronization of each layer, successful construction of fourth-order Lyapunov functions and actual controllers was achieved through appropriate variable substitutions and virtual controllers design. Overall, this paper provides a new perspective and methodology for the synchronization analysis and research of interconnected stochastic nonlinear systems.}, archive = {J_TASE}, author = {Wanying Guo and Tianhang Chang and Wenxue Li and Xing Liu}, doi = {10.1109/TASE.2024.3508019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9559--9570}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synchronization for coupled stochastic nonlinear systems with time-varying and non-strongly connected structure}, volume = {22}, year = {2025}, }
• (2025). Scheduling single-arm cluster tools with an equipment front-end module subject to wafer residency time constraints. TASE, 22, 9544--9558. (www)

  @article{ , abstract = {Semiconductor manufacturing widely employs cluster tools that comprise three critical components: a vacuum module (VM), a loadlock module (LLM), and an equipment front-end module (EFEM). In operating such tools, LLM plays a pivotal role. Additionally, wafer fabrication in VM faces strict wafer residency time constraints, making the coordination among the modules particularly challenging. This paper addresses the cyclic scheduling problem of single-arm cluster tools with EFEM and wafer residency time constraints. It focuses on the tool efficiency in module cooperation and schedulability. We explore cooperative strategies for the robot operating within VM to access loadlocks, elucidating the influence of EFEM and LLM on VM. Based on these strategies, we establish necessary and sufficient conditions for the existence of a feasible periodic schedule, providing a foundational basis for the systematic analysis. For schedulable scenarios, we derive an efficient algorithm to identify the feasible and optimal schedule in terms of tool cycle time. The performance and efficiency of the proposed algorithm are validated through experimental verification. Note to Practitioners—In wafer fabs, single-arm cluster tools with EFEM are widely adopted. Studies that neglect the impact of EFEM on scheduling cluster tools are often inapplicable to real-world scenarios. Wafer residency time constraints are essential for maintaining wafer quality. As VM and EFEM operate in different pressure environments, the cooperation via LLM makes it challenging to meet residency time constraints. To address this problem, this paper analyzes the impact of cooperative strategies on the tool performance, and the influence of EFEM and LLM on VM as well. Under these strategies, we examine the schedulability conditions, helping engineers understand how processing parameters affect scheduling feasibility. If a feasible schedule exists, an efficient algorithm is derived to find an optimal schedule. Experimental results show the efficiency of the proposed algorithm, making it suitable for embedding into cluster tool controllers for efficient implementation. For cases that are not schedulable, with the results obtained, this work can provide guidance of how to redesign the process such that feasibility can be ensured.}, archive = {J_TASE}, author = {BaoYing Huang and NaiQi Wu and Yan Qiao and WeiWen Guo}, doi = {10.1109/TASE.2024.3507854}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9544--9558}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling single-arm cluster tools with an equipment front-end module subject to wafer residency time constraints}, volume = {22}, year = {2025}, }
• (2025). A set-theoretic adaptive current control design for grid-following inverter-based resources to tackle practically non-ideal control inputs. TASE, 22, 9528--9543. (www)

  @article{ , abstract = {Three-phase grid-following (GFL) inverter-based resources (IBRs) play a vital role as an interface for integrating renewable energy resources and flexible loads, such as electric vehicles, into the power grid. This paper introduces a novel set-theoretic adaptive control scheme for the primary control of three-phase GFL IBRs, designed to mitigate the impacts of uncertainties or non-ideal conditions affecting the control layer. These uncertainties will risk losing the stability and intended operation of three-phase GFL IBRs by potentially influencing the control commands transmitted to pulse width modulators. In order to address this issue, this study proposes an add-on control signal generated through an adaptive architecture to retrofit the existing (pre-designed) state feedback controller of GFL IBRs. As the name implies, this retrofit control strategy entails upgrading or modifying the existing feedback control instead of completely replacing it. The proposed control scheme is based on a set-theoretic adaptive controller design that employs generalized restricted potential functions. A notable aspect of this framework is its ability to ensure that the reference tracking error bound remains below a user-defined threshold, making it “computable” by providing the control design parameters. The stability of the closed-loop system and the approximate reference tracking performance of the proposed control scheme for GFL IBRs are validated through a theoretical analysis employing the Lyapunov theory. Simulation-based and experimental results further confirm the efficacy of the proposed GFL IBR controller. Note to Practitioners—Given the strive to focus on renewables-intensive and modern power grids, inverters must exhibit enhanced intelligence and versatility to accommodate various functionalities. However, uncertainties or non-ideal conditions originating from diverse sources significantly threaten the optimal operation of inverter-based resources. These uncertainties introduce errors into the control loop of grid-following inverter-based resources, potentially compromising their stability and performance. In order to tackle this compelling challenge, this paper proposes a novel set-theoretic adaptive current control scheme for grid-following inverter-based resources using an add-on control signal. The aim is to mitigate the adverse effects of uncertainties affecting the control commands in grid-following inverter-based resources. By incorporating the additional control signals into the feedback controller, reference tracking is assured despite uncertainties. This approach offers a more cost-effective solution by enhancing the existing feedback control instead of entirely replacing it. Lyapunov stability theory provides a theoretical framework for analyzing stability and ensuring the uniform boundedness of output trajectories in grid-following inverter-based resources. Simulation and experimental results confirm the feasibility and effectiveness of the proposed approach in mitigating the uncertainties affecting the control commands.}, archive = {J_TASE}, author = {Mahmood Jamali and Mahdieh S. Sadabadi and Masoud Davari}, doi = {10.1109/TASE.2024.3507616}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9528--9543}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A set-theoretic adaptive current control design for grid-following inverter-based resources to tackle practically non-ideal control inputs}, volume = {22}, year = {2025}, }
• (2025). Practical finite-time compliant control for horizontal pneumatic artificial muscle systems under force-sensorless reflecting. TASE, 22, 9515--9527. (www)

  @article{ , abstract = {Pneumatic artificial muscle (PAM) actuators have passive compliance and vibration absorption capabilities, adapting to high-intensity human-robot interaction movements. Unfortunately, the asymmetric hysteresis of PAMs is prone to produce motion delays and control inaccuracy, and anti-disturbance control is not friendly when applied in exoskeleton robots. Also, most of existing studies on active compliant control are overly reliant on bulky force-sensing, which is limited by sampling accuracy and communication rate. Therefore, it is still a challenge to realize compliant motions of PAM systems, while ensuring the rapid convergence of output signals. To this end, a new practical finite-time compliant controller is designed in this paper, which realizes satisfactory tracking control of horizontal PAM systems. Specifically, the external contact force is estimated by an improved adaptive law, instead of sensor feedback, thus decreasing noise effects. Meanwhile, the proposed controller ensures that the output tracking error converges quickly within known finite time, while reducing computational complexity. In particular, the desired trajectory is updated in real time by a modified admittance model, so as to achieve motion compliance and interaction safety. Compared with the literature, it is the first attempt to provide a compliant control solution for horizontal PAM systems without force feedback information, and ensures the practical finite-time convergence of the output tracking error. The rigorous stability analysis is presented, and the reliability of the proposed method is verified by hardware experiments. Note to Practitioners—Owing to light weight and flexibility, pneumatic artificial muscle (PAM) actuators can better meet the growing demands of human-machine cooperation tasks. In areas such as power assist and rehabilitation training equipment, it is necessary to ensure that robots can properly adjust the motion trajectory according to the applied force, so as to replace unbearable “hard contact” with more dexterous “compliant interaction”. Inspired by this, a new compliant control method is designed in this paper for horizontal PAM systems, which realizes the online adjustment of desired motion trajectories according to contact forces, and ensures the rapid convergence of the tracking error within known finite time. In particular, the proposed method improves an efficient contact force estimation solution, avoiding bulky and expensive force-sensing, while reducing noise effects. Compared with existing results, this paper for the first time presents a compliant control method without sensor feedback of horizontal PAM systems, and improves the motion rapidity, accuracy, and stability in practical tracking and training cases. Meanwhile, rigorous stability analysis is provided by Lyapunov techniques, and the effectiveness of the proposed method is verified by experiments on a self-built PAM testbench. In the future, we will try to apply the proposed method to limb functional training scenarios, aiming to expand its practical prospects and increase efficiency.}, archive = {J_TASE}, author = {Gendi Liu and Shuzhen Diao and Zhuoqing Liu and Xinlin Zhang and Xue Xiao and Song Men and Ning Sun}, doi = {10.1109/TASE.2024.3507281}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9515--9527}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Practical finite-time compliant control for horizontal pneumatic artificial muscle systems under force-sensorless reflecting}, volume = {22}, year = {2025}, }
• (2025). Application of adaptive parallel fast marching method in automatic submarine cable path planning. TASE, 22, 9498--9514. (www)

  @article{ , abstract = {Submarine optical fiber communication cables (subsequently referred to as submarine cables) form the backbone of the Internet’s infrastructure. Damage to these cables can precipitate Internet outages with far-reaching socio-economic impacts. The prevailing practice of manual cable routing is laborious, considering the thousands of kilometers these cables span. It also fails to strike an optimal balance between cost and risk due to its lack of scalability and precision. The Fast Marching Method (FMM), a non-iterative, precise numerical approach capable of solving the Eikonal equation, offers a viable alternative by optimizing the required path between source and destination, considering a summary objective function of costs and risk factors. An interpretation of its solution signifies the optimum value of the objective function between a starting point and all other points. However, the sequential nature of the FMM algorithm suffers from computational limitations and impedes direct parallelization. In this study, we introduce an Adaptive Parallel FMM (APFMM), an innovative approach utilizing adaptive domain decomposition and dynamic multi-resolution analysis. This scalable and widely applicable method can overcome the limitations of existing methods and achieve planning of high-precision, ultra-long-distance (over 14,000 km) submarine cable routes over the Earth’s surface. Simulated experiment results corroborate that APFMM effectively overcomes the computational challenges posed by the sequential FMM when dealing with large datasets. Additionally, it reduces the running time by more than 81% compared to the traditional parallel FMM. This marks a substantial advancement in facilitating efficient, automated, high-precision planning for long-distance submarine cable paths. Note to Practitioners—This paper introduces APFMM, a novel technique based on adaptive domain decomposition and multi-resolution analysis, facilitating high-precision, ultra-long-distance (over 14,000 km) submarine cable path planning. Results from simulated experiments show that APFMM not only overcomes the computational constraints associated with the sequential FMM for large datasets but also cuts the running time by more than 81% relative to the conventional parallel FMM. This breakthrough improvement holds significant implications for practitioners in submarine cable design and construction. With the use of APFMM, designers can plan and optimize cable paths more efficiently, thereby lowering cabling costs and enhancing network resilience. Furthermore, the application of APFMM is not limited to submarine cable path planning and can be employed in other domains involving large-scale data processing and complex path planning, such as electricity cables, gas pipelines, and transportation route planning. While our focus in this paper is primarily on submarine cable path planning, we anticipate practitioners extending the application of APFMM to other use cases, realizing broader utility and benefits.}, archive = {J_TASE}, author = {Xinyu Wang and Zengfu Wang and Moshe Zukerman}, doi = {10.1109/TASE.2024.3507115}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9498--9514}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Application of adaptive parallel fast marching method in automatic submarine cable path planning}, volume = {22}, year = {2025}, }
• (2025). Multimodal fish feeding intensity assessment in aquaculture. TASE, 22, 9485--9497. (www)

  @article{ , abstract = {Fish feeding intensity assessment (FFIA) aims to evaluate fish appetite changes during feeding, which is crucial in industrial aquaculture applications. Existing FFIA methods are limited by their robustness to noise, computational complexity, and the lack of public datasets for developing the models. To address these issues, we first introduce AV-FFIA, a new dataset containing 27,000 labeled audio and video clips that capture different levels of fish feeding intensity. Then, we introduce multi-modal approaches for FFIA by leveraging the models pre-trained on individual modalities and fused with data fusion methods. We perform benchmark studies of these methods on AV-FFIA, and demonstrate the advantages of the multi-modal approach over the single-modality based approach, especially in noisy environments. However, compared to the methods developed for individual modalities, the multimodal approaches may involve higher computational costs due to the need for independent encoders for each modality. To overcome this issue, we further present a novel unified mixed-modality based method for FFIA, termed as U-FFIA. U-FFIA is a single model capable of processing audio, visual, or audio-visual modalities, by leveraging modality dropout during training and knowledge distillation using the models pre-trained with data from single modality. We demonstrate that U-FFIA can achieve performance better than or on par with the state-of-the-art modality-specific FFIA models, with significantly lower computational overhead, enabling robust and efficient FFIA for improved aquaculture management. To encourage further research, we have released the AV-FFIA dataset, the pre-trained model and codes at https://github.com/FishMaster93/U-FFIA. Note to Practitioners—Feeding is one of the most important costs in aquaculture. However, current feeding machines usually operate with fixed thresholds or human experiences, lacking the ability to automatically adjust to fish feeding intensity. FFIA can evaluate the intensity changes in fish appetite during the feeding process and optimize the control strategies of the feeding machine to avoid inadequate feeding or overfeeding, thereby reducing the feeding cost and improving the well-being of fish in industrial aquaculture. The existing methods have mainly exploited single-modality data, and have a high sensitivity to input noise. Using video and audio offers improved chances to address the challenges brought by various environments. However, compared with processing data from single modalities, using multiple modalities simultaneously often involves increased computational resources, including memory, processing power, and storage. This can impact system performance and scalability. To address these issues, we focus on the efficient unified model, which is capable of processing both multimodal and single-modal input. Our proposed model achieved state-of-the-art (SOTA) performance in FFIA with high computational efficiency.}, archive = {J_TASE}, author = {Meng Cui and Xubo Liu and Haohe Liu and Zhuangzhuang Du and Tao Chen and Guoping Lian and Daoliang Li and Wenwu Wang}, doi = {10.1109/TASE.2024.3507098}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9485--9497}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multimodal fish feeding intensity assessment in aquaculture}, volume = {22}, year = {2025}, }
• (2025). Robust MPC with mode-dependent dissipative performance for non-homogeneous MJSs subjected to fuzzy switching mechanism. TASE, 22, 9473--9484. (www)

  @article{ , abstract = {This article aims to address the issue of model predictive control with optimization predictive dynamics for a class of non-homogeneous Markov jump systems. The principal concept is to construct a robust predictive control strategy by developing a fuzzy switching Markov process and a mode-dependent dissipative index to relax performance constraints. In particular, intending to depict the effects of interference on the Markov process, a high-level fuzzy switching rule, which adopts the interference factors as premise variables, is firstly proposed to orchestrate the resulting time-varying transition probabilities. This further constructs a generalized framework that takes time-invariance and high-level deterministic switching of transition probabilities as special cases. Secondly, to surmount the difficulty of acquiring relaxed stability criteria in the designing of controller and improving control performance, a novel dissipative index with mode dependence is proposed as a means of characterizing the varying dynamic system performance. What’s more, taking into account the optimization predictive dynamics-based model predictive control with bounded disturbance, a dissipation-type cost function is developed to promote the analysis of terminal constraint, dissipative performance and stability, which is distinct from $H_{2}/H_{\infty } $ type. Furthermore, the offline and online optimization algorithms are presented for achieving desired performance. Finally, the practicabilities and advantages of the theoretical results are illustrated by a modified DC motor device system model. Note to Practitioners—Markov jump systems are useful for modelling systems with random changes in parameters or structures, making them valuable in practical engineering applications, for instance the continuous stirred tank heater, the HiMAT vehicle and the F-404 aircraft. However, obtaining an accurate transition probability matrix can be challenging due to the influence of the work environment. This article proposes a high-level fuzzy switching mechanism governed Markov process to achieve more practical modeling. Additionally, a mode-dependent dissipative index is proposed to characterize time-varying performance, considering that system performance may vary with system parameters. Model predictive control is widely used in engineering applications due to its ability to handle uncertainty and hard constraints effectively. This article presents an optimization predictive dynamics-based model predictive control scheme to reduce computational pressure and constructs a dissipative-type cost function to promote interference handling analysis. Finally, the application effect of the proposed method was verified through DC motor device system model.}, archive = {J_TASE}, author = {Jian Wang and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3507085}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9473--9484}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust MPC with mode-dependent dissipative performance for non-homogeneous MJSs subjected to fuzzy switching mechanism}, volume = {22}, year = {2025}, }
• (2025). Distributionally robust chance-constrained line planning for railway systems under passenger demand uncertainty. TASE, 22, 9457--9472. (www)

  @article{ , abstract = {In a railway network, making a line plan is a critical optimization problem that determines the daily transportation capacity of the network, aiming at aligning it with passenger demand while minimizing the operational cost. However, the inherent uncertainty in passenger demand often makes the determined line plan infeasible to cover. Meanwhile, additional adjustments to the line plan due to fluctuating demand in the complex railway system can cause unexpected costs. To obtain a robust line plan, we propose a distributionally robust chance-constrained (DRCC) line planning model based on a type- $\infty $ Wasserstein ambiguity set, aiming to generate a line plan that remains feasible with a pre-specified probability while accommodating a given distribution deviation tolerance. We present an equivalent tractable reformulation for the proposed DRCC model by explicitly characterizing the worst-case probability distribution. Furthermore, we develop valid inequalities and a warm start strategy tailored to this model to enhance computational efficiency. The proposed model and solution method are validated through numerical experiments conducted on the Wuhan-Guangzhou high-speed railway corridor. Results demonstrate the effectiveness of the solution acceleration techniques and underscore the advantage of the DRCC model over the robust optimization model and conventional chance-constrained model in reducing operational costs. Note to Practitioners—The motivation for this paper stems from the need to provide a daily line plan for the railway to meet the uncertain passenger demand with a pre-specified probability. We assume that an estimated probability distribution of the demand is available and characterized by finite scenarios with weights, but we do not need the estimation to be perfect/accurate. We propose a distributionally robust chance-constrained line planning model to maintain robustness under the deviation/inaccurate probability distribution. To solve the model efficiently, we develop a warm start method to quickly find a partial initial solution and design valid inequalities where adding them as constraints can shorten the solving procedure by reducing the feasible region of the model’s continuous relaxation. The produced line plan requires less operational cost than the robust optimization model to cover the same amount of demand by the benefits of leveraging the distributional information.}, archive = {J_TASE}, author = {Linyu Liu and Wanlu Yang and Shiji Song and Yuli Zhang}, doi = {10.1109/TASE.2024.3506988}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9457--9472}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributionally robust chance-constrained line planning for railway systems under passenger demand uncertainty}, volume = {22}, year = {2025}, }
• (2025). Asymptotic adaptive tracking control based on disturbance observer for non-strict feedback nonlinear systems. TASE, 22, 9449--9456. (www)

  @article{ , abstract = {This paper proposes an innovative composite anti-disturbance asymptotic tracking control strategy for a class of nonlinear non-strict feedback systems facing unknown nonlinearities and external disturbances. Based on the separation principle, a new adaptive learning disturbance observer (ALDO) is first introduced, designed to accurately estimate the unknown nonlinear dynamics and time-varying disturbances within the system. This ALDO possesses online learning capabilities, enabling it to estimate the unknown upper bound of the disturbance derivative in real time, thus significantly reducing reliance on prior knowledge of the disturbance derivative and ensuring finite-time stability of the disturbance estimation error system. Furthermore, utilizing adaptive backstepping technology, a virtual controller is designed to compensate for mismatching disturbances within the closed-loop system. Additionally, a composite controller is developed to effectively compensate for disturbances and guarantee asymptotic convergence of the tracking control error to zero while maintaining signal boundedness within the system. Finally, the effectiveness and practicality of this control strategy are verified through a practical application case involving an electromechanical system. Note to Practitioners—This article investigates a class of nonlinear non-strict feedback systems affected by multi-disturbances, where the unknown nonlinear dynamics are functions of the full-state variable, rendering traditional adaptive methods inapplicable due to the potential for the algebraic loop problem. In complex engineering environments, where various external disturbances adversely impact system stability and performance, and prior knowledge of such disturbances is elusive, the disturbance observer-based composite control strategy explored in this paper stands out as crucial. The proposed strategy adeptly mitigates or eliminates disturbances via feedforward and feedback channels, enabling asymptotic tracking of the system while safeguarding its stable performance and reliability, holding promising applications in engineering domains like power systems, robot control, and traffic signal systems.}, archive = {J_TASE}, author = {Boyang Zhao and Xiuming Yao}, doi = {10.1109/TASE.2024.3506931}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9449--9456}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asymptotic adaptive tracking control based on disturbance observer for non-strict feedback nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). LSPT-D: Local similarity preserved transport for direct industrial data imputation. TASE, 22, 9438--9448. (www)

  @article{ , abstract = {Accurate imputation of missing data is pivotal in real-world industrial applications. Traditional direct imputers, which utilize basic statistics to replace missing elements, offer a practical solution but struggle to adapt to the complex patterns in industrial data, leaving a gap in the research landscape. This study explores the untapped potential of direct imputers, enhancing their adaptability and capacity to handle complex patterns in industrial data through optimal transport (OT) theory, with a focus on preserving local sample-wise similarity as an exemplar. To these ends, we construct a Local Similarity Preserved Transport (LSPT) problem, with a solution algorithm based on the Frank-Wolfe technique to compute transport cost. Subsequently, we propose the LSPT-D framework, which employs the transport cost of LSPT for distribution matching, directing the gradient flow to the missing data points to update the imputations directly. This strategy maintains local similarity throughout the imputation process thereby enhancing the overall imputation quality. Our experiments demonstrate that LSPT-D outperform various baselines in industrial missing data imputation. Note to Practitioners—Accurate missing data imputation is essential for enhancing the reliability of data analytics and reducing decision-making risks in industrial automation. This study introduces LSPT-D, a non-parametric imputation technique based on OT technology. Unique to LSPT-D is its ability to preserve local similarity during the imputation process, rendering it particularly advantageous for datasets with varying operational phases and load conditions. In industrial applications, LSPT-D not only significantly improves imputation quality compared to various baseline methods but also maintains modest running costs. Additionally, it serves as an exemplar for developing OT-based imputation strategies that capitalize on the inherent properties of data to improve imputation performance. However, LSPT-D operates under the independent and identically distributed assumption and is thus best applied in scenarios where temporal dependencies, such as trends and seasonality, are minimal or have been previously neutralized.}, archive = {J_TASE}, author = {Hao Wang and Xinggao Liu and Zhaoran Liu and Haozhe Li and Yilin Liao and Yuxin Huang and Zhichao Chen}, doi = {10.1109/TASE.2024.3506835}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9438--9448}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {LSPT-D: Local similarity preserved transport for direct industrial data imputation}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance control for nonlinear systems with sensor faults and actuator hysteresis. TASE, 22, 9426--9437. (www)

  @article{ , abstract = {This paper investigates the prescribed performance control (PPC) issue in nonlinear systems involving actuator hysteresis, sensor fault, and unavailable states. Firstly, an inverse model is formulated by leveraging the properties of the analytic inverse of the asymmetric Prandtl-Ishlinskii (PI) model. Subsequently, the approximation capabilities of neural networks are harnessed to address entirely unknown nonlinear terms. Next, a novel adaptive observer is constructed to estimate the system states, hysteresis parameters, and sensor faults information. Moreover, the PPC strategy in conjunction with a shifting function, ensures that the tracking error converges to a specified region. This proposed approach effectively mitigates hysteresis and fault, rendering all signals of the closed-loop system semi-globally uniformly ultimately bounded (SGUUB). Finally, the effectiveness of the algorithm is verified through a simulation and an experiment. Note to Practitioners—Piezoelectric ceramic micro-positioning platforms find extensive application in high-precision positioning technology areas like fluid machinery, automated driving, and automated precision machining production due to their benefits of quick response speed, high displacement resolution, good stiffness, favorable frequency characteristics, and high positioning accuracy. Despite the advantages of piezoelectric ceramic actuators, hysteresis issues can arise, leading to significant positioning errors. Therefore, implementing a suitable control scheme is essential to address this challenge. Additionally, in the automotive and industrial fields, sensor faults can easily happen as a result of different external factors, potentially resulting in serious accidents. Consequently, effective fault management holds significant importance. The performance control method presented in nonlinear systems aims to ensure convergence time, overshoot, convergence accuracy, and other parameters within the predefined range, thereby guaranteeing optimal system performance. This strategy can be extended and applied to piezoelectric driven positioning systems.}, archive = {J_TASE}, author = {Yu Li and Zhengqiang Zhang and Yurong Li and Xue-Jun Xie}, doi = {10.1109/TASE.2024.3506824}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9426--9437}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance control for nonlinear systems with sensor faults and actuator hysteresis}, volume = {22}, year = {2025}, }
• (2025). Transmission line detection through auxiliary feature registration with knowledge distillation. TASE, 22, 9413--9425. (www)

  @article{ , abstract = {The periodic inspection of transmission lines ensures a stable power supply across various regions. Deep learning methods, particularly those based on multimodal approaches, have made significant advancements in this field. Current multimodal fusion algorithms do not prioritize the integration of supplementary information from other modal sources. Moreover, these methods frequently depend on substantial parameterization to achieve optimal performance, severely hampering deployment on mobile devices. To enhance the supportive role of the auxiliary modality and minimize model parameters, we proposed the knowledge distillation-based auxiliary feature registration network (AFRNet-S*) for RGB-T transmission line detection (TLD). The method incorporates an auxiliary feature registration module during fusion, utilizing unique complementarity between primary and auxiliary feature modalities for precise feature registration. For knowledge distillation, we devised response-mimicking distillation, semantic supplementary distillation, and cross-view feature polymerization distillation tailored for the student network (AFRNet-S, without knowledge distillation) to achieve model compression. AFRNet-S can learn comprehensive final spatial, precise semantic, and complementary feature integration information from AFRNet-T through these distillation methods. Extensive experiments conducted on an industrial TLD dataset demonstrate that the proposed AFRNet-S* (AFRNet-S with knowledge distillation) outperforms the existing state-of-the-art methods, thereby demonstrating its generality and effectiveness. Our AFRNet-S*, compared to AFRNet-T, has a reduced parameter count from 44.92M to 16.86M and a decrease in FLOPs from 21.13G to 6.41G. This further optimizes the deployment of our network on edge devices, such as unmanned aerial vehicles (drones), in the future. The code and results of our approach are available at https://github.com/WangYuSenn/AFRNet. Note to Practitioners—This study introduces the AFRNet for TLD, utilizing knowledge distillation (KD) techniques. We achieved a better balance between model performance and parameter count by integrating the KD task into the TLD task, facilitating easier deployment on mobile devices. The proposed AFRNet-S* effectively balances model performance and parameter count, ensuring efficient performance of the TLD task. Addressing RGB-T TLD-specific challenges remains a crucial area for future research. For instance, in scenarios where a target occupies a small area within an image, removing excessive background noise is essential to enhance the efficiency of target localization. Furthermore, we aim to utilize AFRNet-S* on drones for transmission line inspection in the future to address practical application challenges.}, archive = {J_TASE}, author = {Yusen Wang and Wujie Zhou and Xiaohong Qian}, doi = {10.1109/TASE.2024.3506617}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9413--9425}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transmission line detection through auxiliary feature registration with knowledge distillation}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning control of double-layer markov jump systems with PDT-switched transition probabilities. TASE, 22, 9404--9412. (www)

  @article{ , abstract = {This paper proposes a data-based reinforcement Q-learning control method to efficiently solve the $H_{\infty }$ control problem for discrete-time double-layer Markov jump systems (DLMJSs). The transition probabilities are considered to be piecewise-homogeneous and governed by a persistent dwell-time (PDT) switching scheme. First, a model-based stability and performance optimization criterion for DLMJSs is derived. On this basis, a novel reinforcement Q-learning method in the form of linear matrix inequality is presented, such that optimal disturbance attenuation level and $H_{\infty }$ control policy can be calculated from the collected data. At the core of the approach is a two-step learning design associated with the Q-function. The first step is to estimate the kernel matrix by solving the minimization problem. The second step is to find an improved control policy based on the estimated kernel matrix. Furthermore, the rigorous analysis of the stability of closed-loop DLMJSs and the convergence of the designed algorithm are presented, ensuring that closed-loop systems achieve mean-square exponential stability with an optimal disturbance attenuation level. Finally, the applicability and effectiveness of the proposed method in engineering applications are verified by an armature controlled DC motor system model. Note to Practitioners—Due to the complexity of engineering environment, the $H_{\infty }$ control of DLMJSs has garnered extensive attention. Nowadays, DLMJSs find widespread application in various domains including tidal turbine systems, robot manipulators, and smart grids. It should be noted that existing $H_{\infty }$ control methods for DLMJSs typically necessitate precise system dynamics and assume a prescribed disturbance attenuation level. However, in practical applications, these assumptions are difficult to satisfy. In view of these facts, we employ the reinforcement Q-learning control method to solve the $H_{\infty }$ control problem of unknown DLMJSs with PDT-switched transition probabilities. Moreover, the designed control method can achieve both the optimal disturbance attenuation level and the optimal $H_{\infty }$ control policy of the studied systems, providing a feasible strategy for improving control performance in industrial applications.}, archive = {J_TASE}, author = {Hao Shen and Jiacheng Wu and Yun Wang and Huaicheng Yan and Shengyuan Xu}, doi = {10.1109/TASE.2024.3506592}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9404--9412}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning control of double-layer markov jump systems with PDT-switched transition probabilities}, volume = {22}, year = {2025}, }
• (2025). Monitor-based supervisory control of labeled petri nets under sensor attacks. TASE, 22, 9388--9403. (www)

  @article{ , abstract = {In this paper we investigate a supervisory control problem of discrete event systems under attacks. Specially, we consider a type of sensor deception attacks, called replacement attacks, under which the intruder confuses the observation of events by substituting the occurrence of an observation with another. We use labeled Petri nets as the reference formalism to model a discrete event system and represent control specifications in terms of generalized mutual exclusion constraints (GMECs). The concept of a monitor function is proposed to describe the satisfiability of GMECs given an observation by counting the number of occurrences of each label. Due to the existence of attacks, some labels generated by a plant are prone to be altered by an attacker, interfering with a supervisor such that it cannot make correct control decisions. For assisting the monitor function to estimate the number of occurrences of those altered-prone labels, generated by the plant, the notion of label dependency is introduced. Accordingly, a monitor-based supervisor is designed with low online computational effort, avoiding the marking estimation or the reachability analysis of the system. It is verified that the proposed supervisor not only enforces all GMECs no matter whether or not replacement attacks occur, but also keeps the system's behavior as permissive as possible. Note to Practitioners—Cyber physical systems (CPSs) have exhibited multifaceted applications in various fields such as process control systems, smart grids, distributed robotics, autonomous vehicles. Due to the over-reliance on communication networks, CPSs are vulnerable to attacks that can tamper the data collection processes and interfere safety critical decision making processes, resulting in catastrophic damages to the systems. In the frame of discrete event systems, most of existing supervisory control strategies of CPSs under attacks rely on an exhaustive reachability analysis, which is computationally expensive, making such approaches hardly applicable to large systems. In order to address this issue, this work considers a type of sensor deception attacks, called replacement attacks, and proposes a monitor-based supervisor policy, enforcing the control specifications of the systems in the presence of replacement attacks. Without requiring tedious analysis, the designed online supervisor has low computational effort and control decisions only depend on a direct analysis of the observation sequence.}, archive = {J_TASE}, author = {Ziliang Zhang and Gaiyun Liu and Dimitri Lefebvre and Zhiwu Li}, doi = {10.1109/TASE.2024.3506574}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9388--9403}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Monitor-based supervisory control of labeled petri nets under sensor attacks}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant leader-follower controller for uncertain nonlinear multi-agent systems. TASE, 22, 9376--9387. (www)

  @article{ , abstract = {A new distributed adaptive fault-tolerant tracking controller is proposed for consensus control of high-order nonlinear multi-agent systems in the presence of actuator faults and unknown time-varying delays. The system is subject to uncertain disturbances under directed topology. The proposed adaptive method is based on dynamic surface control, where the nonlinear dynamics are unknown, and are approximated by a radial basis function neural network. This system model is subject to delay, fault, uncertainty, disturbances, and singularity of the input signal. The typical constraints on time-delay terms, addressed in the related literature, are relaxed. This relaxation, however, can lead to singularity in the controller. This drawback is addressed by using a specific differentiable function in the control law. It is shown that the generated closed-loop signals are uniformly semi-globally bounded, and the tracking errors converge to a neighborhood of the origin. The effectiveness of the proposed controller is verified by simulations. Note to Practitioners—This work focuses on the distributed control of nonlinear multi-agent time-delay systems in the presence of actuator faults. The proposed control design has potential applications in various real-world network systems such as smart grids, transportation networks, autonomous vehicles, etc. Note that many practical multi-agent systems are subject to time delay and parameter uncertainties, and ignoring these in the control design can negatively impact the system performance and even destabilize it. We demonstrate that the resulting closed-loop system is uniformly semi-globally stable and that the tracking error converges to a neighborhood of the origin. The boundedness assumption on time delay terms, often considered in the multi-agent system literature, is relaxed here. Consequently, the obtained results closely align with practical situations. In addition, the explosion of complexity problem has been avoided by using the dynamic surface control method. The dynamic system represents an unknown-structure dynamic model. This type of dynamic model is applicable to various practical engineering systems. Particularly, robotic systems, chemical reactor systems and spring-mass-damper systems can be described by a model similar to this dynamic system.}, archive = {J_TASE}, author = {Najmeh Zamani and Marzieh Kamali and Javad Askari-Marnani and Hamid Kalantari and Amir G. Aghdam}, doi = {10.1109/TASE.2024.3506123}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9376--9387}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant leader-follower controller for uncertain nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Low-overlap point cloud registration via correspondence augmentation. TASE, 22, 9363--9375. (www)

  @article{ , abstract = {Existing works have made some progress in point cloud registration, but most of them measure performance only on point cloud pairs with high overlap. In practical applications, it is often difficult to ensure that the collected point clouds overlap in large regions due to problems such as occlusion and noise. Therefore, a good low-overlap point cloud registration method is of great practical significance. However, extracting reliable correspondences from point clouds has always been a challenging task, particularly when dealing with low-overlap situation. In this paper, we propose a novel method for low-overlap point cloud registration via efficient correspondence augmentation, called AugLPCR, which not only enhances correspondences with high confidence, but also employs confidence weights to mitigate the impact of outliers. After the augmentation, the correspondences used for the transformation have a large amount of inliers, leading to improved registration performance. Extensive experiments on indoor and outdoor datasets demonstrate that the proposed AugLPCR is capable of maintaining consistent performance and achieve results comparable to or better than the state-of-the-art methods. Note to Practitioners—The motivation of this paper is to address the problem of registering two low-overlap point clouds. Mainstream algorithms for point cloud registration typically assume a sufficient overlap between point clouds. However, in practical scenarios, it is common to encounter scans with inadequate overlap. These conditions often hinder the extraction of reliable correspondences. This paper introduces an effective method for augmenting correspondences to address the problem of low inlier rates within predicted correspondences. While augmenting correspondences with high confidence, it also mitigates the influence of outliers and ambiguous points. Additionally, traditional approaches often divide superpoint regions before matching, but this can lead to the elimination of points in overlapping regions alongside outliers. To address this issue, we adjust the order of superpoint matching and region partitioning. The proposed framework can be easily applied to other correspondence-based point cloud registration models.}, archive = {J_TASE}, author = {Zhi-Huang Lin and Chun-Yang Zhang and Xue-Ming Lin and Huibin Lin and Gui-Huang Zeng and C. L. Philip Chen}, doi = {10.1109/TASE.2024.3506120}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9363--9375}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Low-overlap point cloud registration via correspondence augmentation}, volume = {22}, year = {2025}, }
• (2025). Tracking control for multi-agent systems with model switching and topological switching: A novel dual-switch-based dynamic event-triggered approach. TASE, 22, 9353--9362. (www)

  @article{ , abstract = {The event-triggered tracking consensus of multi-agent systems (MASs) with model switching, topological switching and non-zero leader inputs is concerned in this paper. In previous event-triggered studies on MASs, only a single switching behavior is considered and triggering parameters remain unchanged during switching, which shows more conservatism. In sharp contrast to existing research, this paper deals with two switching behaviors simultaneously, i.e. model switching and topological switching. Accordingly, dual-switch-based dynamic event-triggered rules are proposed to ensure better triggering performance. Further, dual-switch-based event-triggered control protocols are devised for agents. By constructing dual-switch-based multiple Lyapunov functions and utilizing the average dwell time (ADT) method, sufficient conditions are given to ensure the tracking exponential consensus. Finally, two examples of practical mass-spring systems with variable masses are carried out to certify the effectiveness and superiority of our approaches. Note to Practitioners—In real-world MASs, model switching and topological switching often coexist. Considering the limited computational and communication resources, it is of practical significance to conduct event-triggered research on MASs with model switching and topological switching. However, in existing event-triggered research on MASs, triggering control protocols are designed only for one type of switching behavior. To fill the research gap, this paper delves into the event-triggered control of MASs under both model switching and topological switching, offering greater practicality for engineering applications. Meanwhile, the devised triggering protocol is less conservative, ensuring better triggering performance and saving more computing and communication resources. In the simulation, the proposed approach is applied to practical mass-spring control problems under variable masses and switching topologies, and its practical application value is illustrated.}, archive = {J_TASE}, author = {Shuo Zhang and Jinhai Liu and Huaguang Zhang and Wei Wang and Zhigang Zhang}, doi = {10.1109/TASE.2024.3506117}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9353--9362}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tracking control for multi-agent systems with model switching and topological switching: A novel dual-switch-based dynamic event-triggered approach}, volume = {22}, year = {2025}, }
• (2025). Dynamic multi-objective operation optimization of municipal solid waste incineration process based on transfer learning. TASE, 22, 9338--9352. (www)

  @article{ , abstract = {To achieve optimal performance of municipal solid waste incineration (MSWI) process with nonstationary time-varying dynamics, a dynamic multi-objective operation optimization method (DSE-TrMOPSO), based on transfer learning, is proposed in this paper. First, the operation optimization models are established using data stream ensemble learning, where incremental updating and selective ensemble strategies are adopted to cope with changing working conditions. Second, a dynamic multi-objective particle swarm optimization algorithm based on transfer learning (CTrDMOPSO) is designed for optimization calculation. In this algorithm, the hierarchical clustering-based transfer learning strategy is proposed to construct the initial population with high quality. Afterwards, the knee point-based decision making is performed to determine the final setpoints of manipulated variables for index optimization. Then, the feasibility of the designed algorithm CTrDMOPSO is verified on the benchmark problems. Finally, the proposed DSE-TrMOPSO is applied to the MSWI process. The results demonstrate that the proposed method can achieve satisfactory operation performance in terms of combustion efficiency and nitrogen oxides emissions. Note to Practitioners—This study aims to develop an optimal operation method for the MSWI process with nonstationary time-varying dynamics. To achieve this goal, an operation optimization method based on transfer learning is proposed, which includes two key points, the operation optimization modeling and the dynamic multi-objective optimization. In practice, practitioners can implement the proposed method utilizing real-time data streams. The optimization objective models are constructed by data stream ensemble learning, and the time-varying dynamics can be captured with incremental updating and selective ensemble strategies. After that, the optimal setpoints of manipulated variables are derived by the designed dynamic multi-objective particle swarm optimization algorithm as well as intelligent decision making. The hierarchical clustering-based transfer learning strategy can deal with stochasticity and uncertainty with high optimization efficiency. The applicability and superiority of the proposed method are verified by the process data of a real MSWI plant. The proposed method provides valuable insights to realize the optimal operation of MSWI process.}, archive = {J_TASE}, author = {Junfei Qiao and Yingying Cui and Xi Meng}, doi = {10.1109/TASE.2024.3505846}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9338--9352}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic multi-objective operation optimization of municipal solid waste incineration process based on transfer learning}, volume = {22}, year = {2025}, }
• (2025). Cell cryopreservation in a microfluidic chip with vision-based fluid control and region reaching. TASE, 22, 9327--9337. (www)

  @article{ , abstract = {The solution exchange process is crucial in cell cryopreservation, an assistive reproductive technique that enhances reproductive autonomy and helps women overcome infertility challenges. Such a task is time-critical in the sense that the duration of cell exposure to the solutions significantly impacts cell viability. In this paper, a new micromanipulation system has been developed to automate such a task, where the contributions can be summarized as follows. This paper addresses the challenge of tracking cell positions within a microfluidic chip, due to the limitations of the microscope’s field of view (FOV). By utilizing a region control approach with visual feedback, the proposed method ensures that the cell remains centered in the image. Additionally, a real-time tracking method based on correlation filtering is presented, which precisely localizes cells and controls the syringe pump flow rate to mitigate delays and prevent cell loss. Experimental results illustrate the consistent positioning of the micro objects/cells at each step, the satisfactory success rate, and the robustness to the loss of vision feature. Integrated with novel manipulation strategies, our intelligent manipulation system offers a promising solution for in vitro fertilization (IVF), characterized by an embryologist-centered configuration and standardized robotic manipulation. This study aims to standardize clinical cryopreservation by transitioning from manual operation to a more efficient and reliable automated process. Furthermore, this approach can simplify procedures and enhance oocyte viability. Note to Practitioners—The motivation for this study was to propose an automatic method to address the challenges of manual cell cryopreservation, such as high operational complexity, low efficiency, and significant cell damage. Existing automated cell cryopreservation methods have shortcomings as they fail to simultaneously achieve comprehensive cell tracking and avoid damage caused by direct contact between cells and microtools. Additionally, the high complexity of robotic operations makes clinical application difficult. To overcome these limitations, this study introduces a novel robotic micromanipulation method based on microfluidic chip technology. This method enables full-process cell tracking, FOV tracking of the motorized stage, and non-contact manipulation during the exchange process of cells and solutions. Experimental results showed that the proposed method outperformed existing manual methods in terms of oocyte viability. For fertility experts and medical staff, such a robotic micromanipulation system can be combined with other automatic machines to provide a solution to cell surgery, achieving higher accuracy and efficiency.}, archive = {J_TASE}, author = {Shu Miao and Yongyi Jia and Ze Jiang and Jiehuan Xu and Xiang Li}, doi = {10.1109/TASE.2024.3505537}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9327--9337}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cell cryopreservation in a microfluidic chip with vision-based fluid control and region reaching}, volume = {22}, year = {2025}, }
• (2025). TMoE-P: Toward the pareto optimum for multivariate soft sensors. TASE, 22, 9315--9326. (www)

  @article{ , abstract = {Multivariate soft sensors seek to provide accurate estimation of multiple quality variables through the analysis of measurable process variables, representing a significant advance over the traditional focus on single-quality variable sensors within industrial manufacturing. Current progress stays in applying parameter-sharing neural architectures while ignoring two fundamental issues: (1) catastrophic interference, where the indiscriminate sharing of parameters degrades performance due to the discrepancy of objectives; (2) seesaw optimization, where the optimizer overly focuses on one dominant yet simple objective at the expense of others. To address these issues, we reformulate multivariate soft sensors as a multi-objective optimization problem and propose the Task-aware Mixture-of-Experts framework for achieving the Pareto optimum (TMoE-P). Specifically, to handle issue (1), we propose an Objective-aware Mixture-of-Experts (OMoE) module, which consists of objective-specific and objective-shared experts to realize parameter sharing while accommodating the discrepancy between objectives. To handle issue (2), we devise a Pareto Objective Weighting (POW) module, which dynamically balances the weights of learning objectives to approximate the Pareto optimum among competing objectives. Our evaluations on a public soft sensor benchmark showcase TMoE-P’s superior performance, confirming its enhanced accuracy and robustness. Note to Practitioners—Addressing the burgeoning complexity of estimating multiple quality variables in industrial manufacturing processes, this study introduces a novel Task-aware Mixture-of-Experts framework aiming for the Pareto Optimum (TMoE-P). This framework mitigates the issues of catastrophic interference and seesaw optimization by achieving a delicate balance between parameter sharing and maintaining distinctness between objectives, guiding towards the Pareto optimum. The empirical findings affirm the framework’s capability to adeptly handle the complex interplay of variables, potentially enhancing the efficiency and reliability of industrial processes. While initially developed for multivariate data in industrial applications, the TMoE-P framework’s modular and adaptable nature facilitates its extension to a broad spectrum of data structures, optimizers, and neural architectures. Its adaptability offers practitioners a valuable tool to fulfill specific task requirements, making a modest contribution to industrial process control and monitoring.}, archive = {J_TASE}, author = {Licheng Pan and Hao Wang and Zhichao Chen and Yuxin Huang and Zhaoran Liu and Qunshan He and Xinggao Liu}, doi = {10.1109/TASE.2024.3504736}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9315--9326}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TMoE-P: Toward the pareto optimum for multivariate soft sensors}, volume = {22}, year = {2025}, }
• (2025). Component-targeting opportunistic maintenance policy for multi-machine system considering hierarchical structural dependencies. TASE, 22, 9302--9314. (www)

  @article{ , abstract = {Increasing market competition and growing mechanical structural complexity have urged the application of complex systems comprising multi-component machines. Ensuring the economical operation and maintenance (O&M) for such systems is paramount in industry. To achieve this, the component maintenance policy should cover both the machine-level disassembly sequence among critical components and the system-level configuration setting connecting individual machines. However, integrating such hierarchical structural dependencies into maintenance decision-making poses a formidable challenge and remains unexplored in existing literature. Facing this key issue, our study presents a component-targeting opportunistic maintenance policy considering hierarchical structural dependencies (COMP-HSD) to provide real-time cost-effective maintenance schemes. Aiming at deriving the optimal component maintenance group at each maintenance opportunity, this policy builds a decision-making process including triple-level analyzing framework to comprehensively involve the component degradation and the structural dependencies analyses of machine level and system level. Specifically, sequential component-level predictive maintenance (PdM) intervals are initially optimized according to individual health conditions. Then, a maintenance cost savings model is developed to assess the economy of adjusting PdM actions for each component at the current maintenance opportunity, accounting for the impacts of hierarchical structural dependencies. Furthermore, we incorporate maintenance resource limitation into the decision-making procedure to enhance the adaptability to industrial practice. Numerical examples extracted from a partner engine cylinder head manufacturing enterprise certify the significant cost reduction of this COMP-HSD methodology.Note to Practitioners—This work is motivated by the critical problem of calculating a cost-effective component-targeting maintenance policy for a complex production system with multi-component machines. According to the real O&M management of a manufacturing company, it is intractable to reasonably decide when to maintain each critical component in multiple CNC machine tools of a production system, which restricts the maintenance planning economy. This is due to the difficulty of evaluating the impacts brought by hierarchical structural dependencies (i.e., machine-level disassembly sequence and system-level series configuration). Existing approaches are limited to be directly applied to practice since they either focus solely on individual multi-component machine or assume that each machine in a multi-machine system only contains one component, neglecting the holistic evaluation. Besides, the maintenance resource restriction is rarely considered in the literature. To fill this gap, we present the COMP-HSD policy, which simultaneously considers the individual component degradations, systemic maintenance opportunities and complex structural dependencies analysis in the maintenance decision-making process. Referring to this policy, the manufacturers could determine the component maintenance timetable to conduct timely PdM actions with lower cost. Compared with the traditional maintenance policy, our study greatly reduces the total cost, and also offers extending applications such as maintenance tool preparation and spare part ordering according to the component maintenance scheme.}, archive = {J_TASE}, author = {Yutong Ding and Tangbin Xia and Kaigan Zhang and Dong Wang and Meimei Zheng and Ershun Pan and Lifeng Xi}, doi = {10.1109/TASE.2024.3504723}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9302--9314}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Component-targeting opportunistic maintenance policy for multi-machine system considering hierarchical structural dependencies}, volume = {22}, year = {2025}, }
• (2025). Decentralized coordination of multiple buildings with renewable energy resource and electric vehicles. TASE, 22, 9285--9301. (www)

  @article{ , abstract = {With the popularity of electric vehicles (EVs) and renewable energy sources (RES), the flexibility of charging and discharging of EVs and the intermittency of RES have brought challenges to building operations. Considering the mobility of EVs as commuting tools between buildings and the uncertainty of RES, it is of great practical significance to coordinate multiple buildings with RES and EVs on the premise of meeting the state of energy (SOE) requirements of the future trip. We formulate this coordination problem as a stochastic centralized mixed integer linear programming problem. A polyhedral convex set is constructed to describe the SOE uncertainty of EVs. New nonanticipative constraints (NCs) are derived through forward recursion based on constructed scenarios to guarantee the all-scenario-feasibility (ASF) and nonanticipativity of the decision. A Lagrangian relaxation-based decentralized all-scenario-feasible (LR-DASF) algorithm is developed to solve the centralized optimization problem in a decomposition and coordination way. In this method, the optimal ASF solution can be obtained with a fast convergence rate by updating Lagrangian multipliers without solving all subproblems with NCs. The performance of the LR-DASF algorithm is verified by numerical results, which shows that the algorithm can guarantee the ASF of the solution, as well as promote computational efficiency. Note to Practitioners—EVs as energy storage devices bring energy exchanges between buildings accompanying the mobility of EVs which is an opportunity to improve the energy efficiency of multiple buildings. However, as commuting tools, the SOE of EVs must be guaranteed to be larger than the trip requirement over the randomness of RES generation. Furthermore, solving the coordinated optimization problem of multiple buildings with RES and EVs still faces computational complexity challenge due to the spatio-temperal coupling between EVs and buildings, which may lead to costly computational effort in the premise of guaranteeing the feasibility and nonanticipativity of the decision over the uncertainties in practice. Therefore, in order to overcome the above challenges, an LR-DASF algorithm is developed in this paper to solve the coordinated optimization problem of multiple buildings with RES and EVs. Based on the algorithm, for the system operator, it updates and broadcasts the Lagrangian multipliers information to the local coordinators of buildings. For each building, the local coordinator can make ASF decisions based on new NCs with the information obtained from the system operator independently to guarantee the SOE requirement. The method developed in this paper can make faster optimal decisions without perceivable degradation in accuracy and guarantee the SOE requirement of EVs simultaneously, to meet the requirements of feasibility and computational efficiency of decision making in practice. It is conducive to the future application of LR-DASF in the coordinated optimization of buildings and EVs at the city or regional scale.}, archive = {J_TASE}, author = {Fengxia Liu and Zhanbo Xu and Kun Liu and Haoming Zhao and Jiang Wu and Yuzhou Zhou and Xiaohong Guan}, doi = {10.1109/TASE.2024.3504718}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9285--9301}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Decentralized coordination of multiple buildings with renewable energy resource and electric vehicles}, volume = {22}, year = {2025}, }
• (2025). Deadlock control for flexible assembly systems with multiple resource requirements and separately-loaded parts. TASE, 22, 9275--9284. (www)

  @article{ , abstract = {This work addresses the problem of deadlock control in flexible assembly systems (FASs) with multiple resource requirements and parts loaded separately into the system. We use generalized Petri nets to model, analyze, and control these FASs. Two kinds of structural objects in the Petri net model, namely A-circuits and closed $\Omega $ -structures, can characterize deadlocks. Then a necessary and sufficient condition is derived for the liveness of FAS based on these two objects. By using integer linear programs, we compute all the problematic closed $\Omega $ -structures and A-circuits that are related to deadlocks. A specific design of the monitor is developed for each of these objects so as to prevent it from causing deadlock. By composing all these monitors, a deadlock prevention controller is proposed for a subclass of FASs under consideration. Importantly, this work generalizes the liveness condition of ordinary Petri net models to broader non-ordinary ones. Finally, an example can illustrate the usefulness and correctness of the proposed concepts and methods. Note to Practitioners—In the context of FAS, deadlocks are caused by two patterns: “resource circuit-wait” and “event circuit-wait”. The former indicates a circuit-wait relation among shared resources and has been studied extensively in the past decades, while the latter results from the assembly of different parts and has not received much attention. The analysis regarding event circuit-wait is the key to solving a deadlock problem in FAS. To date, there are a limited number of studies addressing the deadlock control problem for FASs with multiple resource requirements. Given such a complex FAS, based on its Petri net model, this paper 1) characterizes the above two circuit-wait patterns by structural objects, 2) presents a sufficient and necessary liveness condition, and 3) develops an effective deadlock prevention controller for a class of FASs under consideration. The proposed deadlock controller can well solve the deadlocks in concerned FAS applications, like automotive industry, semiconductor manufacturing, and so on. Especially, unlike some existing methods, our controller can be applied to the cases where parts of the same product are permitted to load separately into the system.}, archive = {J_TASE}, author = {Yanxiang Feng and Sida Ren and Ye Cao and Keyi Xing and Yikang Yang}, doi = {10.1109/TASE.2024.3504714}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9275--9284}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deadlock control for flexible assembly systems with multiple resource requirements and separately-loaded parts}, volume = {22}, year = {2025}, }
• (2025). Deep learning-based sensor selection for failure mode recognition and prognostics under time-varying operating conditions. TASE, 22, 9252--9274. (www)

  @article{ , abstract = {Failure mode (FM) recognition and remaining useful lifetime (RUL) prediction play pivotal roles in the field of prognostics and health management (PHM), particularly in regard to monitoring machinery degradation. Advances in sensor technology have opened avenues for FM recognition and RUL prediction by leveraging data from multiple sensors. However, previous research has exhibited certain limitations. Some studies have taken all multisensor data as direct inputs, overlooking the potential heterogeneity in the relevance of individual sensor data to machinery degradation. Others have relied on visual inspection and subjective assessments for sensor selection. These approaches struggle to adaptively select sensors, especially in scenarios involving multiple FMs and operating conditions (OCs), and when only partial FM labels are available. To address these challenges, this paper proposes a novel deep-learning network that adaptively selects sensors to jointly recognize FMs and predict RUL under time-varying OCs. The core of this network incorporates an attention-based long short-term memory (LSTM) module. Within this module, adaptive sensor selection weights are generated, leading to the accurate recognition of FMs and the precise prediction of the RUL. In the context of model training, we construct loss functions utilizing semilabeled samples and extract OC-invariant features through domain adaptation, enhancing the accuracy of FM recognition and RUL prediction. To assess the effectiveness and the generalizability of the proposed method, numerical experiments and two case studies involving aircraft engines and bearings are conducted. Note to Practitioners—This paper proposes a deep-learning network designed for adaptive sensor selection in the context of semisupervised FM recognition and RUL prediction, particularly in scenarios involving multiple OCs. To operationalize this method, five key steps are outlined: First, Data Collection: data are gathered from multiple sensors, time-varying OCs, multiple FMs, and failure time of historical units. It is important to note that FM data are available for only a subset of historical units. Second, Data Preprocessing: The sensor data are tailored using the sliding time window technique, generating a substantial number of samples from historical units. Third, Network Construction: Subsequently, the adaptive sensor selection network is constructed. Fourth, Model Training: Loss functions are constructed, and model parameters are estimated through end-to-end training. Finally, Model Application: The method is utilized to recognize FMs and predict RUL for in-service units. This deep learning network can be applied effectively to various degraded machines experiencing multiple FMs and OCs. It is particularly suitable for machines with complex physical mechanisms or unknown failure thresholds.}, archive = {J_TASE}, author = {Yuhui Wang and Andi Wang and Di Wang and Dong Wang}, doi = {10.1109/TASE.2024.3504595}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9252--9274}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep learning-based sensor selection for failure mode recognition and prognostics under time-varying operating conditions}, volume = {22}, year = {2025}, }
• (2025). Large-scale water quality prediction with deep decomposition architecture and auto-correlation. TASE, 22, 9240--9251. (www)

  @article{ , abstract = {Water quality prediction provides timely insights for addressing potential water environmental issues. Transformer-based models have been widely used in water quality prediction. However, the following challenges exist: 1) Noise in the time series of water quality causes nonlinear models to be overfit; 2) It is difficult to identify temporal correlations in complex time series data; and 3) Information utilization is limited in long-term prediction. This work introduces a large-scale water quality prediction model named SVD-Autoformer to address them. SVD-Autoformer combines a Savitzky-Golay (SG) filter, variational mode decomposition (VMD), an auto-correlation mechanism, and a deep decomposition architecture, which is achieved in the renovation of the transformer. First, the SG filter removes noise while retaining valuable data features. SVD-Autoformer employs the SG filter as a data preprocessing tool to reduce noise and prevent nonlinear models from overfitting. Second, VMD extracts major modes of the signals and their respective center frequencies, thus providing richer features for the prediction. Third, the deep decomposition architecture with embedded decomposition modules allows for gradual decomposition during the prediction process. SVD-Autoformer employs the architecture to extract more predictable components from complicated water quality time series for long-term forecasting. Finally, SVD-Autoformer applies the auto-correlation mechanism to capture the temporal dependence and enhance information utilization. Numerous experiments are conducted and the results demonstrate that SVD-Autoformer provides superior prediction accuracy over other advanced prediction methods with real-world datasets. Note to Practitioners—This paper explores the critical aspects of time series water quality prediction, aiming to provide valuable insights for engineers and decision-makers. Traditional water quality prediction methods primarily rely on linear time series approaches and suffer from high computational complexity when dealing with large-scale data. This study is motivated by the transformer architecture with highly parallel computing capability and innovatively proposes deep decomposition architecture to extract more predictable components. In practice, to handle massive data with low time complexity, we introduce an auto-correlation mechanism. We conduct experiments using real-world datasets to demonstrate that this method achieves superior water quality prediction accuracy. Additionally, the method has been deployed in a real-world water quality prediction platform. Our future work includes its applications to different real-world datasets arising from electric power, intelligent transportation, and meteorological rainfall prediction.}, archive = {J_TASE}, author = {Jing Bi and Mingxing Yuan and Haitao Yuan and Junfei Qiao and Jia Zhang and MengChu Zhou}, doi = {10.1109/TASE.2024.3504575}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9240--9251}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Large-scale water quality prediction with deep decomposition architecture and auto-correlation}, volume = {22}, year = {2025}, }
• (2025). Dual-resource constrained flexible job shop scheduling for SMT back-end production and an empirical study of wearable devices. TASE, 22, 9230--9239. (www)

  @article{ , abstract = {Increasing needs for smart production of high-mix consumer electronics products with shortening product life cycles and complex product mix in wearable electronics industry are challenging. Surface mount technology (SMT) back-end manufacturing is semi-automatic and labor-intensive with combinatorial complexity for configurating multiple resources that have caused increasing costs and resource waste. This study aims to develop an effective solution for dual resource constrained flexible job shop scheduling considering realistic production constraints, non-deterministic polynomial-time hardness, and anthropic factors for implementation in real settings. Indeed, the main contributions of the proposed adaptive scheduling solution include the integration of quantum tunneling mutation mechanism and genetic algorithm to minimize the total tardy jobs, the first-day due tardiness, and makespan for high-mix dual resource constrained flexible job shop scheduling that can obtain near optimal solution within practically limited computing time. An empirical study was conducted for validation. The results have shown the practical viability of the developed solution that can be quickly responsive and adaptive to dynamic needs. Note to Practitioners—SMT back-end scheduling solution is developed to empower smart production of high-mix for wearable consumer electronics with shortening product life cycles in practice. The developed solution for dual resource constrained flexible job shop scheduling considering realistic production constraints, non-deterministic polynomial-time hardness, and anthropic factors to minimize the total tardy jobs, the first-day due tardiness, and makespan. The developed solution has been implemented in a world leading consumer electronics company.}, archive = {J_TASE}, author = {Chia-An Chen and Hsuan-An Kuo and Chen-Fu Chien}, doi = {10.1109/TASE.2024.3503412}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9230--9239}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-resource constrained flexible job shop scheduling for SMT back-end production and an empirical study of wearable devices}, volume = {22}, year = {2025}, }
• (2025). TFGait—Stable and efficient adaptive gait planning with terrain recognition and froude number for quadruped robot. TASE, 22, 9218--9229. (www)

  @article{ , abstract = {Gait planning is one of the most critical technologies for quadruped robots. However, far too little attention has been paid to the tight coupling mechanism of gait planning with terrain understanding and energy efficiency. To date, it is still challenging to plan optimal gait strategies that are highly adapted to terrain features with stable and efficient transitions. Accordingly, this paper proposes an adaptive gait control framework for quadruped robots that combines terrain recognition, Cost of Transport (CoT), and the Froude (Fr) number. More specifically, an optimal gait selection strategy for quadruped robots is designed based on different terrain texture features and the CoT characteristics of different gaits. To address the gait transition process induced thereby, an adaptive method for gait parameters based on the Fr number is further proposed, which can make the process more stable. Besides, model predictive control (MPC) and whole-body control (WBC) are employed as the motion controllers for the quadruped robot. Furthermore, simulation and experimental results indicate that the proposed method possesses superior terrain adaptability, energy efficiency, and motion stability during gait transitions, which is beneficial for the quadruped robots to maintain stable motion and reduce energy consumption when performing tasks in changeable terrains.Note to Practitioners—This paper is motivated by the problem of adaptive gait planning for quadruped robots that walks through different terrains. We propose a method that ensures optimal gait selection by robots facing diverse terrains and maintains the stability of gait transition. The proposed control framework, upon testing in a simulated environment, can be directly deployed on real-world robot without further adjustments and allows the robot to traverse various terrains with minimal sim-to-real issues. Hopefully, our proposed method can provide valuable guidance and support for facilitating the enhancement of capabilities in performing prolonged endurance tasks in unstructured environments for quadruped robots.}, archive = {J_TASE}, author = {Aocheng Luo and Qifeng Wan and Yan Meng and Shihan Kong and Wanchao Chi and Chong Zhang and Shenghao Zhang and Yuzhen Liu and Qiuguo Zhu and Junzhi Yu}, doi = {10.1109/TASE.2024.3503277}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9218--9229}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TFGait—Stable and efficient adaptive gait planning with terrain recognition and froude number for quadruped robot}, volume = {22}, year = {2025}, }
• (2025). The input-to-state stability in probability for constrained delay systems with stochastic delayed impulses. TASE, 22, 9205--9217. (www)

  @article{ , abstract = {This paper investigates the problem of the input-to-state stability in probability (ISSiP) for a family of constrained delay systems with stochastic delayed impulses. A new type of admissible edge-dependent average impulsive interval (AED-AII) impulsive signal is presented, in which the concepts of AED-AII impulsive intensity and AED-AII impulsive density are firstly presented to comprehensively describe the impulsive signals. Under the constraint of stochastic impulses, we study the ISSiP of impulsive systems in two different scenarios. One scenario is that the impulsive intensity is stochastic and the impulsive density is limited by a deterministic impulsive interval. Another scenario is that both impulsive intensity and density are subject to stochastic constraints, and the impulsive interval is satisfied by a renewal process. Some novel AED-AII-based criteria for ISSiP are deduced, respectively, where a relationship between impulsive density, impulsive intensity, the jumping probability of impulse mode and Lyapunov function decay rate is established. Then, we extend this study to constrained impulsive switched systems and obtain a class of AED-based criteria to guarantee the ISSiP property under different impulsive conditions. Finally, examples are given to depict the validity of the proposed stability criteria. Note to Practitioners—The problem discussed in this paper is the stability problem of a set of constrained delay systems. Under the constraint of stochastic impulses, for the first time, the intensity and density of impulsive signals are characterized from the perspective of edge-dependent, and input-to-state stability in probability is studied, which is then extended to impulsive switched systems. We propose AED-AII to investigate the different situations of impulsive intensity and impulsive density under random constraints, in order to ensure the ISSiP of the system. Considering the structure and performance of circuit systems and robot manipulators, such as non-autonomous Chua’s circuits with nonlinear resistors, external inputs, and random noise, this method can be applied to systems with random noise.}, archive = {J_TASE}, author = {Lijun Gao and Zihan Liu and Suo Yang}, doi = {10.1109/TASE.2024.3502656}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9205--9217}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The input-to-state stability in probability for constrained delay systems with stochastic delayed impulses}, volume = {22}, year = {2025}, }
• (2025). A dynamic joint production-service scheduling approach by a model-and-data driven algorithm. TASE, 22, 9190--9204. (www)

  @article{ , abstract = {Dynamic production and service scheduling are critical in manufacturing servitization. Unlike previous research assuming unrealistic specific mappings between products and services, this paper studies a more general and complex setting with a two-product-two-service mapping in a product service system, formulating it as a Markov Decision Process model. To solve this model with large-scale cases, designing a data-driven prediction algorithm is a common approach to overcome the curse of dimensionality. However, there might exist inconsistencies in the statistical distribution between the training and testing sets, as the training samples generated by the exact algorithm may not cover the entire state space. To tackle this problem, we propose a novel model-and-data-driven framework. We extract a service priority rule from the model as prior knowledge and embed it into the machine learning algorithm through tailored feature engineering. This strengthens prediction capabilities in unexplored state space regions compared to a purely data-driven approach. Numerical experiments validate our contributions, displaying the potential for embracing industrial automation and exploring the synergy effects for further exploration of the optimal structure as more effective prior knowledge. Note to Practitioners—Dynamic joint scheduling of production and service plays an important role in industrial automation. In this context, a wide range of studies are dedicated to developing efficient algorithms to meet the quick response requirement to tackle large-scale cases beyond the capability of exact algorithms. However, these studies often rely on unrealistic assumptions or have unsatisfactory computational performance, limiting potential cost reduction. To overcome this challenge, we relax unnecessary hypotheses and generalize the problem by allowing for a more generalized production-service mapping, which better reflects the reality. Apart from that, we integrate the prior knowledge into existing data to construct a more powerful approximation algorithm and enrich the prediction capability to accommodate the dynamic environment. With greater computation resources and data samples, the proposed algorithm achieves superior solutions. Furthermore, our research uncovers some key properties of the product service system, which could be leveraged to customize more effective automated scheduling rules.}, archive = {J_TASE}, author = {Yi-Lun Zhang and Zhi-Bin Jiang and Shan Jiang and Li-Ping Zhou}, doi = {10.1109/TASE.2024.3502521}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9190--9204}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A dynamic joint production-service scheduling approach by a model-and-data driven algorithm}, volume = {22}, year = {2025}, }
• (2025). Universal domain adaptation in intelligent fault diagnosis via simulating unseen classes. TASE, 22, 9178--9189. (www)

  @article{ , abstract = {Since the introduction of domain adaptation to fault diagnosis, it has greatly improved the performance of cross-domain fault diagnosis. However, previous methods require constructing independent frameworks for specific domain adaptation scenarios in fault diagnosis, which reduces efficiency and increases consumption. The difficulty lies in the fact that the classes of source and target domains in universal domain adaptation scenarios are not uniform and agnostic, leading to the inability to identify unknown fault modes. To address this issue, we propose a novel Simulating Unseen Classes (SUC) method. It is a general framework that does not depend on a priori knowledge of the classes of target domains. More data with new classes are generated to simulate the unknown fault classes based on source domain at both data and feature levels. Class gap and domain gap between source and target domains are effectively reduced to extract domain invariant features. Adequate experiments are conducted on three benchmark datasets and the results demonstrate that our method can outperform other methods by a large margin. Note to Practitioners—Intelligent fault diagnosis plays an important role in automation system. In practice, the changes in equipment, operating conditions and the environment raise the difficulty of cross-domain fault diagnosis. This paper proposes a novel SUC method for universal domain adaptation in cross-domain fault diagnosis. It is a more general but difficult scenario where the prior of source and target domains is agnostic. More data with new classes is generated effectively based on source domain at both data and feature levels. Class gap and domain gap are reduced to improve the accuracy of fault diagnosis. Our method provides a universal framework that can be readily applied to different domain adaptation (DA) scenarios, including closed set DA, partial DA, open set DA, and universal DA. Experimental results demonstrate the superiority of our method. In the future, we will investigate more challenging issues in fault diagnosis, such as scenarios where the data in the target domain is unavailable during training.}, archive = {J_TASE}, author = {Huikai Shao and Dongdong Jing and Ning Xia and Zixiang Tang and Dexing Zhong}, doi = {10.1109/TASE.2024.3502426}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9178--9189}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Universal domain adaptation in intelligent fault diagnosis via simulating unseen classes}, volume = {22}, year = {2025}, }
• (2025). Multi-task multi-agent reinforcement learning with task-entity transformers and value decomposition training. TASE, 22, 9164--9177. (www)

  @article{ , abstract = {Multi-task multi-agent reinforcement learning aims to control multiple agents to perform well on multiple tasks. It encounters three core challenges: the varying number of agents and entities, the disparities in cooperative behaviors among different tasks, and the training imbalance caused by varying task difficulty levels. To address these issues, we propose a novel framework named Task-Entity Transformer Qmix (TETQmix), which employs pretrained language models for task encoding, utilizes proposed Task-Entity Transformer to handle observations across various tasks, and adjusts task learning weights to achieve balanced multi-task training. Task-Entity Transformer not only enables handling multi-task scenarios with varying numbers of agents and entities, but also leverages cross-attention modules to integrate observation and task embeddings, so that each agent can obtain individual values and decisions for multiple tasks. We then utilize a transformer-based mixer to monotonically combine the individual values, and train the whole network’s parameters using temporal-difference errors. To facilitate multi-task training, we define task regret as the difference between the current-stage return and the candidate best one, and adjust the learning weight of each task based on its task regret. Experiments are conducted on both simulated multi-particle environments and real-world multi-robot systems. Compared with existing baselines, our method not only is superior in multi-task learning efficiency, but also shows promising transfer ability on unseen tasks. Note to Practitioners—The flexibility of multi-agent systems makes them quite fit to multiple tasks. Compared to designing different decision models for different tasks, it is more convenient if one can use just one decision model to resolve multiple tasks. Besides, it can make the maximum utilization of trajectory data coming from similar tasks when the data are integrated for multi-task decision model training. Natural language provides a powerful tool to describe the task context and emphasize the similarities or differences among different tasks. Pretrained language models can encode the task context, based on which the decision model can adjust its output distribution for different tasks and even synthesize the decisions from existing and similar tasks to achieve promising zero-shot and few-shot transfer performance for unseen tasks. With our proposed TETQmix, practitioners are able to realize multi-task capability in multi-agent systems and increase the generalization in a variety of scenarios.}, archive = {J_TASE}, author = {Yuanheng Zhu and Shangjing Huang and Binbin Zuo and Dongbin Zhao and Changyin Sun}, doi = {10.1109/TASE.2024.3501580}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9164--9177}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-task multi-agent reinforcement learning with task-entity transformers and value decomposition training}, volume = {22}, year = {2025}, }
• (2025). EEG-based automated system for reach-and-grasp identification using amplitude envelope enabled multivariate spectral information. TASE, 22, 9153--9163. (www)

  @article{ , abstract = {The amplitude envelope is a crucial parameter to analyse natural systems as it provides useful amplitude modulation (AM) based information. In many cases, power spectral entropy (PSE) of a non-stationary signal is not able to discriminate AM based information. This paper proposes amplitude envelope based spectral entropy (ASE) which quantifies AM related information in the spectral domain and superiority is justified in comparison with PSE. Moreover, ASE is extended for multivariate signal analysis using joint AM based information. Efficacy of ASE is shown in various scenarios. Further, multivariate ASE is utilized for the development of a reach-and-grasp identification system using multichannel electroencephalogram (EEG) recordings. In this system, a novel correntropy based channel selection method is proposed to reduce system complexity. The number of EEG channels are reduced by almost 50% using the proposed channel selection method. The selected channels are decomposed into intrinsic mode functions (IMFs) using multivariate decomposition method. The AM based information present in these IMFs is obtained using multivariate ASE. Support vector machine classifier with radial basis function kernel is utilized to identify the type of grasp. Pearson correlation coefficient-based feature ranking is applied to select the significant features. The highest classification performance is achieved using five features with accuracy, sensitivity and specificity of $72.03~\pm ~2.39$ %, $66.19~\pm ~8.96$ % and $83.31~\pm ~1.67$ % respectively, which is better than compared method. The proposed reach-and-grasp identification method can be used to develop real time systems to avail natural control of neuroprosthetic devices. Note to Practitioners—The multichannel recordings has wide applications in brain-computer interface and human-machine interaction. This paper suggests a novel foundation for the use of information preserved in amplitude envelope of multichannel EEG signals. We have demonstrated the idea in four steps: 1) variation of the proposed amplitude envelope based spectral entropy (ASE) for various modulation cases; 2) extension of ASE for multivariate data analysis and its possible application; 3) correntropy based significant channel selection to reduce the system complexity; and 4) potential use of amplitude envelope based neuronal information to develop an automated system for reach-and-grasp identification.}, archive = {J_TASE}, author = {Aakash Shedsale and Shivam Sharma and Rishi Raj Sharma and Ram Bilas Pachori}, doi = {10.1109/TASE.2024.3500370}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9153--9163}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {EEG-based automated system for reach-and-grasp identification using amplitude envelope enabled multivariate spectral information}, volume = {22}, year = {2025}, }
• (2025). Adaptive oscillation-suppression control for distributed nonholonomic vehicle safe formation with nested input saturation. TASE, 22, 9140--9152. (www)

  @article{ , abstract = {Nonholonomic vehicles in distributed networks are prone to triggering nested velocity and acceleration saturation during reactive safety formations, exacerbating oscillations. This paper proposes a hybrid secure distributed collaborative frame-work, integrating compound adaptive anti-windup strategies with vehicle kinematics and safe geofences to achieve smooth and effective obstacle and collision avoidance while suppressing saturation-induced oscillations. The vehicle’s safe behavior for bypassing obstacles is formed via acceleration envelopes from safe geofences and input saturation, which generate constraint velocity commands. Additionally, a low-trigger and power-adjustable enhanced artificial potential field is integrated into the safety coordination to fine-tune vehicle maneuvers at extremely close distances to hazardous targets, ensuring high reliability. Safe acceleration envelopes and nested kinematic saturation are utilized to design a compound adaptive auxiliary dynamic system, smoothing oscillations induced by dual command constraints during formation. A distributed formation controller is further designed to enable multitasking collaboration in formations. The overall stability is mathematically analyzed, and the method’s superior smoothness and safety in task coordination are validated through simulations and experiments with vehicle clusters. Note to Practitioners—In response to the severe trajectory oscillations caused by saturation triggered by existing reactive avoidance approaches, this paper proposes a novel hybrid safety collaborative control based on the nonholonomic vehicle kinematics that markedly enhances the smoothness and safeness of formations in obstacle environments. The integration of safety acceleration envelopes, as well as low-trigger and adjustable artificial potential functions, markedly mitigates oscillations from reaction saturation compared with the solitary traditional artificial potential functions, as evidenced by simulations and experiments that demonstrate reduced oscillation amplitudes and shorter recovery times when evading hazardous targets using the proposed method. In addition, existing velocity/acceleration nested windups in actual applications are concurrently considered for the first time, and the corresponding compound adaptive anti-windup method is employed to smooth oscillations caused by control saturation. The security and smoothing strategies outlined allow for collaborative operations in more complicated obstacle environments and enable the deployment of larger vehicle clusters in confined spaces, significantly enhancing multi-vehicle collaboration’s economic viability and efficiency. Furthermore, the safety collaborative control framework designed for kinematics is conveniently structured for engineers as a standalone module, which is easily transferrable to commercial robotic products. The composite approach to safeness and smoothness can also be applied in other unmanned and manned collaborative scenarios.}, archive = {J_TASE}, author = {Zhi Zheng and Tao Jiang and Jianxiang Wang and Xiaojie Su and Jiangshuai Huang and Zhenshan Bing and Alois Knoll}, doi = {10.1109/TASE.2024.3500076}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9140--9152}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive oscillation-suppression control for distributed nonholonomic vehicle safe formation with nested input saturation}, volume = {22}, year = {2025}, }
• (2025). Nonlinear observer-based sliding mode control for robot-aided bilateral human-compliant rehabilitation training of upper limb. TASE, 22, 9130--9139. (www)

  @article{ , abstract = {Robotic-assisted rehabilitation therapy has been a promising way in improving upper limb motor function. This paper proposes a multi-mode training control method for a bilateral upper limb rehabilitation robotic system, with which human-compliant rehabilitation training can be provided. Firstly, an admittance controller is built to transform the human-robot interaction force to compliant desired trajectory. Then, by integrating with super-twisting algorithm, a nonlinear observer is designed to estimate the lumped disturbance exerted on the driving revolute joint, including the active force applied by human subject, the force of friction, the model uncertainty, et al. To guarantee that the state of position converges to the desired value in real time, a high-order sliding mode controller combined with the disturbance compensation from the observer is proposed. Additionally, based on the aforementioned several methods, multiple bilateral training modes are constructed for patients in different rehabilitation stages. The overall system including the constructed bilateral rehabilitation robotic system and the proposed control method is verified in several experiments, demonstrating the advantage of the controller on interaction compliance with respect to normal method in addition to the capability of multiple rehabilitation training modes.Note to Practitioners—This work is motivated by the patients’ needs of the compliance and comfort during the human-robot interaction in the robot-aided rehabilitation training process. Thus, a nonlinear observer-based sliding mode controller combined with admittance model is proposed in this paper. The developed control method has the following functionalities: (1) Ensuring the compliance of the desired trajectory via the constructed admittance model. (2) Solving the estimation of lumped disturbance exerted on the robotic system based on nonlinear force observer. (3) Ensuring the trajectory tracking with high accuracy under unknown disturbances via sliding mode controller combined with the observer compensation. The controller can be potentially applied in lots of areas: 1) Human-robot compliant collaboration or interaction, e.g., human-robot cooperative manipulation, robotic surgery; 2) Precise motion of robotic arm under unknown disturbance.}, archive = {J_TASE}, author = {Jianfeng Li and Wenjie Liu and Haozhe Li and Ran Jiao and Mingjie Dong}, doi = {10.1109/TASE.2024.3499329}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9130--9139}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nonlinear observer-based sliding mode control for robot-aided bilateral human-compliant rehabilitation training of upper limb}, volume = {22}, year = {2025}, }
• (2025). Knowledge transfer enabled diverse task scheduling for individualized requirements in industrial cloud platform. TASE, 22, 9117--9129. (www)

  @article{ , abstract = {Nowadays, application providers often prefer to execute their workflows on heterogeneous distributed computing resources deployed on cloud infrastructure to achieve a high level of resilience and cost saving. Optimally scheduling workflow on computing resources is a well-known combinatorial optimization problem, where a trend of using evolutionary algorithm (EA) is emerging rapidly. However, conventional EA optimizes only one problem in a single run and suffers from a high computational burden. In practical scenario, cloud platform needs to handle massive amounts of scheduling requests from users, scheduling different workflows simultaneously is highly challenging. Bearing this in mind, we put forward a novel knowledge transfer enabled EA to schedule diverse workflows in tandem, where domain knowledge of scheduling one workflow is extracted to enhance the scheduling efficiency of other related workflows. In our design, the knowledge source selection and the intensity of performing knowledge transfer are adapted in a synergistic way. Furthermore, search operator is enhanced by exploiting both historical experience and heuristic information. Experimental results on real-life workflows and extensive synthetic applications demonstrate the competitiveness of our approach, in comparison to state-of-the-art contenders. Note to Practitioners—Workflow scheduling is an important requirement for users in cloud computing, whose intractability increases exponentially when the size of problem grows, posing stiff challenges to heuristic methods. Using EAs to tackle workflow scheduling has received increasing attention recently. Suppose workflow scheduling is treated as a optimization task, cloud platform typically needs to handle versatile tasks from numerous users. However, traditional EA optimizes only one task in a single run and unable to handle multiple tasks at the same time. To address this issue, we introduce a novel multi-task solver to resolve different tasks jointly via online learning and exploitation of problem-solving experiences across tasks. The results demonstrate that our proposal significantly outperforms the state-of-the-art peers. It is expected to facilitate the practical efficacy of industrial cloud system which faces multiple workflow scheduling tasks submitted from enormous users.}, archive = {J_TASE}, author = {Jiajun Zhou and Liang Gao and Chao Lu and Yun Li}, doi = {10.1109/TASE.2024.3498064}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9117--9129}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Knowledge transfer enabled diverse task scheduling for individualized requirements in industrial cloud platform}, volume = {22}, year = {2025}, }
• (2025). The impact of two scheduling windows on outpatient management for a clinic charging optional appointment fees with patient no-shows. TASE, 22, 9104--9116. (www)

  @article{ , abstract = {The efficiency of outpatient care deeply affects patient services in all downstream departments within a hospital. West China Hospital (WCH) is the largest public hospital in Southwest China. Nevertheless, WCH still experienced relatively high patient no-shows. WCH offered patients the option to prepay appointment fees to motivate them to keep scheduled appointments. This study aims to investigate the impacts of optional application fees on the scheduling system and, specifically, to gain insights into the scheduling policy: Should WCH use different scheduling windows (SWs) for patients depending on whether they prepay or do not pay the appointment fee? We developed an $M/M /1/N$ queueing model to determine the optimal SWs for two patient types - prepaid or unpaid. Our study shows that WCH should use two different SWs when adopting optional appointment fees. Furthermore, we derived theoretical findings and validated them using actual WCH data. Empirical results show that there could be an average of 20% cost savings if two optimal SWs are used for scheduling. Note to Practitioners—This paper was motivated by the problem of no-shows in the outpatient department at West China Hospital (WCH). The two most common approaches to mitigating patient no-shows are overbooking and scheduling window (SW). However, excessive overbooking may lead to physician overtime and long patient wait time. This paper examined if different SWs should be used to reduce patient no-shows. Due to the high patient demand, WCH provided multiple appointment channels, using one common SW, for outpatient scheduling. WCH yet still experienced relatively high patient no-shows. For the curiosity of outpatient management, WCH considered if an option for patients to prepay appointment fees could motivate them to attend appointments. This paper aims to answer WCH’s questions by examining the overall cost differences between a system with the prepaid option and a system without it. We developed an $M/M /1/N$ queueing model to determine the optimal SWs for two patient types - prepaid or unpaid. Our study shows that WCH should consider using two different SWs when charging optional appointment fees. We derived theoretical findings and validated them using actual data from WCH. Moreover, the empirical analysis shows that the system could achieve an average of 20% cost savings if two optimal SWs are used for outpatient scheduling.}, archive = {J_TASE}, author = {Ying Zhou and Li Luo and Bernard T. Han and Xiaoli He and Zhi Wan and Lu Gan}, doi = {10.1109/TASE.2024.3497952}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9104--9116}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The impact of two scheduling windows on outpatient management for a clinic charging optional appointment fees with patient no-shows}, volume = {22}, year = {2025}, }
• (2025). Event-triggered fixed-time sliding mode control for lip-reading-driven UAV: Disturbance rejection using wind field optimization. TASE, 22, 9090--9103. (www)

  @article{ , abstract = {This paper investigates the fixed-time sliding mode control (FTSMC) problem for a quadcopter unmanned aerial vehicle (QUAV), which is driven by a lip-reading recognition module. The lip-reading recognition module is consisted of a trained deep neural network with the structure of 2D-Conv+GhostNet+TCN. In order to reduce the communication burden between the remote controller and the QUAV as well as reduce the computation burden in running the lip-reading recognition module, the event-triggered mechanism is introduced to the position controller design. The low-bound of the triggering interval is derived explicitly so that the Zeno phenomenon can be excluded. Furthermore, in order to overcome the main obstacle in high-accuracy control of QUAV, this paper launches a novel wind disturbance rejection approach by using wind field model, which is motivated by the physical dynamic characteristics of the practical wind. Specifically, the wind disturbance is estimated in the designed FTSMC by applying a specific wind field equation with preassigned physical parameters. To further reduce the chattering in the controller, a fitting technique is introduced via a local multivariate linear regression. Finally, both simulation and human-in-the-loop experiment results verify the applicability of the proposed control approach for the lip-reading-driven QUAV system. Note to Practitioners—This research is motivated by the need to design lip-reading-driven QUAV. In noisy environments or when silence is required, the efficiency of traditional human-computer interaction methods such as speech recognition is greatly reduced. Especially for people with damaged vocal cords, speech recognition is not achievable. In addition, it is difficulty to realize high-precision anti-interference control of QUAV with lower computational and communication burdens. In order to solve these problems, this research designs a lip-reading recognition module for QUAV control to cope with various complex application scenarios and realizes high-performance control by FTSMC algorithm. The key of this work to save system resources is to introduce the event-triggered mechanism into the position controller of the QUAV. In addition, this paper introduces the wind field model into the QUAV model to realize the wind disturbance suppression. The lip-reading-driven QUAV proposed in this paper have a wide range of applications, such as controlling QUAV in hazardous environments and improving the efficiency of interaction between human and QUAV.}, archive = {J_TASE}, author = {Tian Lan and Jun Song and Zhiwei Hou and Kang Chen and Shuping He and Hai Wang and Jason J. R. Liu}, doi = {10.1109/TASE.2024.3496939}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9090--9103}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered fixed-time sliding mode control for lip-reading-driven UAV: Disturbance rejection using wind field optimization}, volume = {22}, year = {2025}, }
• (2025). An attack-defense game-based reinforcement learning privacy-preserving method against inference attack in double auction market. TASE, 22, 9075--9089. (www)

  @article{ , abstract = {Auction mechanism, as a fair and efficient resource allocation method, has been widely used in varieties trading scenarios, such as advertising, crowdsensoring and spectrum. However, in addition to obtaining higher profits and satisfaction, the privacy concerns have attracted researchers’ attention. In this paper, we mainly study the privacy preserving issue in the double auction market against the indirect inference attack. Most of the existing works apply differential privacy theory to defend against the inference attack, but there exists two problems. First, ‘indistinguishability’ of differential privacy (DP) cannot prevent the disclosure of continuous valuations in the auction market. Second, the privacy-utility trade-off (PUT) in differential privacy deployment has not been resolved. To this end, we proposed an attack-defense game-based reinforcement learning privacy-preserving method to provide practically privacy protection in double auction. First, the auctioneer acts as defender, adds noise to the bidders’ valuations, and then acts as adversary to launch inference attack. After that the auctioneer uses the attack results and auction results as a reference to guide the next deployment. The above process can be regarded as a Markov Decision Process (MDP). The state is the valuations of each bidders under the current steps. The action is the noise added to each bidders. The reward is composed of privacy, utility and training speed, in which attack success rate and social welfare are taken as measures of privacy and utility, a delay penalty term is used to reduce the training time. Utilizing the deep deterministic policy gradient (DDPG) algorithm, we establish an actor-critic network to solve the problem of MDP. Finally, we conducted extensive evaluations to verify the performance of our proposed method. The results show that compared with other existing DP-based double auction privacy preserving mechanisms, our method can achieve better results in both privacy and utility. We can reduce the attack success rate from nearly 100% to less than 20%, and the utility deviation is less than 5%. Note to Practitioners—Privacy protection in trading markets, such as advertising, crowdsensing, and spectrum, is crucial. Traditional approaches like differential privacy have been unable to entirely guard sensitive data against inference attacks. To address this, we introduce a novel privacy-preserving mechanism for double auction markets. Our approach employs an attack-defense game model, where noise is added to bidders’ valuations and then used to launch an inference attack. This process allows for the evaluation of the noise’s effectiveness and iteratively refines the privacy protection method. Transformed into a reinforcement learning model and optimized through a DDPG network, our mechanism reduces computational complexity. It has been shown to significantly diminish the success rate of inference attacks, while maintaining a minimal utility deviation. Practitioners in auction-based markets can leverage our approach to enhance privacy protection without negatively impacting market performance. By integrating our mechanism into their operations, auctioneers can foster a safer and more efficient trading environment.}, archive = {J_TASE}, author = {Donghe Li and Chunlin Hu and Qingyu Yang and Yuhao Ma and Feiye Zhang and Dou An}, doi = {10.1109/TASE.2024.3496869}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9075--9089}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An attack-defense game-based reinforcement learning privacy-preserving method against inference attack in double auction market}, volume = {22}, year = {2025}, }
• (2025). Diagnosability verification and enforcement for unbounded petri nets by online supervisors. TASE, 22, 9061--9074. (www)

  @article{ , abstract = {This paper addresses the problems of diagnosability verification and enforcement of discrete event systems modeled with unbounded Petri nets. Diagnosability in such systems is critical for ensuring reliability and maintaining operational integrity, yet current methods often struggle with the complexity introduced by unboundedness and potential deadlocks. Given an unbounded labeled Petri net that may reach deadlocks, a quiescent basis coverability graph is established to verify the diagnosability of the considered system. This procedure employs a deterministic finite state automaton, called an extended verifier, derived from the proposed quiescent basis coverability graph. It is shown that an unbounded Petri net is diagnosable if and only if the verifier does not contain a class of cycles, called repetitive F-cycles. This result also provides necessary and sufficient conditions for diagnosability enforcement by developing an online supervisor. Further, the designed supervisor is maximally permissive and also circumvents a plant entering deadlocks by firing non-fault sequences. Examples are presented to demonstrate the proposed method. Note to Practitioners—Fault diagnosis and diagnosability enforcement are critical for the development and operation of highly automated systems covering computer-integrated production processes, intelligent traffic, computer and communication networks, smart gird, etc. This work touches upon this problem from the perspective of discrete event systems that are modeled with unbounded labeled Petri nets. The feasibility and applicability of the reported method stem from the usage of a structurally compact representation of a considered plant such that the computational cost of a real-world system is acceptable. The graphical representation of Petri nets as well as the proposed quiescent basis coverability graph make the method easy to use and manipulate. Moreover the sufficient and necessary conditions of diagnosability enforcement can be readily verified by the supervisory theory, facilitating its adoption by practitioners.}, archive = {J_TASE}, author = {Shaopeng Hu and Yihui Hu and Ding Liu and Maria Pia Fanti and Zhiwu Li}, doi = {10.1109/TASE.2024.3496861}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9061--9074}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Diagnosability verification and enforcement for unbounded petri nets by online supervisors}, volume = {22}, year = {2025}, }
• (2025). Dynamics-based motion control for a hybrid-driven continuum robot with continuously variable stiffness. TASE, 22, 9049--9060. (www)

  @article{ , abstract = {While the hybrid driving method effectively addresses the contradiction between inherent compliance and the finite load-bearing capability of continuum robots, integrating multiple actuations poses challenges in modeling and control. This article introduces a dynamics-based robust uncertainty estimation and control (DRUEC) method for hybrid-driven continuum robots with continuously variable stiffness to tackle the fast internal dynamic variations and enhance motion tracking accuracy. Initially, a conventional kinematic formula is established to transfer all local force and position vectors into the global coordinate. Subsequently, the Euler-Lagrange methodology is employed to construct the entire dynamic model within the actuation space. For improving programming efficiency and independence from system parameter identification technologies, explicit expressions of matrices in the constructed dynamic model are derived by using the chain rule and properties of homogeneous coordinate transformation. In addition, a novel robust uncertainty estimator (RUE) is proposed to estimate modeling errors arising from the unmodeled dynamics and parameter perturbations. Various experiments are implemented based on a hybrid-driven continuum robot with two segments. Comparative results show the effectiveness of the proposed scheme over the classical methods. Note to Practitioners—The motivation for this article is to enhance the motion tracking accuracy of hybrid-driven continuum robots. Due to their inherent compliance, continuum robots exhibit a great adaptation capability to restricted environments. However, a conflict between intrinsic compliance and positioning accuracy of the endpoint limits their practical applications. To tackle this issue, the hybrid driving method offers an accessible solution by decoupling stiffness regulation from position adjustment. This enables different stiffness levels at the same posture, allowing the continuum robot to withstand varying loads without experiencing significant deformations. Nevertheless, the incorporation of multiple actuations also complicates modeling and control due to the increased coupling and nonlinearity. Moreover, the effective operation of continuum robots needs to timely deal with the effects induced by the rapid internal dynamic changes and relatively large movement speeds. These facts imply that the commonly used quasi-static models or kinematics-based methods are insufficient. Therefore, this article is dedicated to improving the control accuracy from the perspective of dynamics-based control approaches.}, archive = {J_TASE}, author = {Jun Yang and Edward Harsono and Haoyong Yu}, doi = {10.1109/TASE.2024.3496773}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9049--9060}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamics-based motion control for a hybrid-driven continuum robot with continuously variable stiffness}, volume = {22}, year = {2025}, }
• (2025). Multivariable adaptive super-twisting sliding mode resilient control for uncertain nonlinear CPSs against actuator and sensor attacks. TASE, 22, 9039--9048. (www)

  @article{ , abstract = {This paper proposes an adaptive super-twisting sliding mode resilient control method for uncertain nonlinear cyber-physical systems (CPSs) subject to actuator and sensor attacks. Two super-twisting algorithms with adaptive gains are designed based on the compromised system states to counteract the effect of cyber-attacks to ensure the system states converge to a small region near zero. Furthermore, the gain overestimation problem of the sliding mode resilient controller is also discussed. Finally, a formal analysis of the closed-loop system is derived using Lyapunov function techniques. Simulation results on a two-link manipulator system validate the effectiveness of the proposed resilient control strategy. Note to Practitioners—The motivation of this paper is to address the secure issues of the nonlinear cyber-physical systems, including applications in robotic manipulators, vehicle systems, and power systems. Since some devices of CPSs, such as sensor and controller components, communicate via networked channels in practice, the original system states and control signals of CPSs may be tampered by cyber-attacks, leading to unavailable system states and severe uncertainties with unknown bounds. Although the super-twisting sliding mode control (ST-SMC) method is an effective tool for dealing with uncertainties, the control gains may be overestimated due to the intermittent nature of cyber attacks, resulting in severe chattering caused by the discontinuous term of the ST-SMC method. Furthermore, it is extremely challenging when only compromised system states are available for the controller, potentially generating false commands and destroying the entire system. Therefore, based on Lyapunov theory, a novel multivariable adaptive sliding mode resilient control method with two adaptive super-twisting algorithms is proposed to mitigate the effects caused by cyber-attacks. The effectiveness of the proposed method is validated using a two-link manipulator. Note that the proposed method has the potential for broader engineering applications in real-world scenarios.}, archive = {J_TASE}, author = {Yannan Bi and Fei Wang and Pansheng Ding and Tong Wang and Jianbin Qiu}, doi = {10.1109/TASE.2024.3496755}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9039--9048}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multivariable adaptive super-twisting sliding mode resilient control for uncertain nonlinear CPSs against actuator and sensor attacks}, volume = {22}, year = {2025}, }
• (2025). Inverse optimal adaptive neural power regulation control for variable-speed wind turbines with load mitigation. TASE, 22, 9029--9038. (www)

  @article{ , abstract = {Traditional optimization control algorithms need to solve the Hamilton-Jacobi-Bellman (HJB) equation in real time, which requires a time-consuming training procedure and is difficult to apply in actual engineering facilities. To break this limitation, an inverse optimal adaptive neural pitch control scheme is first presented for variable-speed wind turbines (VSWTs) to achieve a comprehensive performance of power regulation and load mitigation in this paper. First of all, to facilitate the design of the controller, the non-affine VSWT model is converted into an affine model by employing the filter regulation error technology. Subsequently, the stability of the overall control system is demonstrated via the Lyapunov theory and the regulation error ultimately converges to an adjustable region near zero. Then, for a meaningful cost functional, the practical inverse optimality of the control system is realized. Finally, the OpenFAST (Fatigue, Aerodynamics, Structure and Turbulence) platform is utilized to compare the proposed scheme with existing pitch controllers on an NREL WP 1.5-MW three-blade VSWT to verify the feasibility of the presented scheme. Note to Practitioners—VSWTs are receiving increasing research attention due to their performance in maximizing power generation and mitigating effects of wind fluctuations. Under the turbulent wind condition, it has great practical significance to achieve stable power quality and alleviate fatigue loads on valuable components of VSWTs. How to improve the power quality and reduce the operation and maintenance cost of the turbine is a challenging problem. This paper presents an inverse optimal adaptive neural network control method for VSWTs under the high-speed turbulent wind condition.}, archive = {J_TASE}, author = {Shaodong Wang and Qinmin Yang and Shuzong Xie and Wenchao Meng}, doi = {10.1109/TASE.2024.3496743}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9029--9038}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Inverse optimal adaptive neural power regulation control for variable-speed wind turbines with load mitigation}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time fuzzy control for MIMO coupled systems with unknown structure and control direction: Application to robotic arm. TASE, 22, 9013--9028. (www)

  @article{ , abstract = {Due to model uncertainty, system coupling and external disturbance, the mathematical structure of control system is often constructed imprecisely. This unknown structure implies the unknown system dynamic and control input direction. Hence, it is difficult for unknown-structure Multiple Input Multiple Output (MIMO) coupled control systems to prescribe the settling time and control accuracy. To solve this problem, a novel prescribed-time robust direction-adjusting control method is proposed by adaptive Takagi-Sugeno (AT-S) fuzzy approximation. More specifically, by constructing an adaptive T-S fuzzy approximator under relaxed conditions, the mismatched uncertainty of systems can be transformed into the bounded matched uncertainty. It is a pre-condition for prescribed finite-time stability. After the prescribed time terminal, an adaptive robust control designed on AT-S fuzzy model is used for prescribed-accuracy convergence. Besides, compared with existing studies, the proposed method is independent of the structure information and initial control direction. Simulations and experiments verify its effectiveness. Note to Practitioners—The robot arm system with no or lowaccuracy Lagrange dynamic identification is a typical unknownstructure MIMO coupled system. It is difficult to achieve fastconvergence and high-accuracy control for this practical system, especially with no empirical pre-adjustment of the initial input direction. To solve this practical problem, a novel prescribedtime direction-adjusting adaptive T-S fuzzy control method is proposed. Firstly, the robotic arm system is decoupled into multiple joint subsystems by virtual dimension reduction strategy to avoid rule explosion, where system coupling is also converted to bounded unmatched uncertainty. Then, the inappropriate initial direction of the actuator input signal is addressed by control direction-adjusting algorithm. Meanwhile, based on adaptive TS fuzzy approximation theorem, the mismatched uncertainty is transformed into the bounded matched uncertainty. In the end, according to the prediction of approximation accuracy, a prescribed-accuracy robust controller is designed on AT-S fuzzy model. The proposed method is effective in a practical robotic arm control experiment without identification of Lagrange dynamics. Moreover, compared with other related methods, the control precision of the proposed method is higher.}, archive = {J_TASE}, author = {Wen Yan and Tao Zhao and Edmond Q. Wu}, doi = {10.1109/TASE.2024.3496699}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9013--9028}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time fuzzy control for MIMO coupled systems with unknown structure and control direction: Application to robotic arm}, volume = {22}, year = {2025}, }
• (2025). A tube-NMPC approach for robust control of glucose in type 1 diabetes mellitus. TASE, 22, 9001--9012. (www)

  @article{ , abstract = {Blood glucose concentration (BGC) is a significant indicator of human health and plays an essential role in maintaining a healthy life. Type 1 diabetes mellitus (T1DM) is an immune-mediated deficiency disease that causes a substantial deficit in insulin secretion. However, hyperglycemia can lead to a range of complications. Additionally, uncertainties such as model mismatch, exercise, and stress can affect BGC, impacting overall health. To address these challenges, this study advocates a data-driven tube-based nonlinear model predictive control (Data-Driven Tube-NMPC) approach. Firstly, a genetic algorithm-based long short-term memory (GA-LSTM) neural network is formulated to build a prediction model, where its hyperparameters are optimized via a genetic algorithm (GA). Secondly, a tube-based nonlinear model predictive control (Tube-NMPC) strategy is proposed to introduce the concept of a tube, tightening constraints to reduce model mismatch and mitigate the effects of factors such as individual differences, exercise, and stress. Additionally, the feasibility, constraint satisfaction, and stability of the algorithm are analyzed. Finally, simulation results demonstrate that, compared to conventional model predictive control (MPC), this approach more effectively controls the BGC of T1DM patients in the presence of perturbations. Note to Practitioners—Due to the model mismatch problem and factors such as individual difference, exercise and pressure, the robust control of BGC during the human blood glucose process by artificial pancreatic system (APS) will be affected to some extent. In order to achieve the robust control of BGC in a limited time, this paper designs a Tube-NMPC strategy. The proposed control scheme effectively regulates BGC in T1DM patients after food intake, ensuring both safety and timeliness. In practice, practitioners should obtain accurate data. The prediction accuracy is improved by automatically determining the hyperparameters of the LSTM by a GA. Secondly, a Tube-NMPC algorithm is employed to incorporate the notion of tubes, thereby constricting constraints to mitigate the impact of uncertainty. Subsequently, the efficacy of the proposed control strategy is substantiated by validating the life trajectory of a patient diagnosed with T1DM.}, archive = {J_TASE}, author = {Runda Jia and Xiaoyi Zhao and Shulei Zhang and Xia Yu}, doi = {10.1109/TASE.2024.3494819}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {9001--9012}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A tube-NMPC approach for robust control of glucose in type 1 diabetes mellitus}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered formation control of multi-agent systems with non-uniform time-varying communication delays. TASE, 22, 8988--9000. (www)

  @article{ , abstract = {In this study, we address the challenge of time-varying formation control in multi-agent systems (MASs) in the presence of time-varying intra- and inter-agent communication delays. To tackle time-varying delays, we equip each agent with a bank of distributed observers to estimate its own and its neighbors’ states. We apply dynamic periodic event-triggered mechanisms to both sensor-to-observer (S-O) and controller-to-actuator (C-A) channels, aiming to reduce unnecessary data transmissions in the network by relying on locally triggered sampled data in a distributed fashion to enhance resource efficiency. In the design stage, we transform the state formation control problem into an asymptotic stability problem. Using the Lyapunov-Krasovskii functional (LKF) approach, we design the event-triggering parameters such that the closed-loop system of all agents is stable and agents reach the desired formation. Numerical simulations demonstrate that our approach achieves a balance by reducing inter-agent communication frequency while maintaining the desired formation. Finally, we illustrate the effectiveness and advantages of this approach through experiments on a real-world robotic system. Note to Practitioners—In practical applications of multi-agent systems, the use of a communication network introduces some challenging issues. To name a few, periodic sampling with a high frequency relies on heavy transmission of information between components, which may result in network congestion. Factors such as limited bandwidth, signal attenuation, and packet losses contribute to delays in networked MAS. Additionally, network security, protocols, buffering, processing, and transmission times play significant roles. Since network-induced delays depend heavily on variable network conditions, they are generally non-uniform and time-varying. This paper proposes a solution for formation control in MASs, considering communication delays, and holds practical implications across various industries. It can enhance coordination for tasks such as warehouse logistics and collaborative manufacturing in autonomous robotics. Drone swarms can benefit from more efficient and reliable movement coordination, impacting surveillance and precision agriculture. In industrial automation, synchronization among machines or robotic arms can be improved for increased efficiency. A noteworthy aspect of this paper is the validation of our results through experiments on a real-world multi-robot system, demonstrating broad applicability.}, archive = {J_TASE}, author = {Milad Abbasi and Horacio J. Marquez}, doi = {10.1109/TASE.2024.3494658}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8988--9000}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered formation control of multi-agent systems with non-uniform time-varying communication delays}, volume = {22}, year = {2025}, }
• (2025). Modeling of dynamic systems with hysteresis using predictive gradient-based method. TASE, 22, 8976--8987. (www)

  @article{ , abstract = {A new modeling method of dynamic systems with rate-dependent hysteresis is proposed in this paper. In this method, a hysteresis model with simple exponential structure is proposed to describe the features of rate-dependent hysteresis. Subsequently, the properties of the proposed hysteresis model are analyzed. Then, a Hammerstein model embedded with the proposed hysteresis model is established to describe the behavior of dynamic systems with rate-dependent hysteresis. Afterward, a predictive gradient-based modeling method is proposed to determine the parameters of the new model. In addition, the convergence analysis of the predictive gradient based modeling method is analyzed. Then, the proposed identification method is applied to modeling of electromagnetic scanning micromirror chips. Finally, the comparison between the proposed novel modeling scheme and other typical nonlinear modeling methods is illustrated. Note to Practitioners—To describe the characteristics of rate-dependent hysteresis in electromechanical systems, both rate-dependent hysteretic operator-based models and non-smooth differential equation-based hysteresis models have complex model structures. However, the exponential-type hysteresis model proposed in this paper not only has a simple structure but can also describe the more complex characteristics of rate-dependent hysteresis. In addition, for rate-dependent hysteresis with multiple local extremes, the proposed modeling method based the predictive gradients can avoid the modeling process being stuck in local extremes, thereby obtaining fast convergence and accurate modeling results.}, archive = {J_TASE}, author = {Guo Chai and Yonghong Tan and Qingyuan Tan and Ruili Dong and Changzhong Ke and Ya Gu and Tianyu Wang}, doi = {10.1109/TASE.2024.3494596}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8976--8987}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling of dynamic systems with hysteresis using predictive gradient-based method}, volume = {22}, year = {2025}, }
• (2025). Learning anticipatory decision for distributed systems with robustness guarantees. TASE, 22, 8965--8975. (www)

  @article{ , abstract = {This paper investigates anticipatory decision for unknown distributed systems with robustness concerns. Anticipatory decision focuses on action selection before observations appear at temporal scales. Firstly, anticipatory decision forms sequential feedback with min-max performance guarantees, while causality comes from time series analysis. Next, distribution, robustness and time consistency partition the optimization into spatial and temporal sub-games. The spatial sub-games dispel conflicts on distribution and robustness, while the temporal ones ensure stability and performance through time consistency. Finally, we propose a multi-step reinforcement learning algorithm under causality analysis and game theoretical framework. Numerical results demonstrate the effectiveness of the approach, and practical experiments show potential real-world applications. Note to Practitioners—This framework focuses on anticipatory decision for distributed systems, which suffer from distributed communication, unknown dynamics, environmental disturbances and state observation loss. Our framework has various application scenarios, e.g., internal surgical robots, low-light autonomous driving and non-GPS navigation, and these scenarios mainly involve dynamic environments and weak signal feedback. For example, decision-making in autonomous driving requires not only reacting to current environmental conditions but also anticipating future scenarios and uncertainties due to poor visibility. Most results deal these issues with model-driven approaches, while unknown dynamics render these methods inapplicable. For implementation, we propose a multi-step reinforcement learning algorithm for anticipatory decision framework with stability and robustness guarantees, and details mainly contain three parts: 1) We collect data during offline phase, and form the data structure, namely, current-next observation pair with multi-step decision and accumulated reward; 2) Strategies and value functions are approximated with neural networks through Monte-Carlo methods; 3) The strategy is deployed as sequential feedback in practical systems, and predicts multi-step decisions with single-step state observation. Finally, we select robot consensus with optical sensors as the implementation demo.}, archive = {J_TASE}, author = {Peijiang Liu and Xindi Yang and Hongliang Ren and Hao Zhang and Zhuping Wang}, doi = {10.1109/TASE.2024.3493912}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8965--8975}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning anticipatory decision for distributed systems with robustness guarantees}, volume = {22}, year = {2025}, }
• (2025). A learning-based assembly sequence planning method using neural combinatorial optimization with satisfactory generalization ability. TASE, 22, 8952--8964. (www)

  @article{ , abstract = {This paper proposes a specific and effective real-time sequence planning method using robot manipulators to complete complex assembly tasks. Many previous studies developed different traversal methods to obtain the optimal assembly sequence. Besides, a number of algorithms were proposed to enhance flexibility when the conditions or rules were changed in various sequence optimization problems. However, these state-of-the-art (STOA) methods necessarily require modifications when task details are changed. Consequently, to further improve the generalization ability and improve the performance of the sequence optimization, a neural combinatorial optimization algorithm combined with a self-learning strategy is proposed for assembly sequence planning. In addition, obstacle avoidance and the non-collision constraints between workpieces in the assembly process are considered. According to the experiment results, the new method is superior to the STOA methods in terms of optimization efficiency. More importantly, the proposed method has satisfactory generalization ability for different assembly tasks.Note to Practitioners—This paper studies assembly sequence planning problems for different real-world applications in industrial and home service fields. Many assembly sequence planning solutions have been widely utilized before. However, the generalization ability of the previous methods is not satisfactory since the re-adjust process is required when the workpiece number or collision condition changes in different tasks.Motivated by the above reasons, this paper develops a learning-based assembly sequence planning solution to resolve complex assembly problems without parameter re-adjustment processes. Users can directly apply the developed workpiece identification and localization method to obtain the sensing information. Then, the newly designed collision-free cost function should be programmed as the core of the assembly sequence optimization. Next, the proposed neural combinatorial optimization (NCO) with the sensing information and target configuration as inputs can provide the optimal assembly sequence by self-learning. The learned NCO-based method can be directly applied to diverse planning tasks, even with different workpiece numbers. Users can also refer to the experimental examples in this paper for the extension of the proposed method to their own applications.}, archive = {J_TASE}, author = {Ruiming Hou and Sheng Xu and Chenguang Yang and Jianghua Duan and Xinyu Wu and Tiantian Xu}, doi = {10.1109/TASE.2024.3493617}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8952--8964}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A learning-based assembly sequence planning method using neural combinatorial optimization with satisfactory generalization ability}, volume = {22}, year = {2025}, }
• (2025). Secure distributed adaptive control of nonlinear multi-agent systems. TASE, 22, 8937--8951. (www)

  @article{ , abstract = {This paper addresses the problem of secure consensus tracking control for nonlinear leader-follower multi-agent systems, with a specific focus on safeguarding followers’ private information from external network eavesdroppers or internal untrusted neighbors. To tackle this problem, we first employ the dynamic linearization approximation technique to transform the nonlinear system models into equivalent linear forms that involve unknown time-varying pseudopartial derivatives. Then, a novel model-free secure distributed adaptive control (MFSDAC) framework is proposed using an encoding-decoding mechanism and Paillier encryption. Within this framework, we develop a secure distributed control scheme using a recursive form and design an adaptive updating law with a modified projection to estimate the time-varying pseudoparial derivatives. To enhance security, we introduce an adjustable parameter and a random integer into the distributed communication protocol, effectively preventing the disclosure of followers’ data during both network transmissions and controller evaluations. Additionally, parameter selection rules for the controller, quantizer, and adaptive updating law are provided, along with convergence analysis and guarantees against quantizer saturation. Finally, numerical simulations confirm that the proposed MFSDAC framework successfully achieves leader-following tracking and secure data transmission, even in the presence of external network eavesdroppers or internal untrusted neighbors. Note to Practitioners—Multi-agent systems (MASs) provide a versatile framework for modeling and understanding various real-world applications, including autonomous systems, traffic management, and distributed sensor networks. Designing effective control strategies for MASs is essential to enhance cooperation and coordination among agents, strengthen system-level resilience and adaptability, and tackle complex tasks that surpass the capabilities of individual agents. A key challenge in this area is ensuring network security and protecting individual privacy, especially in the presence of external eavesdroppers and untrusted internal neighbors. To tackle this challenge, we propose a model-free secure distributed adaptive control framework for nonlinear leader-follower MASs. This framework incorporates a confidential communication protocol that leverages homomorphic encryption to protect sensitive information. The proposed framework has undergone rigorous stability analysis and been validated through numerical simulations, demonstrating its feasibility and effectiveness in achieving secure consensus control of nonlinear MASs.}, archive = {J_TASE}, author = {Yongxia Shi and Ehsan Nekouei}, doi = {10.1109/TASE.2024.3493136}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8937--8951}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure distributed adaptive control of nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Security control for cyber-physical systems with optimal dynamic stealthy actuator attacks. TASE, 22, 8925--8936. (www)

  @article{ , abstract = {This paper studies the security control problem for a class of cyber-physical systems with optimal dynamic stealthy actuator attacks. The minimum principle is used to design actuator attacks which satisfy an objective function presented in the paper. The ellipsoidal outer approximation is adopted to quantify the potential impact of stealthy actuator attacks. Meanwhile, the attack-free performance is guaranteed through the L $\infty $ performance index function. It provides an analysis tool based on the linear matrix inequality technique to redesign the state feedback matrix and the observer gain matrix of the plant, which ensures that the volume of its ellipsoidal outer approximation is minimal while meeting the L $\infty $ performance index function. Eventually, the simulation example is presented to demonstrate the effectiveness of the proposed control algorithm. Note to Practitioners—The security control problem is considered for discrete-time cyber-physical systems by using an ellipsoidal outer approximation. The proposed control scheme not only solves the problem of optimal dynamic actuator attacks, but also quantifies the potential impact of stealthy attacks. In addition, simulation results are given. In future research, we will address the control problem of continuous-time cyber-physical systems with dynamic stealthy attacks and its practical application.}, archive = {J_TASE}, author = {Kangkang Sun and Xiaochun Zhang}, doi = {10.1109/TASE.2024.3492727}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8925--8936}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Security control for cyber-physical systems with optimal dynamic stealthy actuator attacks}, volume = {22}, year = {2025}, }
• (2025). Sawtooth-characteristic-based resilient control design for wireless networked control systems under denial-of-service attacks. TASE, 22, 8912--8924. (www)

  @article{ , abstract = {The resilient control design for wireless networked control systems (WNCSs) under Denial-of-Service (DoS) attacks is investigated based on the sampled-data theory in this study. Firstly, by employing a logic processor module in the control center, WNCSs under DoS attacks are modeled as aperiodic sampled-data systems (SDSs). This integrates the duration of DoS attacks into the sampling intervals, simplifying the controller design without the need for additional handling of DoS attacks. Then, a sawtooth-characteristic-based hierarchical integral inequality is presented to leverage the underused sampling sawtooth characteristic in existing studies in WNCSs under DoS attacks. Next, a high-order two-sided looped-functional caters for the proposed hierarchical integral inequality is constructed and the sawtooth-dependent free-weighting matrices are introduced into the constraint of system equation to further preserve the information of SDSs. These improvements of methodological promote a reduction in the conservatism of the stability criteria. Consequently, a resilient controller design scheme is presented based on the system substitution-based de-coupling method, resulting in a controller that exhibits enhanced counteraction against DoS attacks. The efficacy of this controller design scheme is demonstrated through a satellite control system and a load frequency control power system, showcasing the advantages of the resilient controller. Note to Practitioners—This research offers significant insights into enhancing the resilience of WNCSs against DoS attacks, which is a prevalent challenge in industrial network security. The developed solution is particularly beneficial in industries where maintaining uninterrupted control communications is critical, such as satellite systems and power systems that are vulnerable to malicious DoS attacks. The logic processor module provides a practical approach to dynamically monitor the strength and duration of DoS attacks. Based on this, the relevant system is re-modeled as an aperiodic SDS. This integration with the existing sampling theories ensures that the system makes good use of current technologies and facilitates the design of resilient controllers that can withstand DoS attacks. On the one hand, the application of this enhanced control design allows for a more robust system that maintains functionality under cyber attack conditions, thereby protecting critical operations and potentially reducing downtime and associated costs. On the other hand, this approach reduces the conservatism of controller design guidelines, allows for more efficient utilization of network resources, and improves system response. However, this approach relies heavily on the accuracy and timeliness of the logic processor’s response to DoS attacks. This may not be entirely reliable in environments with highly irregular or complex network threats.}, archive = {J_TASE}, author = {Ying Zhang and Xing-Chen Shangguan and Yong He and Chen-Guang Wei and Chuan-Ke Zhang}, doi = {10.1109/TASE.2024.3492180}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8912--8924}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sawtooth-characteristic-based resilient control design for wireless networked control systems under denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Estimating lyapunov region of attraction for robust model-based reinforcement learning USV. TASE, 22, 8898--8911. (www)

  @article{ , abstract = {This article addresses the robustness of unmanned surface vehicles (USV) using model-based reinforcement learning (MBRL). A novel MBRL approach, Lyapunov probabilistic model predictive control (LPMPC) is proposed to simultaneously learn both the probabilistic model of a USV and its corresponding estimated Lyapunov region of attraction (ROA) under one reinforcement learning framework. Unlike the existing MBRL USV systems with less consideration of robustness and safety, our method naturally learns a general indicator of system stability based on the probabilistic model’s belief and employs it to guide its policy. Evaluated by different navigation tasks in a simulation driven by real boat data, LPMPC demonstrated significant advantages in both control robustness and task completion against various levels of environmental disturbances compared with the baseline approach without Lyapunov ROA’s guidance. Note to Practitioners—Modelling the system stability without human prior knowledge is challenging in the domain of USV. This work proposed a data-driven method to iteratively learn a task-relevant stability model of USV in a probabilistic view. Based on the evaluation of a real boat data-driven simulation, the learned stability model contributed to superior driving skills in different USV scenarios by properly indicating and avoiding potentially risky states. In future research, we plan to expand the definition of risks in different tasks, such as loss of control, overlarge sway, and excessive energy consumption and investigate the proposed approach in real-world USV.}, archive = {J_TASE}, author = {Lei Xia and Yunduan Cui and Zhengkun Yi and Huiyun Li and Xinyu Wu}, doi = {10.1109/TASE.2024.3492174}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8898--8911}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Estimating lyapunov region of attraction for robust model-based reinforcement learning USV}, volume = {22}, year = {2025}, }
• (2025). Novel sliding innovation filter inspired fault detection for hydrofoil attitude control systems. TASE, 22, 8886--8897. (www)

  @article{ , abstract = {In this paper, a novel approach for detecting anomalies in the non-linear fully-submerged hydrofoil attitude control system (HACS) is proposed, even in the presence of time-varying disturbances. To address the trade-off between robustness against disturbances and optimality in terms of estimation error, the extended sliding innovation filter (SIF) is employed as a state estimator for the target non-linear HACS. By utilizing a switching gain with a sliding boundary layer, the SIF inherently possesses a degree of robustness to estimation issues that may involve fault conditions or factors of disturbances. A residual framework is subsequently established to achieve state tracking and comparison. The residual evaluation for the fault detection (FD) scheme is then easily conducted using statistical methods such as the modified Z-Score and the peak signal-to-noise ratio (PSNR). Finally, the effectiveness of the developed FD strategy is substantiated through experiments conducted on a hardware-in-loop (HIL) platform. Comparative analysis with state-of-the-art robust UKF algorithms reveals the impressive fault detection proficiency of the proposed strategy. Note to Practitioners—This paper was motivated by the challenges of state estimation and FD for hydrofoil crafts under stochastic ocean wave disturbances. The extended SIF-based approach enhances robustness in the estimation and FD against disturbances by introducing a dynamic layer. Moreover, the adaptive layer indicates system anomalies as significant changes, which can assist engineers in promptly identifying anomalies. Furthermore, the modified statistics used in the FD scheme effectively reduce interference in the results. The developed FD method is easy to implement without linearizing the non-linear target system. The experimental validation conducted on the dSPACE platform utilizing the PCH-1 model illustrates the practicality of the proposed strategy for pertinent practitioners.}, archive = {J_TASE}, author = {Tao Wang and Dezhi Xu and Bin Jiang and Peng Shi and Levente Kovács}, doi = {10.1109/TASE.2024.3492040}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8886--8897}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Novel sliding innovation filter inspired fault detection for hydrofoil attitude control systems}, volume = {22}, year = {2025}, }
• (2025). Generalized distributed optimal coordination for multiagent systems via weak coupling hierarchical control framework. TASE, 22, 8870--8885. (www)

  @article{ , abstract = {In this article, we reformulate the distributed optimal coordination problem for multi-agent systems to broaden its applicability across a wider range of coordination scenarios, thereby introducing a Generalized Distributed Optimal Coordination (GDOC) problem. In GDOC, the inter-agent relationships evolve from equality (consensus) to affinity (coordination), while local cost functions are unified as blends of parameters and shared basis functions, enabling cohesive network optimization. To address the GDOC problem, we propose a weak coupling hierarchical control framework for heterogeneous multi-agent systems. This framework consists of three layers: a signal generator, a tracking controller, and a speed regulator. For the generator, a transformed consensus protocol is designed for agents to estimate the global cost function and feasible set in a distributed manner, with the gradient projection method applied to minimize the objective function locally. For the controller, an observer-based output feedback control law is designed through system decomposition. For the regulator, a dynamic adaptive parameter is introduced to adjust the updating speed of the reference signal based on the agent’s relative tracking ability. The proposed framework not only preserves the universality of hierarchical control but also addresses the limitation of topdown structural open-loop control by introducing a regulator to form a bottom-up feedback loop. Finally, the effectiveness of the proposed framework is verified by Lyapunov stability theory analysis and simulation experiments. Note to Practitioners—In numerous task scenarios, the coordinated control of multi-agent systems involves solving optimization problems. This paper proposes GDOC to mathematically characterize these scenarios in a unified way, with the goal of controlling each agent’s output to converge towards the minimum point of the aggregate cost functions while maintaining preset inter-agent relationships. To achieve this, the affine transformation matrix is introduced to describe these inter-agent relationships under diverse coordination scenarios. Furthermore, the cost function adopts a form involving parameters and shared basis functions, facilitating interaction and iteration within the cost function. Correspondingly, an engineering-friendly control framework is proposed to address the GDOC problem. This framework consists of a reference signal generator, a tracking controller, and a speed regulator, each of which can be designed separately. The speed regulator is a new addition, aimed at adjusting the updating speed of signals according to the physical dynamic response capability of each agent, thereby establishing an indirect bottom-up feedback loop. This control framework can be applied to addressing GDOC problems in situations with switching cost functions and multiple solutions.}, archive = {J_TASE}, author = {Yu Feng and Fuyong Wang and Zhongxin Liu and Fei Chen}, doi = {10.1109/TASE.2024.3492018}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8870--8885}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Generalized distributed optimal coordination for multiagent systems via weak coupling hierarchical control framework}, volume = {22}, year = {2025}, }
• (2025). A new double-integration-enhanced RNN algorithm for discrete time-variant equation systems with robot manipulator applications. TASE, 22, 8856--8869. (www)

  @article{ , abstract = {Discrete time-variant equation systems represent a typical and complex problem across various disciplines. With the increasing complexity of systems in various fields, traditional methods have been unable to effectively deal with the current discrete time-variant equation systems, especially in the dynamic engineering problem. Generally speaking, traditional methods are typically limited to considering discrete time-variant equation systems in ideal state, and there is a lack of deep research about more intricate disturbance states. This paper introduces a new recurrent neural network (RNN) algorithm, termed the discrete-time double-integration-enhanced RNN (DT-DIE-RNN) algorithm, for handling discrete time-variant equation systems (including discrete time-variant linear and nonlinear equation system) under discrete square-time-variant disturbance. Firstly, the continuous-time double-integration-enhanced RNN (CT-DIE-RNN) algorithm is presented for solving discrete time-variant linear and nonlinear equation systems by using double-integral-type error function. Secondly, the corresponding discrete-time RNN algorithm is presented, and the convergence and precision of such an algorithm are theoretically analyzed. Finally, the effectiveness and superiority of the proposed DT-DIE-RNN algorithm for solving discrete time-variant linear and nonlinear equation systems are supported by comparative numerical experiments, and these theoretical results are further verified by robot manipulator applications.Note to Practitioners—Generally speaking, previous algorithms are unable to guarantee the operation of robot manipulator stably with discrete square-time-variant disturbance. In this study, we design a new RNN algorithm which possesses stronger anti-disturbance capability for solving discrete time-variant equation systems, and such the algorithm is applied to the tracking control of robot manipulator. In the discrete-time environment, in order to solve the discrete time-variant problem smoothly and stably, it is necessary to ensure that the error in each time period is within an acceptable range. The basic guidelines of this paper are as follows. First of all, a double-integration RNN algorithm is constructed, i.e., discrete-time double-integration-enhanced RNN (DT-DIE-RNN) algorithm. Then the effectiveness and superiority of the DT-DIE-RNN algorithm are verified by numerical experiment and engineering simulation, respectively. In summary, the proposed RNN algorithm can be seen as a new breakthrough in the research field of discrete-time RNN algorithm.}, archive = {J_TASE}, author = {Yang Shi and Wei Chong and Xinwei Cao and Chao Jiang and Ruxin Zhao and Dimitrios K. Gerontitis}, doi = {10.1109/TASE.2024.3491913}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8856--8869}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A new double-integration-enhanced RNN algorithm for discrete time-variant equation systems with robot manipulator applications}, volume = {22}, year = {2025}, }
• (2025). Probability-based force control for flexible ureteroscopy robots. TASE, 22, 8844--8855. (www)

  @article{ , abstract = {In robotic flexible ureteroscopy, great care should be taken to prevent excessive collision force at the ureteroscope tip to avoid urinary tissue damage. However, the passive flexure of a flexible ureteroscope from reaction forces with the surrounding tissue introduces great uncertainties in force control. To address this, a probability-based force control method is proposed in this paper. A morphological wavelet-based statistic test is proposed to detect collision by identifying the change point of the probability distribution of force signal, which is measured by a fiber optical sensor at the ureteroscope tip. The force signal and its change points are applied as an admittance model inputs to generate position command. It is augmented to the robot controller for minimizing the collision force. Meanwhile, a probabilistic model approximates the real interactive system, and combines Bayesian inference for safely online learning the admittance parameters. Experimental results on a flexible ureteroscopy robot show that collision can be detected instantly, and the force is reduced remarkably during the advancement of the robotic ureteroscope. Note to Practitioners—This paper is motivated by the rapidly increasing applications for robotic flexible ureteroscopy. As ureteroscope enters into urethra, urinary bladder, ureter, and reaches renal pelvis, it sometimes penetrates soft tissue, especially in the case of no ureteral access sheath placement. It is important to control the ureteroscope tip force for surgery safety. Therefore, a force control method is proposed in this paper. Experiments conducted on a flexible ureteroscopy robot demonstrate that the proposed method is a protective measure to mitigate the risk of inner surface perforation.}, archive = {J_TASE}, author = {Yinan Deng and Tangwen Yang and Yuelin Zou and Jianchang Zhao and Jianmin Li and Bin Yao and Guoli Song}, doi = {10.1109/TASE.2024.3491139}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8844--8855}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Probability-based force control for flexible ureteroscopy robots}, volume = {22}, year = {2025}, }
• (2025). Lexicographic optimization-based priority ascending strategy for feasibility judgment and soft constraint adjustment. TASE, 22, 8828--8843. (www)

  @article{ , abstract = {Feasibility judgment and soft constraint adjustment play vital roles in process optimization and control. Priority ascending strategy is one of the most widely used methods to deal with feasibility analysis problem and has been popularly deployed in many industrial commercial control software including DMC3 of AspenTech and RMPCT of Honeywell. However, our recent research has identified that multiple optimal solutions may exist during the feasibility analysis of a specific priority and current methods lack the capability to retain all these solutions for subsequent priorities’ optimization. Motivated by this, our work proposes a novel lexicographic optimization-based priority ascending strategy for feasibility judgment and soft constraint adjustment. Firstly, we theoretically demonstrated that multiple optimal solutions could exist during the feasibility judgment and soft constraint adjustment of a specific priority. An illustrative example is provided to validate this theoretical finding. Subsequently, an inequality equation group which can encompass all multiple optimal solutions is introduced into the optimization process of priority ascending strategy. This ensures that all optimal solutions at a specific priority can be preserved in subsequent priorities’ optimization. Finally, an inequality equation group covering all optimal solutions of the last priority’s optimization is applied to the subsequent economic self-optimization and target tracking, guaranteeing the optimality and completeness of the entire steady-state optimization process. The proposed approach effectively minimizes the relaxation of controlled variable constraints, enhancing safety and maximizing economic profits. Extensive experimental validation confirms the effectiveness and reliability of the proposed approach.Note to Practitioners—The two-layer model predictive control finds extensive application in practical process control software like DMCPlus and DMC3 of AspenTech and RMPCT of Honeywell, demonstrating reliability and practicality. However, from a theoretical analysis standpoint, this algorithm exhibits minor deficiencies. This paper highlights a small specific issue termed “multi-solution incompleteness” and offers a simple solution. Addressing this problem could notably enhance the performance of relevant control software and elevate the safety and reliability of controlled plants. This paper is anticipated to interest researchers focusing on control theory and applications, along with engineers dedicated to utilizing practical control software.}, archive = {J_TASE}, author = {Jianbang Liu and Yaqing Jv and Zhaowei Wang and Yi Zhang and Haojie Sun and Hongyu Zheng}, doi = {10.1109/TASE.2024.3490620}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8828--8843}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lexicographic optimization-based priority ascending strategy for feasibility judgment and soft constraint adjustment}, volume = {22}, year = {2025}, }
• (2025). Robust active fault diagnosis for linear stochastic systems within bayesian decision framework. TASE, 22, 8818--8827. (www)

  @article{ , abstract = {This paper is concerned with the active fault diagnosis (AFD) problem for a class of linear stochastic systems with unknown prior probabilities of modes. In AFD within the Bayesian decision framework, the unknown prior probabilities typically make diagnosis performance compromised. A novel robust AFD approach is developed to settle the AFD problem under uncertainties in prior probabilities of modes for linear stochastic systems. The input design goal is to minimize the misdiagnosis probability on the basis of the robustness for uncertainties in prior probabilities of modes. The input design problem is characterized as a min-max optimization problem. As there is no closed-form expression for the misdiagnosis probability, its a novel upper bound family is deduced as an alternative. The extensively utilized Bhattacharyya upper bound is proven to be included in this upper bound family. The tightness of the upper bounds in the proposed family is analyzed. A novel two-layer optimization method is proposed to solve the input design problem to global optimality. Two numerical examples are presented to illustrate the effectiveness and superiority of the developed robust AFD approach. Note to Practitioners—The majority of the existing stochastic AFD results are developed within the Bayesian decision framework, where prior probabilities of system modes are required to be known. However, the prior probabilities are difficult to obtain in practice, which makes these AFD approaches impractical. A major challenge in stochastic AFD is to evaluate the probability of misdiagnosis due to the presence of multivariate integrals in its expression. An extensively utilized solution is to exploit its Bhattacharyya upper bound as an alternative. Unfortunately, there is looseness in the Bhattacharyya upper bound under certain circumstances, which is likely to make the misdiagnosis probability evaluated inaccurately and render diagnosis performance unsatisfactory. In this paper, the authors develop a novel robust AFD approach with a new upper bound family on the misdiagnosis probability to improve diagnosis performance for linear stochastic systems with unknown prior probabilities of modes. The developed AFD approach is robust for uncertainties in prior probabilities of modes. The input design problem is characterized as a min-max optimization problem, which is solved to global optimality by a novel two-layer optimization method. The upper bounds in the proposed family are tighter than the Bhattacharyya upper bound under certain conditions. For nonlinear stochastic systems with unknown prior probabilities of modes, the proposed robust AFD may not be applicable. In future research, we will focus on the robust AFD problem for nonlinear stochastic systems with unknown prior probabilities of modes.}, archive = {J_TASE}, author = {Yaqi Guo and Xiao He}, doi = {10.1109/TASE.2024.3489720}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8818--8827}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust active fault diagnosis for linear stochastic systems within bayesian decision framework}, volume = {22}, year = {2025}, }
• (2025). Defending UAV networks against covert attacks using auxiliary signal injections. TASE, 22, 8805--8817. (www)

  @article{ , abstract = {Unmanned aerial vehicle (UAV) networks, which carry vital information, are prone to various attacks, and hence security issues are a major concern. In this paper, we design and implement a novel covert attack detection and secure control scheme, which operates between the UAV (the physical layer) and ground control station (GCS) (the cyber layer). Covert attacks can alter the UAV states, and yet cause the signals seen by the controller to appear unchanged, resulting in these attacks being more difficult to detect, and hence more dangerous compared to other types of attacks. To unmask the covert attacks, we construct and inject auxiliary signals to both the controller output and the UAV input. The auxiliary signals cause information of the attack to appear in the controller input, which is then fed to a detection observer to detect the attack. Next, we propose an integrated estimation and secure control scheme, comprising a reconstruction observer (which is a sliding mode observer (SMO)) that estimates the system states and attack signal, and an output-feedback controller that utilizes the estimated signals. We perform a series of transformations to the system, such that the design parameters of both reconstruction observer and secure controller are placed in a framework that is solvable using Linear Matrix Inequalities (LMIs). We also prove that the proposed integrated secure controller causes the output tracking errors to satisfy an ${{\mathcal {H}}}_{\infty }$ performance index. We also rigorously analyze the system performance, and present the necessary conditions for the scheme to be feasible. Finally, simulations are conducted to verify the effectiveness of the proposed scheme. Note to Practitioners—This paper presents a method to detect covert attacks in the UAV network, and to mitigate against those attacks. Covert attacks are more malicious since they are difficult to detect. The proposed method in this paper consists of auxiliary signal injection and a detection observer that will expose and detect the attacks, and a reconstruction observer and secure controller that will estimate the attacks and mitigate its effect on the plant. The controller and observers are designed using Linear Matrix Inequalities to minimize the ${{\mathcal {H}}}_{\infty }$ gain from the attack on the plant performance. In addition, this paper also investigates the conditions that the plant must satisfy such that the proposed scheme is feasible, and presents them in an easily verifiable form. Finally, the paper also analyses the performance of the proposed scheme in all scenarios - namely before the attack occurs, when the attack occurs but is not yet detected, and after the attack is detected.}, archive = {J_TASE}, author = {Xianghua Wang and Chee Pin Tan and Youqing Wang and Xiangrong Wang}, doi = {10.1109/TASE.2024.3489609}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8805--8817}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Defending UAV networks against covert attacks using auxiliary signal injections}, volume = {22}, year = {2025}, }
• (2025). Head-dominant enhancement with local count for better human detection in crowds. TASE, 22, 8794--8804. (www)

  @article{ , abstract = {In crowded scenes, it is difficult to extract discriminating human features due to occlusion. Some human detectors have improved this issue by introducing head detection. However, a complex problem still exists in associating full-body detection with its corresponding head detection. Instead of learning the association, we propose a Head-dominant Enhancement Module (HDEM) that uses the full-body proposal to regress the head bounding box. To embed the head information into the target human feature, we further propose a Consistent Weighted (CW) loss. Additionally, existing Non-Maximum Suppression (NMS) algorithms do not consider density changes with the selection of detection boxes, which leads to false and missed detection. Similar to human visual habits in occluded scenarios, we propose a Count-aware Dynamic Threshold Module (CADTM) that utilizes head context information to predict local count, which is associated with crowd density. CADTM can solve the inherent defects of Greedy-NMS in crowded scenes by adjusting the Intersection over Union (IoU) threshold dynamically. Ultimately, through the combination of HDEM and CADTM, we achieve state-of-the-art performance on CrowdHuman with a small computational cost. Our method achieves 4.6% AP gains, 2.2% MR-2 gains, and 3.2% JI gains over a Cascade R-CNN baseline. Furthermore, the proposed method is flexible and can be used with most proposal-based detection frameworks and various IoU-based NMS. Note to Practitioners—The motivation for this study arises from a prevalent issue encountered in human detection applications, particularly in densely populated areas such as shopping malls, streets, and subway stations, where occlusion poses a significant challenge. Employing a generic object detector results in numerous missed detections, thereby significantly compromising the overall performance of the detector. This research proposes a convenient plugin that achieves significant performance improvements at a very small cost. To validate its efficacy, our proposed method is thoroughly evaluated on proposal-based detection frameworks. The experimental results demonstrate the robustness of our approach and its ability to adapt to diverse crowded scenarios.}, archive = {J_TASE}, author = {Shoudong Han and Huilin Ding and Zhiling Han and Heng Li}, doi = {10.1109/TASE.2024.3488856}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8794--8804}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Head-dominant enhancement with local count for better human detection in crowds}, volume = {22}, year = {2025}, }
• (2025). Fuzzy adaptive resilient output control for nonlinear multi-agent systems under byzantine agents. TASE, 22, 8783--8793. (www)

  @article{ , abstract = {This article studies the fuzzy adaptive resilient output control for nonlinear multi-agent systems (MASs) under Byzantine attacks. Firstly, fuzzy logic systems (FLSs) are adopted to approximate the unknown nonlinear dynamics, then a fuzzy state observer is constructed to identify the unmeasured states in the controlled system. The mean-subsequence-reduced (MSR) technique is introduced to filter the unavailable neighbor information of each agent. On the basis of the distributed backstepping technique, a special Nussbaum function is utilized to construct an attack compensation scheme. Combining the compensation scheme and MSR technique, an observer-based robust fuzzy adaptive resilient output control design algorithm is proposed. The developed resilient control method enables the system outputs follow a given protocol under Byzantine attacks. Finally, the feasibility and effectiveness of the proposed control method and theory can be illustrated via the simulation results. Note to Practitioners—In this article, an observer-based resilient fuzzy adaptive output control issue is studied for nonlinear multi-agent systems subjected to Byzantine attacks. Fuzzy state observer is constructed to identify the unmeasured states. Since there exist the Byzantine attacks in the controlled system, the neighbor information of each agent cannot be fully used in the controller design. Thus, compared with the existing related results, the mean-subsequence-reduced technique can be introduced in this article to filter the unavailable neighbor information of each agent. Besides, an attack compensation control method via a special Nussbaum function is designed to resist the effect of uncertainties and malicious data caused by attacks. Simulation results ensure the feasibility of the developed resilient control approach and theory.}, archive = {J_TASE}, author = {Kewen Li and Lanlan Song and Yongming Li}, doi = {10.1109/TASE.2024.3488820}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8783--8793}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fuzzy adaptive resilient output control for nonlinear multi-agent systems under byzantine agents}, volume = {22}, year = {2025}, }
• (2025). Power modulation enables reduced motor power requirement of ankle assistance exosuit. TASE, 22, 8773--8782. (www)

  @article{ , abstract = {Actuation efficiency is a significant consideration for soft exosuits. It has important significance for reducing system weight and improving the effectiveness of walking assistance. Inspired by a fact that desired assistance from exosuits varies over the gait stage, this study developed a power modulation unit (PMU) to adapt to human walking characteristics, aiming to reduce the power requirement of exoskeletons. The underlying principles consist of: 1) dividing the gait phase into assistance period and idle period by different gait characteristics, where the terminal stance phase is defined as the assistance period and other phases are defined as the idle period; 2) motor power is stored in elastic elements during the idle period with mechanical advantage; 3) an amplified power bursts during the assistance period. Preliminary experiments were carried out on an ankle exosuit with human users. Experimental results indicate that the proposed PMU can amplify the output power by 3.1 times while affecting little on wearers’ normal gait, which implies the potential of the developed PMU for use in wearable actuation systems. Note to Practitioners—This work was motivated by the trade-off between multipath assistance and lightweight actuation. That is, multipath assistance typically necessitates multiple motors, consequently adding to the overall weight of the system. In this work, we developed a PMU to adapt to human walking characteristics (motor power is stored in elastic elements during the idle period and released when needed), aiming to reduce the requirement for peak actuation power. Experimental results show that the PMUs can provide an average power amplification ability of 3.1 times and does not interfere with normal human gait. The developed PMU holds great potential in walking-assisted exosuits or other devices for human movement assistance.}, archive = {J_TASE}, author = {Mingming Zhang and Kaiqi Guo and Zhiyi Gao and Jianhuang Wu and Yi-Feng Chen and Mingjie Dong}, doi = {10.1109/TASE.2024.3488714}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8773--8782}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Power modulation enables reduced motor power requirement of ankle assistance exosuit}, volume = {22}, year = {2025}, }
• (2025). Observer-based finite-time fuzzy adaptive resilient control for uncertain nonlinear systems against deception attacks and unknown dead zones. TASE, 22, 8762--8772. (www)

  @article{ , abstract = {This study explores the issue of observer-based adaptive finite-time resilient control for a class of uncertain strict-feedback nonlinear systems that have unknown sensor deception attacks and unknown dead zones. Due to the states of the considered system being unmeasured, a new fuzzy state observer is constructed via compromised output and fuzzy logic systems (FLSs). Moreover, the adverse impact of the deception attacks lies in that the output information is distorted, and the exact output signal is unavailable for control design. Based on the Nussbaum gain technique, a novel fuzzy finite-time resilient adaptive output feedback control scheme is developed, which incorporates both the attacked and estimated state variables. The adaptive output feedback control results reveal that all signals in the overall system remain stable and bounded. Compared to the current state feedback control results for nonlinear systems besieged by sensor-based attacks, the resilient adaptive output feedback control approach has the advantage of addressing the control issue of nonlinear systems with unavailable state measurements. Additionally, it eliminates the need to assume that the sign of the attack weight is positive. Finally, the validity of the designed controller is demonstrated through a simulation example. Note to Practitioners—In the industry, adaptive finite-time output feedback control issue of nonlinear systems exists in many different systems, such as intelligent transportation management, mobile robot networks, surveillance and monitoring. Since the above systems operate in a network environment, the security problems of the systems cannot be ignored. Hence, considering the unmeasured states, the unknown nonlinear functions, and the unknown time-varying attack gains existing simultaneously in the studied systems, it is a challenging and meaningful task to achieve the desired security control objectives. On the other hand, leveraging the Nussbaum gain technique, this article presents a novel fuzzy finite-time resilient adaptive output feedback control scheme for the nonlinear systems under the deception attacks. This method offers a practical solution for industrial applications.}, archive = {J_TASE}, author = {Jipeng Zhao and Guang-Hong Yang}, doi = {10.1109/TASE.2024.3488693}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8762--8772}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based finite-time fuzzy adaptive resilient control for uncertain nonlinear systems against deception attacks and unknown dead zones}, volume = {22}, year = {2025}, }
• (2025). Adaptive multi-robot cooperative localization based on distributed consensus learning of unknown process noise uncertainty. TASE, 22, 8738--8761. (www)

  @article{ , abstract = {The unknown process noise covariance matrix (PNCM) problem inducing by poor calibration or time-varying environment has not been addressed in the 2-D multi-robot system. This problem will severely deteriorate the distributed cooperative localization consistency and accuracy, and is troublesome to solve due to small magnitude of the 2-D robot’s PNCM. In this paper, the above issue is addressed by the following two steps. Firstly, the motion model of the 2-D robot is reconstructed to form a more estimable PNCM, from which a small-scale PNCM estimation algorithm is derived. Then the cooperative strategy consisting of a Kullback-Leibler average strategy and a recovery strategy is proposed to guarantee global PNCM estimation consensus and convergence, even if only partial robots access absolute measurement information. Theoretical consensus and convergence analyses are presented and comprehensive simulation and experimental tests are conducted to verify the effectiveness and superiority of the proposed algorithm. Note to Practitioners—This work is motivated by the inaccurate PNCM problem existing in the 2-D homogeneous multi-robot system, whose PNCM is very small in magnitude. The pose accuracy of the 2-D mobile robot, which relies on the sensor precision, is generally not enough to estimate such small-scale PNCM. Most of the existing PNCM estimation algorithms are regarding to simple target tracking models whose PNCMs are relatively large in magnitude. Furthermore, a few small-scale PNCM estimation algorithms make crucial assumptions about the PNCM, which limits their practicality. This paper proposed a novel small-scale PNCM estimation algorithm and an efficient cooperative strategy to facilitate global PNCM estimation consensus and convergence, without making any assumptions about the PNCM. The consensus and convergence analyses are provided to further demonstrate the effectiveness of the proposed adaptive cooperative localization algorithm. The proposed algorithm has been evaluated via simulation, public dataset and physical experiment.}, archive = {J_TASE}, author = {Chao Xue and Han Zhang and Fengchi Zhu and Yulong Huang and Yonggang Zhang}, doi = {10.1109/TASE.2024.3488319}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8738--8761}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive multi-robot cooperative localization based on distributed consensus learning of unknown process noise uncertainty}, volume = {22}, year = {2025}, }
• (2025). Walk before you can run: Sampling-rate-aware sequential knowledge transfer for multirate process anomaly detection. TASE, 22, 8725--8737. (www)

  @article{ , abstract = {Anomaly detection plays a vital role in monitoring industrial processes. Despite the extensive development of deep anomaly detection approaches, many assume uniform sampling rates for process variables, yet multirate phenomena are common in practice. The resultant large-scale missing data and varying dynamics of variables with different sampling rates pose a challenge to conventional anomaly detection methods. To this end, this paper developed a SAmpling-Rate-aware sequential Knowledge trAnsfer (SARKA) model for detecting the anomalies in multirate industrial processes. First, the raw multirate dataset is chunked into multiple data blocks, such that samples in each block correspond to a specific sampling rate and can be more readily processed by deep neural networks. Then, a Sampling-rate aware Slow Variational Auto-Encoder (S2VAE) model is developed, in which a sampling-rate-aware slowness principle is devised and integrated to enable both data block- and instance-wise personalized dynamic features characterization. Besides, to alleviate the scarce sample problem in the low-sampling-rate data block due to the multirate phenomena, a Sequential Knowledge Transfer (SKT) strategy is devised to convey the knowledge from the high-rate data to facilitate low-rate data modeling and improve the overall monitoring performance. Experimental results from a real-world coal mill in a thermal power plant multirate process demonstrate the effectiveness of the proposed method.Note to Practitioners—Due to the challenges resulting from large-scale missing data and complex varying dynamics, multirate process anomaly detection is a crucial task in modern industries. To tackle the challenge, this paper presents SARKA, which consists of two modules, i.e., S2VAE and SKT. Among them, S2VAE serves as the base model, which integrates a novel devised slowness principle under the probabilistic framework for improved sampling-rate-aware dynamic features characterization. Besides, SKT is elaborated to bridge the multiple S2VAE base models by conveying the modeling knowledge from those high-rate data to low-rate data, as the amount of the low-rate data is generally scarce to train a valid base model. Therefore, the base model can sufficiently capture both the static and dynamic features of the multirate process, and the SKT further enhances the modeling of the low-rate data, yielding improved performance on the overall multirate process.}, archive = {J_TASE}, author = {Zheng Chai and Jiaye Wang and Chunhui Zhao}, doi = {10.1109/TASE.2024.3488272}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8725--8737}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Walk before you can run: Sampling-rate-aware sequential knowledge transfer for multirate process anomaly detection}, volume = {22}, year = {2025}, }
• (2025). Robotic assembly strategy with wrist force sense for narrow clearance peg-in-hole. TASE, 22, 8709--8724. (www)

  @article{ , abstract = {Narrow clearance peg-in-hole assembly is a typical industrial assembly scene. Due to the motion accuracy limitation of industrial robot system and the uncertainty of the assembly circumstance, robotic autonomous assembly of narrow clearance peg in hole assisted by machine vision and force feedback has always been a research focus in this field. However, few studies focused on accomplishing it in a short time while maintaining minimal forces. On the other hand, in some extreme scenarios, such as nuclear, camera lenses may age quickly when exposed to high-dose radiation, requiring a non-visual peg-in-hole method. In this article, aiming at peg-in-hole assembly tasks without vision, and inspired by wrist force sense of human, a robotic narrow clearance peg-in-hole location and insertion methods with wrist force sensation are proposed. The collaborative robot obtains the accurate location of the hole by touching the hole edge several times, named “force-sense-based location”. And then, the peg is automatically inserted into the hole by the collaborative robot with four basic motions named “approach”, “slippage”, “rolling” and “drilling”. A geometric theoretical model was deduced to guarantee the feasibility of the developed method. To avoid excessive contact force, all four motions were designed with admittance hybrid/position control in different directions. The experimental results demonstrate the effectiveness of the proposed method, while the key factors affecting their performance were analyzed. Finally, we compared the methods with three peg-in-hole methods and comprehensively analyzed their performances in terms of assembly time, assembly force, and sensitivity to initial bias. This article provides a new effective method for automatic assembly of narrow clearance peg-in-hole tasks.Note to Practitioners—The efficiency of product assembly has significantly increased thanks to the emergence of robotics. Robots are progressively taking the role of humans in many manual assembly procedures. Peg-in-hole assembly is a common assembly activity accounting for about 40% of the total assembly work. The majority of peg-in-hole assembly techniques start by locating the holes visually, then make modifications based on high-frequency force input. However, vision is frequently prohibited in many real-world assembly scenarios, such as those that actually happen in the dark, low light, or radiation. This paper proposed the use of a UR5e collaborative robot mounted with a six-dimensional force/torque (FT) sensor to complete the peg-in-hole assembly task. The collaborative robots are excellent for applications requiring frequent interactions such as peg-in-hole assembly for high security and great price-performance ratio. A robotic peg-in-hole locating algorithm based on touch sense is presented. This method does not require the assistance of visual equipment like cameras or laser-type displacement transducer. Instead, the location and contact force of the component in various contacts are used to estimate the part’s positional information. The presented method can locate different rigid cylindrical parts and performs fairly precise positioning. The algorithm’s tolerance for position error uncertainty allows it to handle chamfered peg-in-hole assemblies with a narrow clearance ( $25\sim 90\mu $ m). The mathematical derivation ensures the reliability and adaptability of the assembly approach. But a limitation of this approach is that it can only presently be used for cylindrical peg-in-hole assembly with flat end faces. In the future, to expand the application of our method, we will carry out theoretical derivation, control strategy design for more parts with typical shapes, as well as the placement and insertion task of irregular parts.}, archive = {J_TASE}, author = {Yecheng Mei and Ruiya Li and Yuegang Tan and Ding Zhu and Tianliang Li and Zude Zhou}, doi = {10.1109/TASE.2024.3487988}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8709--8724}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic assembly strategy with wrist force sense for narrow clearance peg-in-hole}, volume = {22}, year = {2025}, }
• (2025). Adaptive tuning of dynamic matrix control for uncertain industrial systems with deep reinforcement learning. TASE, 22, 8695--8708. (www)

  @article{ , abstract = {Dynamic matrix control (DMC) has been field-validated in many industrial practices, highlighting the critical importance of fine-tuning parameters for optimal performance. However, the tuning of well-performed parameters is challenging because the relationship between parameters and the performance of DMC is intricate to characterize for industrial systems with uncertainty. An adaptive tuning approach based on deep reinforcement learning (DRL) is proposed to optimize the performance of DMC for uncertain systems in this paper. The approach can online tune the horizons and weighting matrices of DMC in real time adaptive to the state and uncertainty of the systems. Compared with offline tuning approaches, the proposed approach does not need to tradeoff optimality for robustness. The proposed approach utilizes various state-of-the-art DRL algorithms, e.g., value-based and actor-critic-based, to develop online parameter tuning policies that can adapt to system uncertainty. A piecewise reward function is designed to improve the performance and stability of the agent. A novel predictor-switching criterion is developed to address the horizon inconsistency in the receding optimization process. The proposed approaches are validated by the moisture control task in industrial cigarette drying process. Note to Practitioners—This paper is motivated by the adaptive tuning problem of dynamic matrix control (DMC) in uncertain industrial systems. For other nonlinear industrial scenarios, practitioners should first design a nonlinear model predictive controller suitable for the controlled object. Then, they can refer to the proposed tuning algorithm to improve the controller performance. Specifically, regarding the setting of state and action sets, please refer to the technical details provided in this paper. The reward function can be flexibly set according to the needs of practitioners for the controller, e.g., improving the dynamic performance of the controller or saving controller energy consumption. The design processes of the tuning algorithms can refer to Algorithm 1 and Algorithm 2. There are two reasons why the proposed algorithm can be easily transferred to nonlinear industrial scenarios. First, the proposed algorithm does not restrict the model predictive controller or controlled object type. Second, the requirement for horizons or weighting matrices tuning widely exists in the practical applications of various model predictive algorithms.}, archive = {J_TASE}, author = {Yang Zhang and Peng Wang and Liying Yu and Ning Li}, doi = {10.1109/TASE.2024.3487878}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8695--8708}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tuning of dynamic matrix control for uncertain industrial systems with deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Adaptive inverse optimal control for linear human-in-the-loop systems with completely unknown dynamics. TASE, 22, 8683--8694. (www)

  @article{ , abstract = {To improve machines’ intelligence, it is necessary for the machines to learn human’s behavior. In this paper, we make a reasonable hypothesis that a human behaves like a linear quadratic regulator whose cost function is unknown to the machine when performing a task. In addition, the system dynamics in many real applications is completely unknown. Therefore, our purpose is to search for an equivalent cost function to the human only from control input and system state data for continuous-time linear human-in-the-loop (HiTL) systems with completely unknown dynamics. An adaptive inverse optimal control (IOC) method is proposed for this purpose, which can help the machine conduct a better understanding for the human behavior and makes it possible to reproduce a similar optimal controller in other environments. Noticing the difficulty of directly obtaining the weighting matrix, an adaptive integral concurrent learning (ICL) algorithm is developed to identify the system matrices and human feedback gain matrix online, which removes the persistent excitation (PE) conditions. Then, the weighting matrix is determined via solving a convex programming problem. Finally, simulation results on the lane-keeping assist system of an intelligent vehicle are presented to demonstrate the validity of the proposed adaptive IOC algorithm. Note to Practitioners—In practice, it is hoped that the machine can work like a human such that it can replace the human to complete certain tasks. However, it is not easy to design corresponding algorithms for the machine because many tests need to be carried out for selecting appropriate parameters. Instead, an effective method is to teach the machine learn the human’s demonstrated behavior. It is noteworthy that the environment (system dynamics) may be not prior knowledge and only system state and control input are measurable. To this end, an adaptive IOC method is developed for imitation learning the human’s behavior, which is implemented online but requires only limited data. The proposed approach can be used in autonomous driving vehicle, service robot, and medical rehabilitation, etc. In future research, we will extent the proposed method to more complex environment.}, archive = {J_TASE}, author = {Mi Wang and Huai-Ning Wu}, doi = {10.1109/TASE.2024.3487857}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8683--8694}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive inverse optimal control for linear human-in-the-loop systems with completely unknown dynamics}, volume = {22}, year = {2025}, }
• (2025). Intermediate observer-based fault-tolerant control for continuous-time switched affine systems: Application to power converters. TASE, 22, 8670--8682. (www)

  @article{ , abstract = {In this paper, the fault estimation and fault-tolerant control problems are addressed for a class of continuous-time switched affine systems with actuator faults and bounded disturbances. Two novel observer-based approaches are developed to address the fault estimation problem for switched affine systems. The first one refers to a dynamic proportional-integral observer design method, where the presented fault estimate constitute proportional and integral terms to enhance the accuracy of the fault estimation, the common assumption that the measurement output derivative needs to be measurable is eliminated. The second one is an intermediate variable observer, which relaxes the observer matching condition. The output estimation error feedback term is added to the intermediate variable observer to improve the estimation performance. Then, by introducing a switching multi-shifted-point-dependent Lyapunov functional, both a fault-tolerant controller and a new robust output-dependent switching law are jointly designed to compensate the fault effects in the closed-loop switched affine systems and to ensure the practical exponential stability of augmented system, where the convergence region consists of multiple regions and the center point is around some shifted points. The traditional switching quadratic Lyapunov function method is generalized by the designed method. A practical study of a DC-DC boost converter and a numerical example are provided to illustrate effectiveness and validity of the developed fault-tolerant control design method. Note to Practitioners—Power electronics are very common in practical systems, which are usually modeled as a class of switched affine systems. In real applications, faults inevitably occur, which may lead to undesirable behavior and damage to the system. Therefore, how to achieve better fault-tolerant control objectives to guarantee the normal operation of the system with faults is a hot topic. It is practically important to address the fault-tolerant control problem for switched affine systems, where actuator faults and bounded disturbances exist simultaneously. In addition, on account of the existence of affine terms, the controller synthesis of switched affine systems is more complicated than switched linear systems. Based on the special structure of switched affine systems, two kinds of novel fault observers are proposed, where the dynamic proportional-integral observer is proposed to improve the estimation accuracy and speed by utilizing the current output information, and the common supposition that the output derivative needs to be measurable is eliminated. Furthermore, to avoid the limitation of observer matching conditions, an improved intermediate variable observer is designed to estimate faults. Different from the conventional method, the intermediate variable parameters can be selected separately for the corresponding fault channel of each subsystem and the error feedback term of output estimation is added to the intermediate variable observer to enhance the estimation performance. In addition, it is challenging to design a fault-tolerant controller and an output-dependent switching law to guarantee that the augmented system is practically stable and robust to bounded disturbances. The results demonstrate that the designed fault-tolerant control scheme has a definitive practical value.}, archive = {J_TASE}, author = {Fang Liao and Yanzheng Zhu and Michael V. Basin and Donghua Zhou}, doi = {10.1109/TASE.2024.3487635}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8670--8682}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intermediate observer-based fault-tolerant control for continuous-time switched affine systems: Application to power converters}, volume = {22}, year = {2025}, }
• (2025). Economical and reliable energy management for networked microgrids in a multi-agent collaborative manner. TASE, 22, 8659--8669. (www)

  @article{ , abstract = {Reliability is a fundamental requirement of power systems. However, uncertainties from renewable energy generators and demand loads bring challenges to the economical and reliable operation of power distribution networks. This paper focuses on an energy management problem for networked microgrid systems (NMSs), aiming at establishing energy management policies for individual microgrids within NMSs to enhance the long-term economy and stability of the entire system. The key challenge is that each microgrid has independent decision-making capability and how to collaborate all microgrids’ decisions in a distributed multi-agent manner such that the entire system goal is optimized. We introduce the mean and variance of the exchanged power between the NMS and the main grid as the indicators of NMS operation profits and power fluctuations, respectively. The energy management problem is formulated as a mean-variance team stochastic game (MV-TSG) model. Because the variance metric is neither additive nor Markovian in a dynamic setting, the dynamic programming principle does not hold for this problem. We analyze the MV-TSG from a view of sensitivity-based optimization and propose a multi-agent policy iteration method by introducing a sequential update scheme. Furthermore, to tackle MV-TSGs with larger scales or unknown environmental parameters, we extend the idea of trust region optimization and develop a novel multi-agent deep reinforcement learning algorithm. The performance of our approaches is verified through numerical experiments using a real dataset in power distribution networks. Note to Practitioners—Energy management is essential for economical and reliable operation of networked microgrid systems (NMSs). NMSs exchange power with the main grid, where positive exchanged power indicates that the NMS is selling power to the main grid and generating profits, while negative exchanged power indicates that the NMS is purchasing power and incurring costs. The long-term average exchanged power between the NMS and the main grid reflects the economical operation of NMSs. However, fluctuations in exchanged power can lead to power quality issues such as frequency deviation and voltage flicker, which affect the reliability of NMS operations. In this paper, we study the economical and reliable operation of NMSs with distributed energy management policies, formulating the problem as a mean-variance team stochastic game (MV-TSG). We propose a multi-agent policy iteration method to address MV-TSGs. We further develop a multi-agent deep reinforcement learning algorithm to tackle MV-TSGs with large scales and in scenarios where environmental models are unknown. Experiments utilizing real data demonstrate that our algorithms effectively promote collaboration among microgrids to control exchanged power. As the weight coefficient increases, power fluctuations are reduced, achieving power smoothing and peak shaving simultaneously in long-term scheduling. Consequently, our algorithms can facilitate the trade-off between economy and reliability in NMS operations. These results provide valuable insights into the energy management challenges of NMSs in power distribution networks.}, archive = {J_TASE}, author = {Junkai Hu and Li Xia and Jianqiang Hu and Haoran Wu}, doi = {10.1109/TASE.2024.3487292}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8659--8669}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Economical and reliable energy management for networked microgrids in a multi-agent collaborative manner}, volume = {22}, year = {2025}, }
• (2025). Dynamic subclass-balancing contrastive learning for long-tail pedestrian trajectory prediction with progressive refinement. TASE, 22, 8645--8658. (www)

  @article{ , abstract = {Pedestrian trajectory prediction is critical for understanding human behavior. The prevailing approaches employ neural networks to predict trajectories from large amounts of trajectory data. However, pedestrian trajectory data exhibits a long-tail distribution, which presents challenges in accurately predicting the future trajectories of tail samples. Previous research utilized contrastive learning and loss reweighting to tackle the long-tail distribution challenge in trajectory prediction. Although this approach enhanced the tail samples’ performance, it reduced the head samples’ performance. In order to address this limitation, we propose a trajectory prediction framework based on dynamic subclass-balancing contrastive learning in this work. Firstly, we obtain general motion patterns by clustering future trajectory data. We use the adaptive motion pattern refinement block to refine the general motion patterns, providing accurate guidance for the model and thus facilitating the recognition of tail motion patterns. Subsequently, we propose dynamic subclass-balancing contrastive learning to address the long-tail distribution issue of trajectory data on the encoder, which includes subclass-balancing clustering and dynamic dual-level contrastive learning. Subclass-balancing clustering is employed on the head trajectory data to achieve subclass balance across the dataset. Afterward, we perform dynamic dual-level contrastive learning for motion features to achieve instance balance and optimize the feature space. Finally, we use enhanced motion features to adjust the predicted trajectories through the trajectory proposal refinement block, achieving progressive refinement. This addresses the long-tail distribution issue of trajectory data on the decoder and improves the model’s generalization capability. Experimental results demonstrate that our method outperforms state-of-the-art long-tail trajectory prediction methods in addressing the long-tail distribution issue, improving the performance on both head and tail samples. The code will be released at https://github.com/YanCCZU/DSBCL-PRM.Note to Practitioners—This work aims to tackle the long-tail distribution issue of pedestrian trajectory prediction while improving the model’s generalization capability. Existing methods mitigate the impact of the long-tail distribution issue on the encoder using contrastive learning. However, their overemphasis on the tail samples through loss reweighting has reduced the head samples’ performance. This work proposes a dynamic subclass-balancing contrastive learning module, which classifies head samples into several subclasses, each with a similar sample number in the tail classes. It performs dynamic dual-level contrastive learning based on class and subclass labels to achieve subclass and instance balance, improving the performance in head and tail samples. We utilize general motion patterns from the training set to guide the prediction of future trajectories. Moreover, we propose a progressive refinement strategy consisting of two refinement blocks to mitigate the impact of the long-tail distribution issue on the decoder and improve the model’s generalization performance. First, we adaptively refine motion patterns based on the difference between observed trajectories and historical motion patterns to provide accurate guidance. Then, we adjust the original predicted trajectories using the enhanced motion features, mitigating the impact of the long-tail distribution issue on the decoder while improving the model’s generalization and adaptability in unknown scenes. Our method’s simplified and effective model design ensures excellent real-time performance. Consequently, it is well-suited for deployment of edge devices in areas such as autonomous driving, intelligent surveillance, and social robotics. The proposed method enables accurate prediction of infrequent future trajectories in various scenarios, thus supporting safer decision-making.}, archive = {J_TASE}, author = {Biao Yang and Kai Yan and Chuan Hu and Hongyu Hu and Zhitao Yu and Rongrong Ni}, doi = {10.1109/TASE.2024.3487255}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8645--8658}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic subclass-balancing contrastive learning for long-tail pedestrian trajectory prediction with progressive refinement}, volume = {22}, year = {2025}, }
• (2025). Hierarchical reinforcement learning for swarm confrontation with high uncertainty. TASE, 22, 8630--8644. (www)

  @article{ , abstract = {In swarm robotics, confrontation including the pursuit-evasion game is a key scenario. High uncertainty caused by unknown opponents’ strategies, dynamic obstacles, and insufficient training complicates the action space into a hybrid decision process. Although the deep reinforcement learning method is significant for swarm confrontation since it can handle various sizes, as an end-to–end implementation, it cannot deal with the hybrid process. Here, we propose a novel hierarchical reinforcement learning approach consisting of a target allocation layer, a path planning layer, and the underlying dynamic interaction mechanism between the two layers, which indicates the quantified uncertainty. It decouples the hybrid process into discrete allocation and continuous planning layers, with a probabilistic ensemble model to quantify the uncertainty and regulate the interaction frequency adaptively. Furthermore, to overcome the unstable training process introduced by the two layers, we design an integration training method including pre-training and cross-training, which enhances the training efficiency and stability. Experiment results in both comparison, ablation, and real-robot studies validate the effectiveness and generalization performance of our proposed approach. In our defined experiments with twenty to forty agents, the win rate of the proposed method reaches around ninety percent, outperforming other traditional methods. Note to Practitioners—With artificial intelligence rapidly developing, robots will play a significant role in the future. Especially, the swarm formed by many robots holds promising potential in civil and military applications. Promoting the swarm into games or battles is rather riveting. The reinforcement learning method provides a plausible solution to realize the battle of robotic swarms. There are still some issues that need to be addressed. On one hand, we focus on the uncertainty caused by the battlefield nature and the environment which limits our ability for the implementation of swarms. On the other hand, we solve the problem that the decision process combined with commands and actions is a hybrid system, which cannot be directly reflected in the confrontation of swarms. Overall, our approaches throw light on artificial general intelligence and also reveal a solution to interpretable intelligence.}, archive = {J_TASE}, author = {Qizhen Wu and Kexin Liu and Lei Chen and Jinhu Lü}, doi = {10.1109/TASE.2024.3487219}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8630--8644}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical reinforcement learning for swarm confrontation with high uncertainty}, volume = {22}, year = {2025}, }
• (2025). Distributed economic dispatch algorithm with quantized communication mechanism. TASE, 22, 8618--8629. (www)

  @article{ , abstract = {Due to the limited bandwidth and energy of communication channels among agents in practical applications, the communication-efficient distributed optimization method has emerged as a pressing research topic in recent years. The distributed economic dispatch problem with restricted data communication/finite communication bandwidth is investigated in this study, where the communication among agents can be described as a strongly connected directed network. For this purpose, a robust push-pull distributed optimization algorithm with a dynamic scaling quantization mechanism is developed based on the gradient tracking technique. A novel surplus variable is designed to prevent the accumulation of quantization errors, and then, a heavy-ball momentum is introduced to speed up convergence performance. In addition, a linear convergence rate of the developed approach is deduced for the strongly convex and Lipschitz smooth cost function. Finally, we offer two instances for illustration. Note to Practitioners—This paper proposes a robust quantization-based algorithm for the economic dispatch problem, in which the broadcasting information is quantized before sending to its neighboring generators. Therefore, this method reduces duplicate transmission of agents and improves the use of communication resources. Furthermore, the developed method can be extended to similar constrained optimization problems, such as the resource allocation problem in wireless networks, and the network utility maximization problem in the Internet.}, archive = {J_TASE}, author = {Xiasheng Shi and Changyin Sun and Chaoxu Mu}, doi = {10.1109/TASE.2024.3487214}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8618--8629}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed economic dispatch algorithm with quantized communication mechanism}, volume = {22}, year = {2025}, }
• (2025). Active disturbance rejection based adaptive dynamic surface control for nonlinear systems. TASE, 22, 8607--8617. (www)

  @article{ , abstract = {This paper presents an adaptive dynamic surface control (DSC) method for nonlinear systems with multiple disturbances, parameter uncertainties, and unknown nonlinear dynamics, using a reduced-order extended state observer (ROESO). Compared to related methods, this proposed approach offers several distinct advantages: (i) The method does not require knowledge of the upper bound function for unknown nonlinear dynamics, allowing the controlled plant to not be bounded-input bounded-state, thus expanding the applicability of the DSC method; (ii) By utilizing known model information, an ROESO is designed to handle mismatched uncertainties and disturbances, reducing the load on the observer and enhancing its capability to suppress unknown nonlinear dynamics; (iii) This method employs an adaptive output-feedback DSC approach, which is more practical and easier to implement than state-feedback methods; and (iv) The observer gain, adaptive gain, and DSC gain are simultaneously optimized using a particle swarm optimization algorithm. Additionally, detailed stability analysis is provided, and simulations and comparative experiments are conducted on a rotational system to demonstrate the efficacy and superiority of the proposed method. Note to Practitioners—Most real systems are nonlinear and subject to a variety of matched and unmatched uncertainties and disturbances. Adaptive control is an effective method for dealing with parametric uncertainty, but requires that the uncertainty can be represented linearly in terms of unknown parameters and cannot handle exogenous disturbances. As an alternative active disturbance attenuation method, disturbance/uncertainty estimation and attenuation techniques have a two-degree-of-freedom control structure. However, for nonlinear systems, they simply estimate and compensate for the nonlinear characteristics of the system as disturbances, which, although effective, can lead to problems such as excessive inputs in the actual control, thus affecting the performance of the system. Backstepping control is one of the most powerful tools for handling nonlinear systems to deal with mismatched disturbances and has been well used in various fields, but it usually suffers from the problem of “explosion of complexity”. Thus, how to deal with various matched and unmatched uncertainties and disturbances in a nonlinear system is still a challenge. Motivated by these considerations, this paper presents an adaptive output-feedback dynamic surface control method for a class of uncertain nonlinear systems with multiple mismatched uncertainties and disturbances.}, archive = {J_TASE}, author = {Lan Zhou and Yongbo Sun and Yong He and Hongyi Li and Xian-Ming Zhang}, doi = {10.1109/TASE.2024.3487131}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8607--8617}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active disturbance rejection based adaptive dynamic surface control for nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Graph-based imitation learning for real-time job shop dispatcher. TASE, 22, 8593--8606. (www)

  @article{ , abstract = {We propose an advanced real-time dispatcher for minimizing the makespan of job shop scheduling problems (JSSPs), which are NP-hard combinatorial problems. The proposed dispatcher can be applied to large-sized unseen problems without additional learning. Several studies recently proposed a scalable dispatching agent using a graph neural network (GNN) and reinforcement learning (RL). However, we observe that they have not considered suitable Markov decision process (MDP) and GNN structure to solve JSSPs. Therefore, we incorporate scheduling theory and properties to define the state, action, and GNN model. We especially define the action set and state transition so that active schedules can be exclusively generated, define node features in a dynamic manner, and use a neighbor type-aware Graph Attention Network (GAT) model with length-agnostic neighbor sets. We also investigate the use of imitation learning (IL) to learn the dispatcher instead of the RL. We evaluate the effectiveness of our dispatcher on benchmark instances and dynamic environments. Note to Practitioners—This work aims to develop an advanced real-time dispatcher using GNN. It can handle dynamic scheduling environments. The focus is on JSSPs without constraints, commonly seen in real manufacturing systems. To improve the GNN-based dispatchers proposed by recent studies, we refine the dispatcher using a novel state, action, and a GNN structure. Our dispatcher demonstrates state-of-the-art performance compared to other real-time dispatchers on JSSP benchmark instances and several customized instances, which consist of up to 20 machines and 300 jobs. Additionally, we assess the dispatcher’s performance in dynamic JSSP environments, including dynamic job arrival, machine breakdown, and stochastic processing time. Note that, to apply the proposed dispatcher for real-world fields, you have to prepare only the information about predefined machine orders and currently remaining processing times for each job. Also, you do not need a simulator if you are going to utilize a trained dispatcher. In the future, extended study is needed to apply it to sequence-dependent setup constraints, due-related objectives, etc.}, archive = {J_TASE}, author = {Je-Hun Lee and Hyun-Jung Kim}, doi = {10.1109/TASE.2024.3486919}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8593--8606}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Graph-based imitation learning for real-time job shop dispatcher}, volume = {22}, year = {2025}, }
• (2025). Rethinking discrepancy analysis: Anomaly detection via meta-learning powered dual-source representation differentiation. TASE, 22, 8579--8592. (www)

  @article{ , abstract = {Industrial environments pose distinctive challenges for anomaly detection, primarily stemming from the complexities associated with high dimensionality and the dynamic nature of data patterns over time. These properties determine that the model’s proper convergence on unlabeled data is unpromising, consequently leading to less efficient discrimination of anomalies in previous anomaly detection (AD) works. To address this problem, we present AnoDual, a novel, meta-learning AD framework. From the perspective of data reconstruction, we introduce the multi-memory enhanced VAE reconstructor M2ER, which learns to extract the most salient patterns in unlabeled noisy data through a self-supervised manner. This design eases impacts from potential anomalous components during data reconstruction, and enhances the discernibility of anomalies. To address performance degradation caused by the numerical deviation based AD scheme in most existing works, we design a dual-source self-supervised discriminator DSD, which examines characteristics in the domain of representations. This model actively assesses discrepancies between data pairs and representation pairs in parallel, and conducts AD on a fine-grained scale. In this way, anomalies that used to be unnoticed due to a less prominent numerical deviation can be spotted. Besides, we propose a meta-learning powered training pipeline to enable model training even when no real label is available, which is common in the industry. Extensive experiments on five large-scale real-world industrial datasets suggest that AnoDual achieves an average F1-Score with a substantial increment of 3.39 %, outperforming the latest state-of-the-art baseline. Note to Practitioners—A generative model plus a numerical threshold based detection approach currently takes a significant share in both academia and the industry. However, the performance of this workflow is not promising in actual applications, with multiple factors contributing to this situation. The proper convergence of such generative models is difficult when the training material contains noisy samples - an over-expressed generative model would result in less significant reconstruction discrepancies for anomalies that are hard to notice. In addition, selecting a numerical threshold, which is used to spot anomalies, requires multiple laborious attempts, and can hardly adapt to an ever-changing pattern in industrial environments. These circumstances make it challenging to apply prior works in practical production, which, in turn, urges the need to develop an effective methodology to address the need for industrial anomaly detection. This manuscript includes a novel, meta-learning powered framework AnoDual, which is tailored for industrial scenarios. This framework discards the conventional design of comparing the reconstruction error numerically, but introduces a solution based on the differentiation of the representations. Besides, the multi-head attention enhanced variational autoencoder also leads to a much more pronounced discrepancy for anomalous samples, which benefits their successful detection. Providing a flexible and robust way to detect anomalies on deployed IoT assets, this work can be further transformed to serve applications in many other domains.}, archive = {J_TASE}, author = {Muyan Yao and Dan Tao and Peng Qi and Ruipeng Gao}, doi = {10.1109/TASE.2024.3486688}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8579--8592}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rethinking discrepancy analysis: Anomaly detection via meta-learning powered dual-source representation differentiation}, volume = {22}, year = {2025}, }
• (2025). LeTO: Learning constrained visuomotor policy with differentiable trajectory optimization. TASE, 22, 8567--8578. (www)

  @article{ , abstract = {This paper introduces LeTO, a method for learning constrained visuomotor policy with differentiable trajectory optimization. Our approach integrates a differentiable optimization layer into the neural network. By formulating the optimization layer as a trajectory optimization problem, we enable the model to end-to-end generate actions in a safe and constraint-controlled fashion without extra modules. Our method allows for the introduction of constraint information during the training process, thereby balancing the training objectives of satisfying constraints, smoothing the trajectories, and minimizing errors with demonstrations. This “gray box” method marries optimization-based safety and interpretability with powerful representational abilities of neural networks. We quantitatively evaluate LeTO in simulation and in the real robot. The results demonstrate that LeTO performs well in both simulated and real-world tasks. In addition, it is capable of generating trajectories that are less uncertain, higher quality, and smoother compared to existing imitation learning methods. Therefore, it is shown that LeTO provides a practical example of how to achieve the integration of neural networks with trajectory optimization. We release our code at https://github.com/ZhengtongXu/LeTO. Note to Practitioners—LeTO is driven by the goal of developing an imitation learning algorithm capable of generating safe and constraint-satisfying robotic behaviors. The idea of imitation learning is to enable the robot to learn from human demonstrations of certain tasks. Subsequently, the robot is able to autonomously perform the learned tasks on its own. Thanks to the powerful representational and fitting capabilities of neural networks, imitation learning can let robots perform complex manipulation tasks. However, neural networks often exhibit a certain level of uncertainty and lack theoretical safety guarantees. For robotic systems, it is crucial that robot behaviors meet specific constraints; otherwise, the system may not be sufficiently reliable. Therefore, we introduce LeTO, an approach that integrates trajectory optimization with neural networks to generate actions that not only achieve manipulation tasks, but also comply with constraints. This improves the interpretability, safety, and reliability of robot policies acquired through imitation learning, facilitating their deployment in scenarios with high safety requirements.}, archive = {J_TASE}, author = {Zhengtong Xu and Yu She}, doi = {10.1109/TASE.2024.3486542}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8567--8578}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {LeTO: Learning constrained visuomotor policy with differentiable trajectory optimization}, volume = {22}, year = {2025}, }
• (2025). ParetoSSL: Pareto semi-supervised learning with bias-aware gradient preferences for fruit yield estimation. TASE, 22, 8554--8566. (www)

  @article{ , abstract = {Fruit counting is a fundamental task for fruit yield estimation. Though semi-supervised counting methods have received increased attention in recent years, due to the high data utilization of unlabeled data, they suffer from two limitations. Firstly, difficult weight selection is a limitation, as these methods rely on manually selected fixed weights for both the supervised learning loss and the consistency learning loss, resulting in limited performance. Secondly, biased pseudo-labeling is another limitation, as they may predict biased pseudo-labels that result in small consistency learning losses, leading to training being dominated by supervised learning with large losses. To tackle these two limitations, in this paper, we propose a novel method named ParetoSSL to automatically derive weights of losses from the perspective of multi-task learning. Specifically, ParetoSSL formulates a multi-objective optimization problem for weight derivation by maximizing the similarity between weighted gradients of losses and a customized gradient preference vector, in which, the vector can guide weight derivation. Moreover, to relieve the effect of pseudo-label biases on consistency learning, we propose a bias-aware gradient preference vector. This vector considers gradient biases brought by the pseudo-label biases, which will down-weight the supervised learning loss while high-weighting the consistency learning loss. Meanwhile, to improve the robustness of ParetoSSL, an inequality equation regarding the norm of the gradients of the consistency learning loss is designed to control the range of gradient biases. Extensive experiments are conducted on the Clustered-Fruit dataset and Fruit-2019 dataset to evaluate the effectiveness of ParetoSSL on semi-supervised counting. Experimental results show that our ParetoSSL is superior to state-of-the-art methods. Note to Practitioners—This work is motivated by the emerging need for semi-supervised counting methods in fruit yield estimation. The difficulty of selecting loss weights for training semi-supervised counting algorithms is exacerbated by the pseudo-label bias issue that pseudo-label biases mislead the weight derivation while maximizing the similarity between the weight gradient and gradient preference vectors. The proposed bias-aware gradient preference vectors help users derive loss weights automatically and save the time of choosing loss weights for model fine-tuning. The proposed ParetoSSL is generic as it can be employed as a fruit yield estimation component of crop management support systems, while at the same time being applied to counting frameworks of other objects.}, archive = {J_TASE}, author = {Xiaochun Mai and Meilu Zhu and Yixuan Yuan}, doi = {10.1109/TASE.2024.3486381}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8554--8566}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ParetoSSL: Pareto semi-supervised learning with bias-aware gradient preferences for fruit yield estimation}, volume = {22}, year = {2025}, }
• (2025). Charge or pick up? optimizing E-taxi management: A dual-stage heuristic coordinated reinforcement learning approach. TASE, 22, 8533--8553. (www)

  @article{ , abstract = {In recent years, the rapid adoption of electric vehicles (EVs) in the taxi industry has transformed traditional taxi-hailing systems into electric taxi (E-taxi) hailing systems. As a result, it is crucial to develop effective strategies for optimizing E-taxi management by considering both passenger-taxi matching and charging planning. In this paper, we first formalize the E-taxi management optimization problem as a Markov decision problem with dynamic state and heterogeneous action. We then propose a dual-stage heuristic coordinated reinforcement learning (RL) approach that incorporates advanced feature selection and heuristic allocation strategies. Our approach consists of two main stages. In the first stage, we introduce the feature-guided state dimensionality stabilization proximal policy optimization (PPO) method to address dynamic state dimensions by a feature selection method, and enabling E-taxis to decide whether to charge or pick up passengers. In the second stage, we propose a heuristic coordinated assignment method to further allocate charging stations and passengers for the E-taxis, and provide the RL network in the first stage with rewards based on the results. This approach effectively tackles the challenge of RL processing of heterogeneous action spaces (charge and pick up). We evaluate our proposed method in a real-world E-taxi environment and find that it significantly enhances the experience for both E-taxis and passengers. Specifically, due to our method’s rational planning for passenger pick-up and charging, E-taxis can increase their revenue by 20% compared to traditional RL methods or random scheduling approaches. As for passengers, since the taxis have more efficiently planned their charging behavior, the probability of their orders being answered increases by 15%, while their waiting time is reduced by 55%. These achievements contribute to the advancement of E-taxi management strategies and promote the widespread adoption of electric vehicles, ultimately supporting the transition to a more sustainable transportation system. Note to Practitioners—The increasing adoption of electric vehicles in the taxi industry has led to the need for effective E-taxi management strategies that consider both passenger-taxi matching and charging planning. In this study, we introduce a dual-stage heuristic coordinated reinforcement learning approach that addresses these challenges by integrating a feature-guided state dimensionality stabilization proximal policy optimization method and a heuristic coordinated assignment method. Our approach offers several practical benefits for E-taxi service providers, drivers, and passengers. For E-taxi service providers, the proposed method improves E-taxi dispatch efficiency, resulting in a more effective use of available resources and potentially increasing overall revenue. For E-taxi drivers, our approach leads to better planning of charging and passenger pick-up decisions, increasing their earnings by 20% compared to traditional methods, and reducing the average occurrence of low battery status from more than 4 times every 10 hours to less than 1 time. Passengers, on the other hand, experience improved service quality due to the more efficient E-taxi management. The probability of their orders being answered increases by 15%, and their waiting time is reduced by 100%. These improvements contribute to an enhanced user experience and may encourage further adoption of E-taxis as a sustainable transportation solution. The proposed method can be integrated into existing E-taxi hailing platforms, such as DiDi and Uber, to enhance their dispatch and charging management capabilities. As the global trend towards sustainable transportation continues to grow, our approach provides valuable insights and a practical solution for the efficient management of E-taxi fleets in modern urban environments.}, archive = {J_TASE}, author = {Donghe Li and Chunlin Hu and Qingyu Yang and Pengtao Song and Feiye Zhang and Dou An}, doi = {10.1109/TASE.2024.3486342}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8533--8553}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Charge or pick up? optimizing E-taxi management: A dual-stage heuristic coordinated reinforcement learning approach}, volume = {22}, year = {2025}, }
• (2025). Stability and h∞ control synthesis of electric ground vehicle lateral dynamics through LPV sampled-data systems. TASE, 22, 8519--8532. (www)

  @article{ , abstract = {For electric ground vehicles, yaw moment gain-scheduling controller cannot adequately characterize vehicle handling information, especially longitudinal acceleration and calculation cost. This will increase effects of the front steering angle on vehicle stability and make the vehicle handling performance worse. Thus in our work, through LPV sampled-data systems, a yaw moment controller considering longitudinal acceleration is proposed to maximize handling performance and meanwhile minimize computational cost of vehicles. First, a novel integral inequality with the cubic term of integral interval length instead of the reciprocal one is constructed by non-orthogonal polynomials. This can fully utilize slack matrix variables and additional information about sawtooth structural sampling pattern of vehicle handling. Then, based on the constructed integral inequality, a cubic-term-dependent discontinuous exponent discontinuous Lyapunov-Krasovskii functional (LKF) including additional vehicle handling information is designed. To better estimate the upper bound of LKF derivative further, multiple convex function approximation (MCFA) approach is investigated. It can be applied to handle the two-variable polynomial negative definite problem by separating the variable interval into multiple sub-intervals. Thus, sufficient conditions with better performance of vehicle handling are derived for the feasibility of an $H_{\infty } $ yaw moment sampled-data controller. In addition, an iterative algorithm is constructed through the inner convex approximation solution technique. And the parameter-dependent bilinear matrix inequalities can be further transformed into linear ones. Hence, it is easy to obtain the desired yaw moment sampled-data controller gain. Finally, the validity and superiority of our developed approaches can be verified, by comparing with previous results and implementing practical experiments. Note to Practitioners—In existing yaw moment controllers, vehicle longitudinal acceleration is not considered. This will destroy vehicle handling stability. Hence a yaw moment controller considering longitudinal acceleration is proposed through LPV systems. Meanwhile, to sufficiently make use of the bandwidth of vehicle, sampled-data controller is designed. As is well known, if the allowable maximum sampling intervals of sampled-data systems are increased, the computational cost of the vehicle can be reduced. To maximize vehicle handling performance and allowable maximum sampling intervals, a novel integral inequality is proposed and an appropriate LKF constructed. Because the constructed LKF derivative is a two-variable polynomial, MCFA is proposed to handle its negative definite problem. The sufficient stability condition via MCFA approach is bilinear matrix inequalities, which are difficult to deal with. Thus, an inner convex approximation algorithm is developed. Then the optimal yaw moment controller of vehicle lateral dynamics can also be obtained without using nonlinear solvers. Hence, a certain degree of freedom can be ensured. In summary, a yaw moment sampled-data controller can maximize vehicle handling performance while minimizing vehicle computational cost by the proposed integral inequality, MCFA approach and an inner convex approximation algorithm in this work.}, archive = {J_TASE}, author = {Wenchengyu Ji and Yulian Jiang and Bin Zhao and Chao Cheng and Shenquan Wang}, doi = {10.1109/TASE.2024.3486281}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8519--8532}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stability and h∞ control synthesis of electric ground vehicle lateral dynamics through LPV sampled-data systems}, volume = {22}, year = {2025}, }
• (2025). EEGCiD: EEG condensation into diffusion model. TASE, 22, 8502--8518. (www)

  @article{ , abstract = {Electroencephalography (EEG)-based applications in Brain-Computer Interfaces (BCIs), neurological disease diagnosis, rehabilitation, and other areas rely on the utilization of extensive data for model development. Nevertheless, this raises concerns regarding storage and privacy, since model development needs a significant amount of data, and EEG sharing discloses sensitive information such as identity and health. To address this challenging problem, we provide the paradigm of EEG condensation, aiming to generate a synthetic sample set that is highly information-concentrated yet not visually similar. Correspondingly, we propose a novel dataset condensation framework where the knowledge of the original EEG dataset is condensed into diffusion models, named EEGCiD. Specifically, EEGCiD first utilizes a deterministic denoising diffusion implicit model (DDIM) to store the information of the original dataset and optimizes the condensation latent codes z to obtain the EEG condensation dataset. Further, to enhance the modeling of EEG knowledge in DDIM, we design a transformer architecture incorporating the spatial and temporal self-attention block (STSA) to replace the traditional U-Net backbone. In the condensation phase, EEGCiD randomly initializes a subset of samples from the original dataset to obtain the condensation latent codes z through the forward process in DDIM. Then, it optimizes z by matching the feature distributions in multiple EEG decoding models between the synthetic samples and the original dataset. Extensive experiments across three EEG datasets demonstrate that the condensation dataset from the proposed model not only achieves superior classification performance with limited sample sizes, but also effectively prevents membership inference attacks (MIA). Note to Practitioners—This paper aims to investigate a novel EEG generation paradigm that extracts representative synthetic samples from large-scale datasets. Existing studies in EEG generation primarily concentrate on generating real-like signals, and some work claims that the generated EEG can serve as a substitute for the original dataset to achieve privacy preservation. In the EEGCiD framework, the deterministic DDIM is pre-trained with the original dataset to store the knowledge. Besides, an ensemble feature matching strategy is proposed to condense the information from the original dataset into a small latent code set. Experiments on three datasets demonstrate that EEGCiD addresses two fundamental challenges: 1) obtaining superior classification performance within a small dataset (limited storage capacity); 2) avoiding potential privacy issues during EEG sharing and transmission.}, archive = {J_TASE}, author = {Junfu Chen and Dechang Pi and Xiaoyi Jiang and Feng Gao and Bi Wang and Yang Chen}, doi = {10.1109/TASE.2024.3486203}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8502--8518}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {EEGCiD: EEG condensation into diffusion model}, volume = {22}, year = {2025}, }
• (2025). Event-triggered T-S fuzzy load frequency control with variable probabilistic release for renewable energy integrated power systems. TASE, 22, 8492--8501. (www)

  @article{ , abstract = {This paper explores a T-S fuzzy-model-based approach for load frequency control in network-based power systems under random false data injection attacks, taking into account the integration of electric vehicles and photovoltaic power generation systems. Amidst uncertainties in parameters and nonlinear items of integrated power systems, a T-S fuzzy model is formulated, enabling convenient design and analysis of the rolling horizon optimal control (RHOC) strategy with variable control gains to ensure mean-square asymptotic stability in power systems. Considering the broad dispersion of new energy generation systems across the power grid, the use of network communication has become crucial. To overcome the limitation of network bandwidth while ensuring frequency stability, a novel event-triggered mechanism employing a varying probabilistic release strategy (VPRS) within grouped triggered data-packets is proposed for transmitting power frequency signals. Initially, a conventional event-triggered mechanism is established to generate primary triggering packets, stored in a buffer until designated packet capacity of the group is reached. Then, following the probabilistic updating algorithm in RHOC strategy, the transmission task is executed upon the arrival of the final triggered packet within each group. The efficacy of the proposed method is validated through a numerical example. Note to Practitioners—This paper introduces a T-S fuzzy-model-based RHOC strategy for power systems with renewable energy, addressing challenges posed by limited communication resources and random false data injection attacks. The innovative event-triggered mechanism with a VPRS optimizes network utilization for power systems, while the RHOC strategy with variable control gains guarantees mean-square asymptotic stability despite uncertainties and nonlinearities within the power network. While the efficacy of the approach is demonstrated through numerical examples, practitioners should take into account practical constraints, such as scalability issues and the complexities involved in real-world implementation. Future research will focus on enhancing scalability and addressing implementation challenges to facilitate broader adoption in power systems.}, archive = {J_TASE}, author = {Zhou Gu and Yujian Fan and Fan Yang and Engang Tian}, doi = {10.1109/TASE.2024.3486093}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8492--8501}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered T-S fuzzy load frequency control with variable probabilistic release for renewable energy integrated power systems}, volume = {22}, year = {2025}, }
• (2025). Quantized output feedback tracking control for discrete-time periodic markov jump systems with packet loss compensation. TASE, 22, 8480--8491. (www)

  @article{ , abstract = {The $H_{\infty }$ static output feedback tracking control issue for discrete-time periodic Markov jump systems with quantization and packet loss is explored. The packet loss follows Bernoulli random distribution, which assumes that the packet is discarded in a probabilistic manner before being transmitted to the controller. On this basis, considering the restricted network bandwidth, a novel quantization-based packet loss compensation scheme using single exponential smoothing approach is firstly given to help offset the influence of network congestion and missing packets. Then, an output feedback tracking controller is firstly designed to minimize the tracking error between the system output and the given reference model output. Aiming at the loss of mode information, the tracking controller designed is partially mode-dependent. Furthermore, by giving a mode-dependent Lyapunov function with periodicity, the sufficient condition for the existence of this controller is derived to ensure the stability of the tracking error system with $H_{\infty }$ performance. Ultimately, the effectiveness and practicality of the developed technique are demonstrated through an example of a single-link robotic arm model. Note to Practitioners—In real life, periodic systems generated by random mutations can be seen everywhere, such as economic systems. This type of system undergoes structural or parameter changes within a single operating period due to sudden changes in the external environment. Periodic Markov jump systems (PMJSs) can effectively describe complex systems with both periodic and stochastic characteristics. To address the adverse effects of quantization and packet loss, a new packet loss compensation strategy based on single exponential smoothing method and quantization is proposed. On the other hand, research on output feedback tracking control is crucial for fields such as missiles and spacecraft. To ensure that the system can operate according to the specified trajectory, a new design method for a periodic output feedback tracking controller has been proposed. And this type of controller effectively solves the tracking problem of PMJSs with quantization and packet loss.}, archive = {J_TASE}, author = {Fan Zhang and Mingang Hua and Feiqi Deng and Juntao Fei and Hua Chen}, doi = {10.1109/TASE.2024.3486067}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8480--8491}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quantized output feedback tracking control for discrete-time periodic markov jump systems with packet loss compensation}, volume = {22}, year = {2025}, }
• (2025). Robotic motion planning based on deep reinforcement learning and artificial neural networks. TASE, 22, 8465--8479. (www)

  @article{ , abstract = {Although robotic trajectory generation problem has been extensively investigated, existing solutions are almost customized to specific robot geometry, and generalized schemes are yet to be explored. In this article, a general motion planning framework based on deep reinforcement learning (DRL) and artificial neural networks (ANNs) is proposed for robot with arbitrary geometry. First, a unique screening and grafting mechanism is established to improve the policy learning by exploiting valuable experience sufficiently. Second, based on the reward-oriented characteristics of DRL, a forward progression mechanism is proposed to facilitate the path planning for complicated tasks. Third, a structure consisting of an adventurer and conservator algorithm with automatic optimization and an ANN-based mapper is designed integrally to derive the inverse kinematics solutions without considering the robot geometry. Finally, experimental results have verified the superior performance of the proposed approach.Note to Practitioners—This article is aiming to provide a general method to solve the problem of motion planning for robots via deep reinforcement learning (DRL) and artificial neural networks (ANNs). Compared to the existing approaches, which are highly specialized and limited to robots with specific geometries, and often cumbersome, our method can be easily applied to robots with arbitrary geometries and has good generalization, where to simplify the training based on DRL for diverse practical motion planning tasks, a universal forward progression mechanism is used to partition a complex task into multiple successive simple phases. Furthermore, an ANN-based mapper can cope with the burdensome inverse kinematics of robots with both common and uncommon geometries, especially the ones with redundant degrees of freedom. The proposed method is also validated to be superior to other state-of-the-art methods in real-world experimental studies.}, archive = {J_TASE}, author = {Huashan Liu and Xiangjian Li and Menghua Dong and Yuqing Gu and Bo Shen}, doi = {10.1109/TASE.2024.3486064}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8465--8479}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic motion planning based on deep reinforcement learning and artificial neural networks}, volume = {22}, year = {2025}, }
• (2025). A learning system for deformable object cooperative manipulation. TASE, 22, 8453--8464. (www)

  @article{ , abstract = {Dynamic motion primitives (DMPs) have been widely used in robotics and automation systems because of their rapid deployment capability. Previous research has concentrated on extending coupled dynamic movement primitives (CDMPs) to manipulate rigid objects using a dual-arm robot. However, manipulating deformable objects may fail due to issues with the workspace, manipulability caused by the robot’s layout, and the uncertain dynamics of deformable objects. This research proposes a unique system that combines Learning from Demonstration (LfD) to optimise robot layouts based on human experience and robot manipulability. It also employs a modified CDMPs (MCDMPs) method to manipulate deformable objects. The MCDMPs include a new term to ensure that the deformed object can change its configuration during manipulation. Furthermore, another term is introduced to track the desired trajectory of the deformed object, which is crucial for transporting tasks. We conducted simulations using a mass-spring-damper system for cooperative manipulation to validate the proposed approach. We also employed a dual-arm robot platform to transport a deformed ball with disturbance. The simulation and experimental findings indicate that our method performs well in trajectory tracking and configuration change. Note to Practitioners—This research paper addresses the need for faster deployment of dual-arm robots in various applications, such as industrial and services. We must consider the robot’s installation layout and task programming for rapid deployment. To optimise layouts for tasks quickly, we introduce an optimisation framework which considers the comprehensive performance of dual arms and uses demonstration sampling and learning generalisation. To address the manipulation of deformable objects by dual-arm robots, we use an MCDMPs method that employs the barrier Lyapunov function (BLF) to track changed configurations. Additionally, we introduce a term for reference trajectory tracking component for the manipulated object.}, archive = {J_TASE}, author = {Donghao Shi and Hao Hu and Chenguang Yang and Zhenyu Lu and Qinchuan Li}, doi = {10.1109/TASE.2024.3486063}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8453--8464}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A learning system for deformable object cooperative manipulation}, volume = {22}, year = {2025}, }
• (2025). Motion planners for path or waypoint following and end-effector sway damping with dynamic programming. TASE, 22, 8439--8452. (www)

  @article{ , abstract = {We propose two novel motion planners for a robotic manipulator with a passive end-effector that is free to sway during and after the robot’s motion. The planners utilize Dynamic Programming to generate trajectories that damp the end-effector’s residual sway while ensuring that the boom tip—the point to which the end-effector is attached—follows a collision-free path or time-dependent waypoints. Our use case is a crane of a forwarder machine, a log-loading machine in the forestry industry, with a passive grapple. In the cluttered forest environment, accurate path following and grapple sway damping are critical to increase the operation efficiency and avoid harming the machine and environment. The results of the simulation in a high-fidelity multibody-dynamics simulator showcase the effectiveness of our methodology in achieving exact path following or timed waypoints following and the residual sway damping. In particular, the average of the maximum residual sway is only 1.9°, showing an average reduction of 75%, as compared to fifth, sixth, and tenth order polynomial trajectories, in six test cases, including common paths used by operators to pick and place logs. Monte-Carlo simulations also showed that our planners have very good robustness against payload mass uncertainty. Other merits of our Dynamic Programming trajectories are that they are smooth, computationally inexpensive, and result in reduced residual sway even for nonzero initial sway conditions. Moreover, the generality of our methodology opens a new way to design anti-sway motion planners for construction cranes or quadrotors with a slung payload, in addition to serial manipulators with passive end-effectors.Note to Practitioners—This work was motivated by the problems arising in the operation of log-loading cranes in the forestry industry: the problems of the end-effector’s large sway during crane reconfiguration and the collision between the crane and obstacles, which are detrimental to the efficiency of the operation. Similar issues arise, for example, in construction cranes transporting large hanging objects. We propose a novel methodology to address both problems by generating smooth and computationally inexpensive trajectories for the crane joint motion. The approach begins with the model of the sway motion and the definition of the collision-free path. Then, our Dynamic Programming algorithm generates anti-sway trajectories that satisfy the joint constraints. The results in a high-fidelity simulator show that our motion planners lead to precise path following and significant sway damping, and also confirm its superiority compared to polynomial trajectories, commonly used in industries. Monte-Carlo simulation also confirms our planners’ robustness. Our methodology is also applicable to other dynamic systems with freely hanging objects, such as multi-degree-of-freedom robotic manipulators, construction cranes, and quadrotors carrying a slung payload. A possible limitation is that the methodology is model-based and necessitates finding the sway dynamics model and estimating payload properties.}, archive = {J_TASE}, author = {Iman Jebellat and Inna Sharf}, doi = {10.1109/TASE.2024.3486040}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8439--8452}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Motion planners for path or waypoint following and end-effector sway damping with dynamic programming}, volume = {22}, year = {2025}, }
• (2025). Distributed economic dispatch of microgrids based on ADMM algorithms with encryption-decryption rules. TASE, 22, 8427--8438. (www)

  @article{ , abstract = {Distributed economic dispatch (ED) has emerged as a critical issue in microgrid operations due mainly to the wide application of various clean energy as well as energy storage units. The openness of communication networks in microgrids can lead to privacy breaches, which pose a serious threat to the entire electricity market. As such, this paper presents a distributed ED algorithm based on the alternating direction method of multipliers (ADMM), where a quantization-based encryption and decryption rule is integrated to avoid privacy leakage while iteratively acquiring the optimal ED scheme. By resorting to the property of monotonically convergent sequences, a sufficient condition about the learning rate is profoundly revealed to guarantee the algorithm convergence. Two extended results are presented, respectively, to enhance the convergence rate and meet the requirement of plug-and-play scenarios. Finally, the validity (both privacy and optimality) of the proposed algorithm is verified by using the dual-source trolleybus system in Beijing. Note to Practitioners—This paper develops an engineering-oriented ED algorithm that optimizes the total generation costs of smart grids online while guaranteeing system constraints. Shared network communication undoubtedly plays a significant role in achieving iteratively the optimal solution of distributed algorithms. However, some crucial and sensitive information exchanged via an open and shared network could be eavesdropped by malicious attackers, which could result in a serious security threat affecting the reliability and stability of the smart grid. To overcome such a shortage, an encryption-decryption rule is constructed via a dynamic quantizer. In light of such a rule, the presented algorithm based on ADMM can iteratively acquire the optimal ED solution in a distributed way, realizing the requirements of optimality and privacy. The desired range of the learning rate is disclosed to guide the parameter selection, and two improved versions are proposed to meet more general engineering practice involving plug-and-play scenarios.}, archive = {J_TASE}, author = {Lei Sun and Derui Ding and Hongli Dong and Xiaojian Yi}, doi = {10.1109/TASE.2024.3485922}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8427--8438}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed economic dispatch of microgrids based on ADMM algorithms with encryption-decryption rules}, volume = {22}, year = {2025}, }
• (2025). A flexible job shop scheduling problem considering on-site machining fixtures: A case study from customized manufacturing enterprise. TASE, 22, 8415--8426. (www)

  @article{ , abstract = {The joint optimization of production scheduling and resource constraints is critical to modern manufacturing systems. The number of auxiliary resources (fixtures) is usually insufficient in customized manufacturing. Thus, on-site machining fixtures (Type II fixtures) should be prepared in the workshop to reduce the shortage. In this way, Type II fixtures are production tasks and resource constraints, while Type I fixtures are only resource constraints. The existing studies mainly concentrate on Type I fixtures, whereas the research on Type II fixtures is limited. Therefore, this paper focuses on a flexible job shop with on-site machining fixtures (FJSP-F). Firstly, a mathematical model is developed to minimize total weight tardiness (TWT). Secondly, a job-fixture-machine (JFM) encoding and novel decoding methods are presented to obtain a feasible schedule solution. Thirdly, an improved genetic algorithm (IGA4F) with problem-specific variable neighborhood search (PVNS) is proposed to balance the exploration and exploitation. Finally, the proposed algorithm is tested on 20 instances with comparison algorithms. The results demonstrate that IGA4F is a competitive algorithm in large-scale instances. From the case study results, the performance gains of the TWT and makespan obtained by IGA4F are 49.27% and 28.94% compared to the original schedule solution. Note to Practitioners—The integrated problem of fixture allocation and production scheduling is widespread in highly customized manufacturing enterprises, such as aerospace and shipbuilding. A well-balanced allocation between fixtures and machines can facilitate productivity and resource utilization. In general, Type I fixtures can be used directly if they are idle, and these fixtures are treated as resource constraints. However, due to the limited number of Type II fixtures, they are only available when finished in the workshop. Hence, Type II fixtures are considered production tasks and resource constraints, and the number of these fixtures is dynamic during the production cycle. Therefore, it is necessary for enterprise managers to investigate the effect of Type II fixtures on production scheduling. This paper proposes novel encoding and decoding methods to represent the solution and objective spaces. The evolutionary-based algorithm is proposed to solve the daily order of a real-world enterprise. The obtained results from the proposed algorithm can guide the managers to promote the workshop’s productivity.}, archive = {J_TASE}, author = {Jiahang Li and Xinyu Li and Liang Gao and Qihao Liu}, doi = {10.1109/TASE.2024.3485810}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8415--8426}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A flexible job shop scheduling problem considering on-site machining fixtures: A case study from customized manufacturing enterprise}, volume = {22}, year = {2025}, }
• (2025). Energy efficient scheduling for position reconfiguration of swarm drones. TASE, 22, 8400--8414. (www)

  @article{ , abstract = {Enhancing the energy efficiency of drones, particularly in extending the flight lifetime, has emerged as a crucial area. Position reconfiguration has been explored as a mechanism to achieve this goal for swarm drones. Building on this concept, we investigate how position reconfiguration can be applied within urban wind environments to further extend the lifetime of drone swarms. Despite its potential, efficiently implementing position reconfiguration remains challenging. To address it, we propose an efficient position reconfiguration scheme that reduces the energy consumption imbalance of the swarm and prolongs the lifetime. The scheme includes: (1) a MIP (mixed integer programming)-based optimization method. (2) an approximation algorithm that runs in pseudo-polynomial time and without the need for an optimization solver. The scheme provides a complete position reconfiguration solution that determines (i) the number of position reconfiguration; (ii) when to perform reconfiguration; (iii) who to change positions. Simulation and experimental results demonstrate the effectiveness of our scheme. Note to Practitioners—In urban environments, the significant variation in wind speeds leads to an energy imbalance among swarm drones performing tasks. This paper addresses the practical issue of extending the lifetime of drones in such environments by optimizing position reconfiguration. Specifically, drones operating in high wind speed areas require more energy to maintain hovering, resulting in faster battery depletion. By allowing drones with more remaining energy to exchange positions with those experiencing higher energy consumption, the overall energy usage can be balanced, thus extending the mission duration. We propose an energy-efficient scheduling scheme to determine when and which drones should reconfigure their positions. The scheme strikes a balance between the benefits of reconfiguration and the associated energy costs, preventing unnecessary movement that could waste energy while ensuring drones do not deplete their batteries prematurely. This solution is particularly suited for drone swarms operating in urban environments. Future research could further explore the integration of this scheme into real-time drone fleet management systems.}, archive = {J_TASE}, author = {Han Liu and Mingxin Wei and Shuai Zhao and Hui Cheng and Kai Huang}, doi = {10.1109/TASE.2024.3485681}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8400--8414}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy efficient scheduling for position reconfiguration of swarm drones}, volume = {22}, year = {2025}, }
• (2025). Environment-adaptive online learning for portable energy storage based on porous electrode model. TASE, 22, 8386--8399. (www)

  @article{ , abstract = {The dynamic conditions and internal states of portable energy storage system (PESS), such as temperature, electricity price, state of charge (SOC), and state of health (SOH), significantly impact battery degradation. Current decision-making models for PESS operation often oversimplify the modeling of battery degradation. To address this, we introduce an environment-adaptive online learning framework that effectively integrates deep neural networks and reinforcement learning to exploit and explore external environments (i.e., electricity prices and temperature) and internal dynamics (i.e., battery degradation), providing decision support for PESS operation. This framework dynamically updates battery degradation and decision-making models in real-time, enhancing adaptive responses to external changes. Specifically, we developed a neural network based on porous electrode theory that considers multi-physical factors, such as charging power, initial and terminal SOC, SOH, and temperature to accurately assess battery degradation. This network is embedded within a deep reinforcement learning algorithm, enabling real-time, adaptive decision-making for PESS amidst varying environmental conditions. Furthermore, to navigate complex operational environments, a fine-tuning mechanism is incorporated into the degradation neural network. Application of this framework to the energy arbitrage of PESS in the California power grid demonstrates an average benefit increase of 37% compared to traditional degradation assessment models. Note to Practitioners—In this work, we develop a novel approach to addressing the critical issue of battery degradation in PESS. Existing models often oversimplify degradation, hindering accurate assessments of performance and lifespan projections. More recently, learning-based algorithms have demonstrated outstanding performance in both battery degradation modeling and real-time decision-making. In this sense, we introduce a sophisticated neural network model grounded in a porous electrode model. This model considers multi-physics factors involving charging/discharging power, initial and terminal SOC, SOH, and temperature. Complementing this, we integrate the aforementioned model into an online learning framework, enabling real-time decision-making for PESS. Furthermore, to tackle the challenges of complex operating environments, a fine-tuning mechanism is incorporated into the battery degradation neural network. We validate the effectiveness of the proposed methods through an energy arbitrage application of PESS and also reveal its potential for on-demand applications in energy and transportation systems. This note anticipates a positive impact on PESS management and contributes significantly to the evolution of energy storage systems, offering practitioners invaluable decision support for commercial applications involving battery sharing, trading, and renting.}, archive = {J_TASE}, author = {Guannan He and Yongkang Ding and Zhengrun Wu and Xinjiang Chen and Da Zhang and Jie Song}, doi = {10.1109/TASE.2024.3485541}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8386--8399}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Environment-adaptive online learning for portable energy storage based on porous electrode model}, volume = {22}, year = {2025}, }
• (2025). Oscillation suppression-enhanced cooperative control via refined cooperative disturbance estimation for aerial co-transportation system. TASE, 22, 8371--8385. (www)

  @article{ , abstract = {This article focuses on the oscillation suppression-enhanced cooperative control design for the aerial co-transportation system consisting of two quadrotors and a tethered pipe. The system dynamics are analyzed in depth, which yields a decoupled model under multiple disturbances by utilizing the variation linearization technique and coordinate transformations. Based on this model, a refined cooperative disturbance estimation strategy is developed to capture the angle dynamics of the cables without direct measurements of swing angles. Then the estimation results are used for designing a cooperative control law to guarantee the performance in rapid suppression of the payload oscillation and in accurate positioning of the quadrotors under system uncertainties. The stability and convergence of the overall system is established using Lyapunov theory. Finally, experiments validate and demonstrate the superiority of the proposed method over the existing ones. Note to Practitioners—This paper is motivated by the requirement of safe control schemes for aerial co-transportation systems. The unexpected oscillation of the payload may result in serious accidents, and therefore efficiently suppressing the payload swing is the main concern of the research. Nevertheless, the cascaded underactuation property and the complicated couplings among the drones make it difficult to directly control the payload. Up till now, at the cost of additional weight and more complicated structure, most existing methods relying on extra sensors to detect the states of the payload for feedback control. Accounting for the foregoing problems, this article presents a novel sensorless control scheme for suppressing the payload oscillation. The cable angles are estimated using only the states of the drones. Moreover, cooperative control laws are designed based on the estimated results so that both antiswing and positioning performance are guaranteed. All these aspects are verified by rigorous theoretical analysis and hardware experiments.}, archive = {J_TASE}, author = {Lidan Xu and Hao Lu and Jianliang Wang and Hyondong Oh and Xiang-Gui Guo and Lei Guo}, doi = {10.1109/TASE.2024.3485237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8371--8385}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Oscillation suppression-enhanced cooperative control via refined cooperative disturbance estimation for aerial co-transportation system}, volume = {22}, year = {2025}, }
• (2025). Command filtered adaptive tracking consensus of random nonlinear multi-agent systems. TASE, 22, 8361--8370. (www)

  @article{ , abstract = {This article investigates the topic of adaptive tracking consensus for a family of nonlinear multi-agent systems modeled by random differential equations, which are different from well-known stochastic differential equations. This paper provides some primitive results on random nonlinear multi-agent systems. An improved backstepping method named command filtered control is adopted to derive the adaptive control law, where the convergence of the filtering error is guaranteed. To tackle the serious nonlinearities and uncertainties, a series of dynamic gains are introduced in the design process. The tracking errors for each follower concerning the output of the leader can be regulated arbitrarily to a small enough value by choosing appropriate tuning parameters. All signals of the closed-loop system are analyzed to have bounds almost surely. Moreover, the feasibility of the theory developed in this paper is validated by an example of numerical simulation. Note to Practitioners—Inspired by all kinds of biological clustering or grouping behaviors in our natural world, the research on multi-agent systems has long been popular in both theoretical and engineering scenarios. In addition, colored noises are ubiquitous and negligible when dealing with control problems of different engineering plants. Based on the above observation, this paper was motivated to realize the tracking consensus control of a class of nonlinear multi-agent systems disturbed by colored noise, which is also called random nonlinear systems, with the help of an enhanced adaptive backstepping approach called command filtered control. Although there exist nonlinearities and random noises in each follower agent, the objective of output consensus still could be achieved with the combination of the methods of dynamic gains and command filtering. The application potential of this research is considerable, such as drone formation performance and drone cruise. Our future research will further investigate control problems of random nonlinear multi-agent systems under some practical obstacles such as actuator failures.}, archive = {J_TASE}, author = {Ruipeng Xi and Zhipeng Shen and Hailong Huang and Huaguang Zhang}, doi = {10.1109/TASE.2024.3485167}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8361--8370}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Command filtered adaptive tracking consensus of random nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Event-based fuzzy stabilization controller for uncertain nonholonomic systems with output constraints: Application to mobile robots. TASE, 22, 8351--8360. (www)

  @article{ , abstract = {This article introduces a novel event-triggered fuzzy control scheme for chained nonholonomic systems with output constraints and drift uncertain nonlinearities. By utilizing a barrier Lyapunov function (BLF), the constraint requirements for the $x_{0}$ -subsystem are addressed. On this basis, the state-scaling technique is employed to transform the remaining x-subsystem into an equivalent lower-triangular system, and a systematic event-triggered scheme combining BLF-based backstepping is used to satisfy output constraints and reduce the communication burden. Meanwhile, a fuzzy logic system is applied to approximate the uncertainties. It is demonstrated that the developed stabilization control algorithm ensures the stability of uncertain nonholonomic systems and prevents any violations of output constraints throughout the control process. Besides, the Zeno behavior is successfully avoided. Finally, numerical simulations and experiments conducted on the QBot2e mobile robot are presented to illustrate the effectiveness and applicability of the proposed control algorithm. Note to Practitioners—Nonholonomic systems are prevalent in the engineering field, especially in wheeled mobile robots. The motivation of this article stems from the need to limit the spatial position of an indoor mobile robot, ensuring it operates within a reasonable area while also reducing the communication burden. To meet these requirements, an event-triggered mechanism combined with BLF-based backstepping is developed. This approach not only guarantees specific constraints, but also reduces resource consumption. Furthermore, the scope of the described systems is broadened by employing a fuzzy logic system to approximate the uncertainties of nonholonomic systems. The proposed controller achieves low energy consumption and superior performance, rendering it suitable for practical applications. Two groups of numerical simulations and experimental tests on the QBot2e mobile robot are carried out, verifying the feasibility and applicability of the algorithm. In summary, the event-triggered fuzzy control scheme effectively balances control performance and communication resource usage, addressing the practical control problem of stabilization in uncertain nonholonomic systems.}, archive = {J_TASE}, author = {Mengmeng Liu and Yuqiang Wu}, doi = {10.1109/TASE.2024.3485056}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8351--8360}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based fuzzy stabilization controller for uncertain nonholonomic systems with output constraints: Application to mobile robots}, volume = {22}, year = {2025}, }
• (2025). Fully distributed prescribed-time leader-following output consensus of heterogeneous multi-agent systems with dynamic event-triggered mechanism. TASE, 22, 8341--8350. (www)

  @article{ , abstract = {The fully distributed prescribed-time leader-following output consensus (FDPT-LFOC) of heterogeneous multi-agent systems (HMASs) with a dynamic event-triggered mechanism (ETM) is investigated in this paper. First, a fully distributed prescribed-time (FDPT) observer is developed for each agent to estimate the leader’s state by incorporating an adaptive gain and a time-varying scaling function. The adaptive gain is used to avoid the utilization of global information (e.g., the Laplace matrix eigenvalues, the number of agents) in the observer design, while the time-varying scaling function plays an important role in the PT accurate estimation for the leader’s state. Then, in order to avoid continuous communication among agents while implementing the FDPT-LFOC, a novel dynamic ETM containing a dynamic threshold and the time-varying scaling function is designed. A novel distributed PT (DPT) controller is developed based on the proposed FDPT observer and dynamic ETM. It is shown that the HMASs with the proposed DPT controllers and dynamic ETMs achieve the LFOC within a PT and maintain its consensus performance afterward, while the Zeno behavior can be excluded. Finally, the effectiveness of the proposed controller and dynamic ETM is verified by a simulation example. Note to Practitioners—This paper is motivated by achieving the FDPT-LFOC of HMASs with intermittent communication. Although various finite- and fixed-time consensus controllers have been developed for HMASs, the settling time for the controlled HMASs depends on their initial states and control parameters, which cannot be arbitrarily set according to the task requirements. There, it is difficult to apply these results to many time-critical fields, such as missile guidance, emergency braking, and so on. Moreover, most of the existing PT consensus results are focused on first-order or nonlinear MASs and dependent on continuous communication among neighboring agents, which also restricts their application in practical engineering. Therefore, in this paper, a DPT controller and a dynamic ETM are proposed for HMASs to deal with the above challenges.}, archive = {J_TASE}, author = {Hongpeng Li and Xinchun Jia and Xiaobo Chi and Bin Li}, doi = {10.1109/TASE.2024.3485052}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8341--8350}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fully distributed prescribed-time leader-following output consensus of heterogeneous multi-agent systems with dynamic event-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Distributed adaptive consensus control for nonlinear network systems with event-based switching mechanism against malicious attacks. TASE, 22, 8329--8340. (www)

  @article{ , abstract = {In this paper, the distributed adaptive consensus control problem for nonlinear network systems with unmatched unknown parameters is investigated. The communication channels among subsystems are directed and suffer from malicious attacks. Besides, only part of subsystems can access the reference states. To depict different kinds of attacks, a unified attack model is established from the viewpoint of attacked subsystems. Then, an event-based communication switching mechanism is proposed for subsystems to establish new communication channels, such that attack effects on these channels can be mitigated actively. Noting that malicious attacks and switching communication make transmitted states discrete, it is also a tough issue to design controllers by adopt the backstepping technique. To handle this problem and mitigate residual attack effects, a continuous virtual controller is designed by introducing normalization terms in the adaptive laws. Then, a distributed adaptive control scheme is proposed to guarantee that consensus errors are globally uniformly bounded under arbitrary switching dwell-time and malicious communication attacks. Experimental results are provided to validate the effectiveness of the proposed control scheme. Note to Practitioners—This paper is motivated by the secure consensus control problem for nonlinear network systems under malicious communication attacks. In the practice, the Frequency-Hopping Spread Spectrum (FHSS) technique has been used to defend malicious attacks. However, the switched communication induced by the FHSS technique may destroy the consensus performance. Moreover, how to economize the switching resource and improve the active defense ability of control schemes against malicious attack is still a tough issue. Motivated by these points, an active distributed adaptive consensus control scheme is proposed to defend malicious attacks, which contains an event-based switching mechanism and a novel secure controller. The event-triggered switching mechanism is designed with attack-sensitive functions to determine the switching time instants, such that the switching resources can be utilized more efficiently. By introducing a unified attack model, the attack-defense ability of the secure controller against different kinds of attacks can be improved and analyzed mathematically. Moreover, this controller can also be applied to the switched topologies with arbitrary switching dwell-time. In conclusion, this proposed distributed adaptive control scheme can greatly reduce the effects of malicious attacks and switched piecewise states on the consensus performance and controller design, such that engineers can pay more focuses on selecting suitable control parameters to adjust consensus performance. However, to enable above advantages, some bounded biases are involved in the consensus performance. In the future work, more accuracy attack-sensitive functions, optimization algorithms for parameters, switched secure controllers and unknown nonlinear functions satisfying Lipschitz condition can be considered to decrease above bounded biases and improve the consensus performance under malicious attacks.}, archive = {J_TASE}, author = {Zhen Han and Yue Zhou and Yutong Jiang and Ke Bao and Wenbin Yue}, doi = {10.1109/TASE.2024.3484512}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8329--8340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed adaptive consensus control for nonlinear network systems with event-based switching mechanism against malicious attacks}, volume = {22}, year = {2025}, }
• (2025). An unknown multiplayer nonzero-sum game: Prescribed-time dynamic event-triggered control via adaptive dynamic programming. TASE, 22, 8317--8328. (www)

  @article{ , abstract = {In this paper, the novel prescribed-time dynamic event-triggered control method of an unknown multiplayer nonzero-sum game (MP-NZSG) is designed by using adaptive dynamic programming (ADP). Firstly, a neural network-based identifier is constructed to estimate the unknown system dynamics. Subsequently, a novel ADP-based dynamic event-triggered control approach is advanced to ensure optimality and prescribed-time stability. A critic neural network (NN) is established for each player to approximate the Nash equilibrium solution of the dynamic event-triggered Hamilton-Jacobi-Isaacs (HJI) equation. This network employs a novel weight updating law, based on the experience replay technique, to alleviate the persistence of excitation condition. Furthermore, using the Lyapunov method, the uniform limit boundedness analysis of the neural network approximation error and multiplayer system is validated. Additionally, minimum inter-event time (MIET) is conclusively established to mitigate the notorious Zeno behaviour. Ultimately, the efficacy of the proposed method is rigorously substantiated through comprehensive simulation results. Note to Practitioners—Our research addresses the challenges of multi-component coordinated control, particularly in spacecraft attitude control. To handle these complexities, we propose an innovative adaptive dynamic event-triggered control approach. By integrating adaptive dynamic programming and neural networks, we effectively model and manage unknown system dynamics, enhancing the controller’s adaptability and robustness. Dynamic event-triggered policies are introduced to optimize system performance and reduce computational costs. The ADP-based prescribed time optimal control scheme prioritizes steady-state performance of nonlinear nonaffine systems, ensuring precise task completion within specified timeframes. Additionally, experience replay technology further fortifies the controller’s learning and adaptability to dynamic environments.}, archive = {J_TASE}, author = {Kun Zhang and Zhi-Xuan Zhang and Xiang Peng Xie and Jose de Jesus Rubio}, doi = {10.1109/TASE.2024.3484412}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8317--8328}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An unknown multiplayer nonzero-sum game: Prescribed-time dynamic event-triggered control via adaptive dynamic programming}, volume = {22}, year = {2025}, }
• (2025). Distributed cooperative framework for multiple UAVs safety: A capability-triggered mechanism. TASE, 22, 8303--8316. (www)

  @article{ , abstract = {This article develops a safety-driven distributed cooperative framework (SDDCF) for multiple unmanned aerial vehicles (UAVs) subject to actuator faults in the application of emergency search-and-rescue mission. A capability-triggered decision mechanism is proposed to conquer the challenging situation that the system redundancy cannot satisfy the requirement of fault-tolerant control. By quantitatively analyzing the capability of UAV, a safety threshold is provided, which can be updated adaptively in the light of performance requirement and real-time system capability estimated by a fixed-time fault observer. When the safety threshold is violated, the active performance degradation of the faulty UAVs and communication topology reconfiguration of the multiple UAVs are performed. By virtue of the SDDCF with capability-triggered mechanism, the safety of multiple UAVs system suffering from severe actuator faults is ensured for mission completion. The efficacy of the presented framework is demonstrated by a proof-of-concept emergency search-and-rescue mission in real-world flight experiments. Note to Practitioners—The proposed SDDCF is devoted to reduce the safety risk of multiple UAVs with severe actuator faults in emergency missions, where the mobility and reliability must be balanced carefully. Compared with the existing fault-tolerant control schemes, the SDDCF can ensure the safety even if the actuator faults exceed the system redundancy in a specific mission. Moreover, the practicability of the SDDCF, which can be extended to diverse task scenarios, has been verified in real-world flight experiments. In the future, the abilities of cooperative perception and risk avoidance should be improved to further enhance the safety of multiple UAVs in uncertain environments.}, archive = {J_TASE}, author = {Bin Yang and Sicheng Zhou and Jindou Jia and Ziqi Yang and Kexin Guo and Yi Yang and Xiang Yu and Youmin Zhang}, doi = {10.1109/TASE.2024.3483932}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8303--8316}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed cooperative framework for multiple UAVs safety: A capability-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Bionic design and control of a 12-DoF self-balancing walking exoskeleton. TASE, 22, 8292--8302. (www)

  @article{ , abstract = {Self-balancing walking exoskeletons (SBWEs), which enable paralyzed individuals to walk without assistive devices, have been increasingly employed in rehabilitation training. This paper proposes a Kelvin-Voigt viscoelastic model-based bioinspired viscoelastic compliance controller (BVCC) for a novel SBWE named AutoLEE-II, which features high structural rigidity, low leg inertia relative to center of mass (CoM), and small hip joint axis misalignment between the user and SBWE. First, a novel series-parallel hybrid mechanism is designed for AutoLEE-II. This mechanism, inspired by the lower limbs of humans, reduces hip axis misalignment between the user and SBWE, decreases leg inertia relative to CoM, and improves structural stiffness. Second, a BVCC mimicking biological muscle is proposed to introduce viscoelastic compliance to SBWE to maintain locomotion stability of the SBWE during standing and walking. The BVCC is robust to the variable physical parameters of different users. Finally, self-balancing walking experiments are conducted with AutoLEE-II with empty load, manikin load and human subject load to validate the performance of AutoLEE-II and the proposed compliance controller BVCC. Note to Practitioners—This paper aims to design a self-balancing walking exoskeleton (SBWE) that provides rehabilitation training exercise and walking assistance services for individuals with hemiplegia, paraplegia, and quadriplegia. First, we biomimetically designed the mechanical structure of the SBWE, named AutoLEE-II based on the distribution of human joints and connecting links. The bionic mechanism reduces axis misalignment between the SBWE and users, improves stiffness and reduces the inertia of the legs relative to center of mass. We then designed a bioinspired viscoelastic compliance controller (BVCC) based on the centroid dynamics model, which is robust to the physical properties of the user and introduces the SBWE with active compliance. Finally, self-balancing walking experiments with an empty load, a manikin load and human subject loads are performed to validate mechanical structure of the proposed AutoLEE-II and the locomotion stability of the physical parameter robust BVCC.}, archive = {J_TASE}, author = {DingKui Tian and Yong He and Ziqiang Chen and Feng Li and Jinke Li and Meng Yin and Li Zhang and XinYu Wu}, doi = {10.1109/TASE.2024.3483872}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8292--8302}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bionic design and control of a 12-DoF self-balancing walking exoskeleton}, volume = {22}, year = {2025}, }
• (2025). Optimal control of internally forced switching systems with guaranteed feasibility. TASE, 22, 8278--8291. (www)

  @article{ , abstract = {An efficient dynamic optimization approach for internally forced switching systems is provided in this work. The distinguishing characteristic of the internally forced switching systems is that once the trajectory of the state of a mode hits a switching surface in the state space, the current mode stops operating immediately and then the next mode is activated automatically. To effectively address the optimal control of such systems, first, the continuous control input is approximated by piecewise constant functions utilizing the control vector parametrization (CVP) technique. Sensitivity analysis is subsequently used to derive the gradient of the cost function with respect to (w.r.t.) parameterized control input, while the state transition matrix is introduced for determining the gradients of the objective function w.r.t. switching instants. Second, a dynamic optimization method with the aid of gradient information is presented to locate the optimal solution for internally forced switching systems with a guarantee of rigorously satisfying the path constraints. Third, it is demonstrated that the designed algorithm terminates finitely to generate a feasible solution satisfying the Karush-Kuhn-Tucker (KKT) conditions of the dynamic optimization of internally forced switching systems to a specified tolerance. Finally, the proposed optimization approach is applied to the fed-batch fermentation process to obtain a high concentration of 1,3-propanediol production, while ensuring that the glycerol concentration rigorously satisfies the required boundary during the whole feed process. Note to Practitioners—The motivation of this work is to develop an efficient optimization method for path-constrained internally forced switching systems, which have a wide range of applications in practice, e.g., obstacle avoidance robots, chemical batch feed fermentation, etc. Although research methods for dynamic optimization of switched systems have been well-established, to the best of the author’s knowledge, there is almost no research directly dealing with internally forced switching systems. Since the switching instant of such systems is strongly dependent on the inputs and the state of the system, which makes it difficult to analyze the gradient of the objective function w.r.t. the switching instant. Moreover, it is necessary to consider the inequality path constraints that guarantee the requirements such as system safety and product quality. Therefore, this work aims at the dynamic optimization of the path-constrained internal forced switching system, considering the two important requirements of rigorous satisfaction of path constraints and a finite number of iterations of the algorithm. An efficient dynamic optimization approach is proposed based on the variational method, semi-infinite program technique, and right-handed constraint method, which can simultaneously obtain the optimal switching instants and optimal control inputs while guaranteeing that the path constraints are rigorously satisfied.}, archive = {J_TASE}, author = {Huan Li and Ying Jin and Jun Fu}, doi = {10.1109/TASE.2024.3483748}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8278--8291}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal control of internally forced switching systems with guaranteed feasibility}, volume = {22}, year = {2025}, }
• (2025). A fast finite-time composite control with dynamic prescribed performance for tracking in task space of stewart parallel mechanism. TASE, 22, 8262--8277. (www)

  @article{ , abstract = {To improve tracking control performance in the task space of the Stewart parallel mechanism (SPM) with uncertainties of modeling errors and external disturbances in our developed rust removal sandblasting robot for steel box girder, a fast finite-time composite control method with dynamic prescribed performance (DPP) is proposed. Firstly, by introducing only one exponential power regulation term with a variable exponent coefficient (VEC) into Lyapunov differential inclusion, a VEC finite-time stability theorem for the multi-input multi-output nonlinear system is proposed. Secondly, based on the proposed theorem, a VEC finite-time disturbance observer and a VEC finite-time super-twisting control are respectively designed and combined into a fast finite-time composite control, to enhance system robustness and achieve fast and smooth finite-time convergence for SPM tracking in task space. Finally, the fast finite-time composite control with DPP is formed by devising a dynamic asymmetric prescribed performance function and a modified barrier-function-based constraint transformation function and integrating them into the fast finite-time composite control, to effectively constrain the tracking error into small ranges under time-varying uncertainties, achieve non-overshoot transient performance and high steady-state tracking accuracy, provide continuous and smooth control input torque, and reduce input energy consumption. The effectiveness is validated by simulation and prototype experiments. Note to Practitioners—To improve the control performance of the Stewart parallel mechanism (SPM) with uncertainties of modeling error and external disturbances, a variable exponent coefficient (VEC) finite-time stability theorem is firstly proposed. On this basis, a VEC finite-time disturbance observer and a VEC finite-time super-twisting control algorithm are respectively designed and combined into a fast finite-time composite control method, in order to achieve fast and smooth bounded-time convergence performance for the SPM tracking in task space. Furthermore, a dynamic asymmetric prescribed performance function and a modified barrier-function-based constraint transformation function are respectively devised and introduced into the fast finite-time composite control, aiming to make it further realize non-overshoot transient performance while providing smooth and non-excessive control inputs and reducing input energy consumption.}, archive = {J_TASE}, author = {Yu Tang and Guoqin Gao}, doi = {10.1109/TASE.2024.3483268}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8262--8277}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A fast finite-time composite control with dynamic prescribed performance for tracking in task space of stewart parallel mechanism}, volume = {22}, year = {2025}, }
• (2025). Event-triggered secure control under aperiodic DoS attacks. TASE, 22, 8252--8261. (www)

  @article{ , abstract = {This paper is focused on the event-based secure control issue for cyber-physical systems (CPSs) under aperiodic denial-of-service (DoS) attacks. Malicious DoS attacks disrupt the communication between the controller and the actuator. The finite attack resources of malicious attackers are taken into consideration, and the DoS attacks are characterized using an aperiodic model. In contrast to prior results, the present study tackles the issue of secure controller design by considering the attributes of the DoS attack, instead of employing a switched system approach to address the aforementioned concerns. More specifically, under aperiodic DoS attacks, sufficient criteria are established to guarantee that the closed-loop CPSs can achieve bounded stability. Then, within a time-varying attack period, the relationship between the attack active interval and the attack silent interval is derived. Without satisfying the derived conditions, the system’s stability will deteriorate. Moreover, an event-based secure control scheme under aperiodic DoS attacks is designed. To verify the efficacy of the derived theory, a wheeled mobile robot system under aperiodic DoS attacks is illustrated. Note to Practitioners—CPSs have been widely utilized in various domains, such as aerospace and intelligent transportation. However, the openness of networks provides attackers with numerous opportunities for malicious assaults, consequently leading to a degradation in system performance. Consequently, researching the security issues of CPSs under malicious attacks is of utmost urgency. This paper focuses on the issue of event-triggered secure control for CPSs in the presence of energy-constrained aperiodic DoS attacks. The event-triggered communication mechanism is introduced to reduce the computational burden. The criteria for ensuring the bounded stability of CPSs under aperiodic DoS attacks are proposed. The relationship between the attack active interval and the attack silent interval is derived, which is incorporated into the proposed criteria. A wheeled mobile robot system is given to validate the effectiveness of the proposed method. In the future, an active defense control method will be proposed to counter malicious attacks.}, archive = {J_TASE}, author = {Liyuan Yin and Chengwei Wu and Lezhong Xu and Hongming Zhu and Xiangyu Shao and Weiran Yao and Jianxing Liu and Ligang Wu}, doi = {10.1109/TASE.2024.3483169}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8252--8261}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered secure control under aperiodic DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Defect image sample generation with diffusion prior for steel surface defect recognition. TASE, 22, 8239--8251. (www)

  @article{ , abstract = {The task of steel surface defect recognition is an industrial problem with great industry values. The data insufficiency is the major challenge in training a robust defect recognition network. Existing methods have investigated to enlarge the dataset by generating samples with generative models. However, their generation quality is still limited by the insufficiency of defect image samples. To this end, we propose Stable Surface Defect Generation (StableSDG), which transfers the vast generation distribution embedded in Stable Diffusion model for steel surface defect image generation. To tackle with the distinctive distribution gap between steel surface images and generated images of the diffusion model, we propose two processes. First, we align the distribution by adapting parameters of the diffusion model, adopted both in the token embedding space and network parameter space. Besides, in the generation process, we propose image-oriented generation rather than from pure Gaussian noises. We conduct extensive experiments on steel surface defect dataset, demonstrating state-of-the-art performance on generating high-quality samples and training recognition models, and both designed processes are significant for the performance. Note to Practitioners—This article introduces StableSDG, a method that generates realistic defect images even with limited data. It overcomes the shortcomings of current deep learning approaches that need large datasets to train from scratch. Our solution is to adapt a text-to-image diffusion model for defect generation. The proposed strategy involves two processes: training to adapt token embeddings and model parameters, and generation from partially perturbed defect images. The results show enhanced generation quality and improved accuracy for recognition models trained on the expanded dataset. StableSDG can be practically applied to efficiently enlarge a defect dataset, even when starting with a small amount of data.}, archive = {J_TASE}, author = {Yichun Tai and Kun Yang and Tao Peng and Zhenzhen Huang and Zhijiang Zhang}, doi = {10.1109/TASE.2024.3482362}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8239--8251}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Defect image sample generation with diffusion prior for steel surface defect recognition}, volume = {22}, year = {2025}, }
• (2025). Noise aware path planning and power management of hybrid fuel UAVs. TASE, 22, 8227--8238. (www)

  @article{ , abstract = {Hybrid fuel Unmanned Aerial Vehicles (UAV), through their combination of multiple energy sources, offer several advantages over the standard single fuel source configuration, the primary one being increased range and efficiency. Multiple power or fuel sources also allow the distinct pitfalls of each source to be mitigated while exploiting the advantages within the mission or path planning. We consider here a UAV equipped with a combustion engine-generator and battery pack as energy sources. We consider the path planning and power-management of this platform in a noise-aware manner. To solve the path planning problem, we first present the Mixed Integer Linear Program (MILP) formulation of the problem. We then present and analyze a label-correcting algorithm, for which a pseudo-polynomial running time is proven. Results of extensive numerical testing are presented which analyze the performance and scalability of the labeling algorithm for various graph structures, problem parameters, and search heuristics. It is shown that the algorithm can solve instances on graphs as large as twenty thousand nodes in only a few seconds. Note to Practitioners—The problem and algorithms proposed in this paper are relevant to constrained planning problems in general and specifically to those focused on widespread usage of small aerial vehicles in congested, urban environments. We are concerned here with the path planning of hybrid-fuel aerial vehicles in a noise-aware manner. This is motivated by the increasing usage of aerial vehicles, envisioning a probable future restriction on noise production in certain airspaces and the planning of such vehicles in those airspaces. We explore this novel problem, and present an approach to quickly find the optimal path and power plan in the presence of the noise constraints. The approach here is a discrete one, where the solution is a discrete set of edges and discrete generator settings which must be smoothed to obtain control inputs for a real system. The discrete approach allows solutions to be found quickly while giving up the true optimal trajectory that can be found when considering from a continuous framework. In practice, environment sampling and graph construction will greatly affect time-to-solve and as overall solution quality relative to a continuous approach to the trajectory and generator control.}, archive = {J_TASE}, author = {Drew Scott and Satyanarayana G. Manyam and Isaac E. Weintraub and David W. Casbeer and Manish Kumar}, doi = {10.1109/TASE.2024.3481998}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8227--8238}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Noise aware path planning and power management of hybrid fuel UAVs}, volume = {22}, year = {2025}, }
• (2025). Automated motion control of a microparticle swarm in liquids by a single-solenoid electromagnetic manipulation system. TASE, 22, 8215--8226. (www)

  @article{ , abstract = {In recent years, magnetic field-controlled microparticles have demonstrated their superiority in biomedical applications. To improve the operation efficiency and imaging effect of microparticles, increasing numbers of researchers are devoted to driving a swarm composed of a group of microparticles. In this study, a new strategy for autonomously manipulating microparticle swarms in liquids via an electromagnetic coil system was investigated. Compared with the commonly used rotating magnetic field composed of multiple electromagnetic coils, the oscillating magnetic field drive system formed by a single electromagnetic coil reduces the occupied space meanwhile ensures the sufficient working space, making the microparticle movement more flexible while providing greater feasibility for good compatibility with imaging systems. Through designing the parameters of the input current, a vortex-shaped swarm can be formed, and the movement of the entire swarm can be pulled by controlling the coil position. Owing to the rotation of microparticles, the coefficient of the control input is unknown uncertainty. To solve this uncertainty problem, a sliding mode controller is used, and the chattering caused by the sliding mode surface is reduced using a saturation function. Lastly, the effectiveness of the proposed strategy is verified by simulation and experiments. Note to Practitioners—The motivation of this article is mainly to solve the problem of magnetic microparticle aggregation, aiming to form a stable swarm and automatically control the movement of the swarm. Most of the existing methods of aggregating microparticles to form a swarm are based on the creation of a rotating magnetic field by Helmholtz coils, which leads to the problem of spatial limitation. To solve this problem, this research uses one single solenoid to form a stable swarm, which provides more space for imaging equipment and objects requiring operation. The controlled movement of the stable swarm is achieved by manipulating the solenoid with a sliding mode controller. It reduces the impact of system uncertainties, allowing the swarm to automatically track a predefined path accurately. This will provide a new idea for the application of microparticles in the biomedical field.}, archive = {J_TASE}, author = {Dingran Dong and Jianing Li and Qi Zhang and Dong Sun}, doi = {10.1109/TASE.2024.3481660}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8215--8226}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated motion control of a microparticle swarm in liquids by a single-solenoid electromagnetic manipulation system}, volume = {22}, year = {2025}, }
• (2025). Adaptive dynamic event-triggered asymptotic tracking control for strict-feedback nonlinear systems with a self-adjusting performance function. TASE, 22, 8200--8214. (www)

  @article{ , abstract = {The issue of adaptive dynamic event-triggered asymptotic tracking control for strict-feedback nonlinear systems with unknown functions and full-state prescribed-performance constraints is discussed in this paper. A novel self-adjusting performance function (SAPF) is constructed by merging a continuous function with a finite-time performance function. By associating a transformation function and SAPF, the state constraints problem is recast into analyzing the boundedness of the new variables. Moreover, a new dual dynamic variable-dependent event-triggered mechanism is provided to reduce redundant data transmission. By using a command-filter technique and neural network method, a control scheme is proposed to guarantee the system output asymptotically tracks the reference signal and all system states satisfy specified constraints. Lastly, an applied example is introduced to illustrate the effectiveness of the proposed scheme. Note to Practitioners—In reality, many practical systems, such as flexible manipulators and unmanned underwater vehicles, need to operate in a constrained region. Motivated by this, we design controller for nonlinear systems to guarantee steady-state and transient performance. To reduce the conservatism, a novel SAPF is constructed and then a prescribed performance control strategy is proposed, which can not only achieve the asymptotical tracking of the nonlinear system, but also ensure the constraint performance. In addition, to decrease the transmitted data via communication network, a new dual dynamic variable-dependent event-triggered mechanism is proposed. This method has been illustrated to be feasible via a simulation example.}, archive = {J_TASE}, author = {Haibin Sun and Xiangling Kong and Linlin Hou and Dong Yang and Yunliang Wei}, doi = {10.1109/TASE.2024.3481502}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8200--8214}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive dynamic event-triggered asymptotic tracking control for strict-feedback nonlinear systems with a self-adjusting performance function}, volume = {22}, year = {2025}, }
• (2025). Reinforcement learning-based 3D trajectory tracking control of hypersonic gliding vehicles with time-varying uncertainties. TASE, 22, 8187--8199. (www)

  @article{ , abstract = {In this paper, a robust three-dimensional trajectory tracking control scheme based on reinforcement learning is proposed for the glide phase of a hypersonic gliding vehicle (HGV) with time-varying uncertainties. First, the non-affine nonlinear full-state kinematics and dynamics model of the HGV glide phase is constructed. Then, without linearizing the system, the desired multiplanar reference trajectories for HGVs are planned based on the pseudo-spectral theory under the input constraints, initial conditions, and terminal conditions. Subsequently, the full-state error system is generated by subtracting the reference system state from the actual state of the HGV system with time-varying uncertainty. For the full-state HGV error system with time-varying uncertainty and input constraints, we design a reinforcement learning-based optimal control scheme for its nominal system and establish the equivalence between this optimal control and the robust control of the original HGV error system. A single-evaluation network structure is used in the concrete implementation to reduce the computational cost. A rigorous theory is given to demonstrate the uniform ultimate boundedness of the closed-loop system and the weight error. Finally, we perform simulation traces for reference trajectories with different optimization performances to verify the effectiveness of the proposed method. Note to Practitioners—There are various constraints and uncertainties in the glide phase of HGVs, which is the hinge connecting the initial descent phase and the terminal management phase. How to design robust trajectory tracking controllers for the glide phase of HGVs with complex environments and large span of flight parameters is of great significance to aerial guidance practitioners. In this paper, an RL-based three-dimensional trajectory robust tracking guidance method is proposed for the HGV glide phase system, which can resist time-varying uncertainties and satisfy flight constraints. The uniform ultimate boundedness of the closed-loop system is proved using the Lyapunov method. The proposed tracking algorithm is effective for reference trajectories with different performance indexes.}, archive = {J_TASE}, author = {Biao Luo and Jingyi Sun and Rui Tang and Xiaodong Xu}, doi = {10.1109/TASE.2024.3481422}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8187--8199}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reinforcement learning-based 3D trajectory tracking control of hypersonic gliding vehicles with time-varying uncertainties}, volume = {22}, year = {2025}, }
• (2025). Robust stability analysis of the interactions in multi-terminal HVDC systems using ν-gap metric. TASE, 22, 8174--8186. (www)

  @article{ , abstract = {Stability analysis of multi-terminal HVDC systems is an important and challenging issue for the safety and stable operation of power systems with large-scale renewable energy generations. It is quite difficult to analyze the system stability caused by the interactions among different VSCs with the coupling of DC and AC networks. This paper aims to present a method to analyze the relative stability of different interactions in multi-terminal HVDC systems. By the $\nu $ -gap metric based robust stability theory, firstly we propose a stability index to quantify the stability margin with respect to different paths of interactions. The influence of the interactions among different VSCs on the stability margin can be analyzed quantitatively. Then we present a method to easily calculate the stable region of parameters in the interactions. Extensive examples are given to demonstrate the application and the effectiveness of the proposed method. Note to Practitioners—This paper is motivated by the problem of improving the small-signal stability of multi-terminal high voltage direct current (MTDC) systems, which are widely used to transmit large-scale renewable energy generations. In recent years, many small-signal instability accidents have occurred in actual HVDC projects. The dynamic behavior of MTDC systems is related to the complex interactions among different voltage source converters (VSCs) through the coupling of AC and DC networks. Therefore, we aim to explain the mechanism of oscillations caused by the interactions and the relation between the dynamics and the control parameters. In this paper, we employ the $\nu $ -gap metric based robust control theory to quantify the stability of an MTDC system with respect to different interactions and to calculate the stable region of a particular parameter in the interactions. The results in this paper can explain the influence of the interactions among different VSCs on the stability margin quantitatively. Moreover, the results provide new ideas for the setting and tuning of control parameters in MTDC systems and even in other multi-equipment power systems.}, archive = {J_TASE}, author = {Wanning Zheng and Li Chai and Bing Liu}, doi = {10.1109/TASE.2024.3481009}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8174--8186}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust stability analysis of the interactions in multi-terminal HVDC systems using ν-gap metric}, volume = {22}, year = {2025}, }
• (2025). Data-driven weighted h∞ control of persistent dwell time switched systems with optimal disturbance attenuation guaranteed. TASE, 22, 8162--8173. (www)

  @article{ , abstract = {Persistent dwell-time switched systems (PDTSSs) are being increasingly employed for modeling the dynamics of systems with non-uniform time-dependent switching rules. The existing attempts to design control methods for PDTSSs necessitate a prior knowledge of system dynamics. In this article, we propose a data-driven reinforcement learning (RL) method to solve the weighted ${\mathcal {H}}_{\infty }$ control problem for PDTSSs with completely unknown system dynamics. The proposed method aims to calculate the weighted ${\mathcal {H}}_{\infty }$ control gain for PDTSSs with optimal disturbance attenuation guaranteed by solving a set of linear matrix inequalities related to system data, as opposed to solving it with the exact model information. To ensure the stability of the overall switched system, we establish the criterion for exponential mean-square stability of the closed-loop PDTSSs, as well as analyze the convergence of the proposed data-driven RL algorithm. Finally, the efficacy of the designed algorithms is illustrated via an electric circuit model. Note to Practitioners—First, at a practical level, we consider a more general PDT switching rule that can effectively model the phenomenon of simultaneous fast and slow switching in real systems, which finds wide application in various engineering domains, including capacitor monolithic filters, robot manipulators, and power grids. Then, we aim to address the challenges associated with acquiring accurate system model information and ensuring stable operation with optimal disturbance attenuation performance. On the one hand, the existing $ {\mathcal {H}}_{\infty }$ control methods of PDTSSs only guarantee system stability with a prescribed level of disturbance attenuation, and achieving mean-square stability while maintaining an optimal level of disturbance attenuation remains a challenge. On the other hand, existing $\mathcal {H} _{\infty }$ control methods incorporate precise system dynamics that are difficult or costly to acquire in practical systems. To tackle the aforementioned challenges, we propose a data-driven weighted $\mathcal {H} _{\infty }$ control approach for PDTSSs with unknown system dynamics. This method ensures that the closed-loop PDTSSs exhibit exponential mean-square stability while achieving an optimal disturbance attenuation level. Furthermore, by leveraging RL algorithms, our controller design process only relies on system data rather than precise system dynamics.}, archive = {J_TASE}, author = {Jiacheng Wu and Bosen Lian and Hongye Su and Yang Zhu}, doi = {10.1109/TASE.2024.3480449}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8162--8173}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven weighted h∞ control of persistent dwell time switched systems with optimal disturbance attenuation guaranteed}, volume = {22}, year = {2025}, }
• (2025). Multi-channel time series decomposition network for generalizable sensor-based activity recognition. TASE, 22, 8150--8161. (www)

  @article{ , abstract = {Sensor-based human activity recognition is important in daily scenarios such as smart healthcare and homes due to its non-intrusive privacy and low cost advantages, but the problem of out-of-domain generalization caused by differences in focusing individuals and operating environments can lead to significant accuracy degradation on cross-person behavior recognition due to the inconsistent distributions of training and test data. To address the above problems, this paper proposes a new method, Multi-channel Time Series Decomposition Network (MTSDNet). Firstly, MTSDNet decomposes the original signal into a combination of multiple polynomials and trigonometric functions by the trainable parameterized time series decomposition to learn the low-rank representation of the original signal for improving the extraterritorial generalization ability of the model. Then, the different components obtained by the decomposition are classified layer by layer and the layer attention is used to aggregate components to obtain the final classification result. Extensive evaluation on DSADS, OPPORTUNITY, PAMAP2, UCIHAR and UniMib public datasets shows the advantages in classification accuracy and stability of our method compared with other competing strategies, including the state-of-the-art ones. And the visualization is conducted to reveal MTSDNet’s interpretability and layer-by-layer characteristics. Note to Practitioners—This paper is motivated by solving the problem of decreased classification accuracy caused by differences in human activity, such as the differences in walking patterns between young and elderly people. Models trained on young people’s data are difficult to accurately recognize the activities of the elderly. This is a problem of domain generalization, where the distribution differences between training and testing data require the model’s ability to generalize across domains. We propose a model based on time series decomposition, which helps the model learn universal features and effectively improve its generalization ability. This method can be further extended to time series tasks with varying domain distributions.}, archive = {J_TASE}, author = {Jianguo Pan and Zhengxin Hu and Lingdun Zhang and Xia Cai}, doi = {10.1109/TASE.2024.3480119}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8150--8161}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-channel time series decomposition network for generalizable sensor-based activity recognition}, volume = {22}, year = {2025}, }
• (2025). Learning control affine neural NARX models for internal model control design. TASE, 22, 8137--8149. (www)

  @article{ , abstract = {This paper explores the use of Control Affine Neural Nonlinear AutoRegressive eXogenous (CA-NNARX) models for nonlinear system identification and model-based control design. The idea behind this architecture is to match the known control-affine structure of the system to achieve improved performance. Coherently with recent literature of neural networks for data-driven control, we first analyze the stability properties of CA-NNARX models, devising sufficient conditions for their incremental Input-to-State Stability ( $\delta $ ISS) that can be enforced at the model training stage. The model’s stability property is then leveraged to design a stable Internal Model Control (IMC) architecture. The proposed control scheme is tested on a real Quadruple Tank benchmark system to address the output reference tracking problem. The results achieved show that (i) the modeling accuracy of CA-NNARX is superior to the one of a standard NNARX model for given weight size and training epochs, (ii) the proposed IMC law provides performance comparable to the ones of a standard Model Predictive Controller (MPC) at a significantly lower computational burden, and (iii) the $\delta $ ISS of the model is beneficial to the closed-loop performance. Note to Practitioners—Many engineering systems, such as robotic manipulators and chemical reactors, are described by Control Affine (CA) models, characterized by onlinear dynamics where the control variable enters in a linear way. If only this structural information is available without any additional knowledge, for instance on the order of the system or on the value of its parameters, a black-box identification approach can be followed to estimate the model from data. For these reasons, in this paper we propose a modeling and control design method suited for this class of systems. Specifically, we assume that the system is described by a CA-Neural Nonlinear AutoRegressive eXogenous (CA-NNARX) model. Then, the estimated model is used to design a stable Internal Model Control (IMC) scheme for the solution of output reference tracking problems. The stability, performance, and robustness properties of the proposed approach are studied and tested in the control of a laboratory system. In addition, a simulation analysis shows how IMC represents a valid alternative to the popular Model Predictive Control (MPC) approach, in particular for embedded systems, where the computation power required by MPC can be too high.}, archive = {J_TASE}, author = {Jing Xie and Fabio Bonassi and Riccardo Scattolini}, doi = {10.1109/TASE.2024.3479321}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8137--8149}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning control affine neural NARX models for internal model control design}, volume = {22}, year = {2025}, }
• (2025). High-accuracy adaptive robust fault-tolerant control for quadrotor with actuator uncertainties and aerodynamic drag compensation. TASE, 22, 8123--8136. (www)

  @article{ , abstract = {With the expansion of application range of quadrotors, high-performance safety flight is getting more attention, where the health of actuators is critical. However, based on the commonly used loss of efficiency models, such fault-tolerant control methods are limited in performance to deal with different types of actuator faults in targeted ways. In this paper, by fully utilizing the more detailed and accurate models, the proposed adaptive robust fault-tolerant control has strong fault tolerance ability while maintaining excellent trajectory tracking accuracy. Firstly, the actuator model is established including motor dynamics and propeller model, which can reflect different types of actuator faults to the changes in different physical parameters instead of single-type efficiency factors. Additionally, aerodynamic drag is explicitly considered in quadrotor dynamics for improving control accuracy. Then, adaptive robust control is developed on these bases with comprehensive adaptation mechanism. To be specific, actuator parameters are actively estimated by recursive least square, so that actuator faults can be compensated directly in a targeted way without fault diagnosis. Subsequently, aerodynamic drag is effectively compensated through gradient-type adaptation, while the remaining uncompensated uncertainties are further suppressed by robust feedback. Finally, the comparative experiments demonstrate that the proposed method achieves much higher control accuracy than other compared methods, and it can maintain the same level of accuracy in faulty and fault-free case. Upon sudden faults, the proposed method exhibits the fastest response speed with minimal positional overshoot. Note to Practitioners—This paper aims to improve the flight safety of quadrotors while maintaining good trajectory tracking accuracy. Considering the most critical factor, i.e., actuator faults, it is usually modeled as loss of efficiency in most of literature on fault-tolerant control. However, the causes of faults are diverse, and it is not possible to effectively compensate for all kinds of faults through single-type efficiency coefficients. In fact, the occurrence of faults is associated with a change in a certain parameter of the system. Based on this, by fully utilizing the dynamics model of quadrotor and actuator, the effective online adaptive estimation algorithm is designed specifically for key actuator parameters related to faults. Therefore, more targeted compensation can be achieved for actuator faults caused by different reasons. In addition, those parameters that cannot be identified in advance but has great impact on accuracy, such as aerodynamic drag coefficients, are also explicitly considered and adaptively compensated. As for the remaining uncompensated uncertainties including external disturbances, robust feedback is introduced to ensure stability against them. The experimental results indicate that, for different types of actuator faults, the proposed method can achieve excellent trajectory tracking accuracy comparable to the fault-free cases. The proposed method can also be applied to other kinds of autonomous vehicles, such as underwater vehicles and aerospace vehicles.}, archive = {J_TASE}, author = {Weisheng Liang and Zheng Chen and Bin Yao}, doi = {10.1109/TASE.2024.3479294}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8123--8136}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-accuracy adaptive robust fault-tolerant control for quadrotor with actuator uncertainties and aerodynamic drag compensation}, volume = {22}, year = {2025}, }
• (2025). Complementary sliding mode control using petri probabilistic fuzzy recurrent neural network for active power filter. TASE, 22, 8108--8122. (www)

  @article{ , abstract = {Current loop control of active power filter (APF) is vital for its harmonic suppression. In this paper, a complementary sliding mode controller (CSMC) with a petri probabilistic fuzzy recurrent neural network (PPFRNN) is designed for current control and harmonic suppression of an APF. Compared with the traditional sliding mode control (SMC), CSMC has less chattering and higher control accuracy. By combining the advantages of many kinds of networks, a new PPFRNN scheme is designed to estimate unknown nonlinear terms in the APF dynamic model, so as to reduce the chattering and further improve the performance of sliding mode controller. Simulation and hardware experiments proved the feasibility and superiority of the proposed method, showing it has better harmonic suppression, steady-state and dynamic performance compared with the existing methods. Note to Practitioners—This paper was motivated by the problem of power quality control using active power filter. a PPFRNN based ICSMC is proposed for harmonic suppression of APF. A CSMC is chosen due to the mathematical model of APF is difficult to be obtained in the practical application. However, the selection of parameter of traditional CSMC must balance the chattering problem and controller performance. Hence, a PPFRNN is introduced to reduce the burden of CSMC suppressing uncertainty of APF system, which can alleviate the contradiction between chattering problem and controller performance essentially. Finally, detail simulations and experiments verified the proposed ICSMC has a good harmonic suppression capability and small output chattering.}, archive = {J_TASE}, author = {Juntao Fei and Jiacheng Wang and Lei Zhang}, doi = {10.1109/TASE.2024.3478775}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8108--8122}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Complementary sliding mode control using petri probabilistic fuzzy recurrent neural network for active power filter}, volume = {22}, year = {2025}, }
• (2025). Geometry and force guided robotic assembly with large initial deviations for electrical connectors. TASE, 22, 8095--8107. (www)

  @article{ , abstract = {Electrical connectors (ECs) are extensively employed in industrial scenarios, and their assembly quality is crucial. However, these connectors are often located in confined spaces, which poses challenges of complex lighting conditions and visual occlusions during the execution of robotic assembly tasks. Hence, guiding robots to assemble solely through force/torque (F/T) feedback is an alternative way. However, there is currently limited research on achieving assembly tasks solely through F/T information, especially when the spatial pose of the socket is uncertain, and how to achieve the robotic assembly of ECs remains a difficult problem. To this end, this paper proposes a novel strategy for assembling ECs under large initial pose deviations. Specifically, inspired by observing the assembly process of humans without visual assistance, the robotic assembly is first divided into two stages: by arbitrary surface tracking, the relative pose of the current EC is confirmed and adjusted to a dual-point contact state, i.e. the conversion from non-contact to dual-point contact; by setting heuristic F/T, the alignment of the edge, plane and slots of EC is driven, i.e. the alignment of the plug and socket. Next, we analyse the geometric constraint states at these stages and formulate the corresponding desired contact F/T strategies. To our knowledge, this may be the first attempt to achieve the robotic assembly of ECs under large initial deviations solely using F/T information. Finally, experiments on a UR5 robot indicate that the proposed strategy exhibits robustness to large initial pose deviations and can overcome obstacles caused by friction and jamming to ensure the robotic assembly of ECs. Note to Practitioners—The automatic assembly of ECs in confined spaces with visual occlusions remains an unsolved challenge, especially in aerospace scenarios where manual assembly may not be appropriate. The proposed assembly strategy is based on F/T perception without visual assistance, which can effectively address the perception and assembly processes of plugs under large initial pose deviations, demonstrating robust performance. With the requirement of only a F/T sensor and no need for precise dimensions of the assembly object, this strategy exhibits favourable deployment characteristics for automatic assembly platforms. Furthermore, the proposed strategy can be optimized by the integration of learning-based methods into the perception process, thereby further improving assembly efficiency.}, archive = {J_TASE}, author = {Zhenyu Wang and Xiangfei Li and Huan Zhao and Lingjun Shao and Xiaolong Ma and Huaiwu Zou and Han Ding}, doi = {10.1109/TASE.2024.3477918}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8095--8107}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Geometry and force guided robotic assembly with large initial deviations for electrical connectors}, volume = {22}, year = {2025}, }
• (2025). Object-aware impedance control for Human–Robot collaborative task with online object parameter estimation. TASE, 22, 8081--8094. (www)

  @article{ , abstract = {Physical human-robot interactions (pHRIs) can improve robot autonomy and reduce physical demands on humans. In this paper, we consider a collaborative task with a considerably long object and no prior knowledge of the object’s parameters. An integrated control framework with an online object parameter estimator and a Cartesian object-aware impedance controller is proposed to realize complicated scenarios. During the transportation task, the object parameters are estimated online while a robot and human keep lifting an object. The perturbation motion is incorporated into the null space of the desired trajectory to enhance the estimator precision. An object-aware impedance controller is designed by incorporating the real-time estimation results to effectively transmit the intended human motion to the robot through the object. Experimental demonstrations of collaborative tasks, including object transportation and assembly, are implemented to show the effectiveness of our proposed method. The proposed controller was also compared to a conventional impedance controller through subjective testing and found to be more sensitive, requiring less human effort. Note to Practitioners—This research was motivated by the need to facilitate collaboration between humans and robots in handling heavy or considerable long objects, which can be challenging for a single operator. This paper proposes a physical Human-Robot Interaction (pHRI) approach that enables physical interaction between the human and the robot through the object without additional sensors such as camera or human-machine interfaces. To achieve collaborative task, the separation between the intended human motion and the object dynamics is essential. Most of real-world situations involve uncertain or unknown objects, making it challenging to assume prior knowledge of the target object’s properties. Therefore, this research introduces a real-time approach for estimating the dynamic parameters of unknown objects during collaboration, without requiring additional operational time. Consequently, the design of object-aware impedance controller can be achieved by real-time incorporation of object dynamics. Collaborative transportation and assembly task is demonstrated with whole-body controlled mobile manipulator and 1.5m long object. In future research, we will focus on enhancing the estimation precision through the use of physical informed neural network methods.}, archive = {J_TASE}, author = {Jinseong Park and Yong-Sik Shin and Sanghyun Kim}, doi = {10.1109/TASE.2024.3477471}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8081--8094}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Object-aware impedance control for Human–Robot collaborative task with online object parameter estimation}, volume = {22}, year = {2025}, }
• (2025). Experimental evaluation of haptic shared control for multiple electromagnetic untethered microrobots. TASE, 22, 8069--8080. (www)

  @article{ , abstract = {The precise manipulation of microrobots presents challenges arising from their small size and susceptibility to external disturbances. To address these challenges, we present the experimental evaluation of a haptic shared control teleoperation framework for the locomotion of multiple microrobots, relying on a kinesthetic haptic interface and a custom electromagnetic system. Six combinations of haptic and shared control strategies are evaluated during a safe 3D navigation scenario in a cluttered environment. 18 participants are asked to steer two spherical magnetic microrobots among obstacles to reach a predefined goal, under different conditions. For each condition, participants are provided with different obstacle avoidance and navigation guidance cues. Results show that providing assistance in avoiding obstacles guarantees safer performance, regardless if the assistance is autonomous or delivered through a haptic repulsive force. Moreover, autonomous obstacle avoidance also reduces the completion time by 30% compared to haptic obstacle avoidance and no obstacle avoidance cases, although haptic feedback is preferred by the users. Finally, providing haptic guidance towards the target improves by the 65% the positioning accuracy of the microrobots with respect to not providing this guidance. We also present some illustrative scenarios to generalize the presented haptic shared control strategies to arbitrary formations of N microrobots, while showing the effectiveness of the method for a clinical use-case of endovascular navigation in simulated environment. Note to Practitioners—The recent increasing interest in microrobotics arises from its potential applications in fields like medicine, manufacturing, and environmental monitoring, enabling highly precise control of minimally invasive tools. By enabling users to teleoperate microscale tools with partial autonomous support, these systems facilitate safe access to confined spaces, enhance task efficiency, and enable navigation in otherwise inaccessible environments. Our presented solution serves as an experimental platform to evaluate the efficacy of different combinations of tactile feedback and partial autonomy during safe navigation tasks, with potential applications spanning microsurgery, drug delivery, microscale manufacturing, and environmental remediation. Further practical adaptation of the system will require defining specific application objectives and specifications, along with potential modifications to the actuation system to accommodate environmental constraints of targeted scenarios.}, archive = {J_TASE}, author = {Marco Ferro and Franco N. Piñan Basualdo and Paolo Robuffo Giordano and Sarthak Misra and Claudio Pacchierotti}, doi = {10.1109/TASE.2024.3477308}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8069--8080}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Experimental evaluation of haptic shared control for multiple electromagnetic untethered microrobots}, volume = {22}, year = {2025}, }
• (2025). Capacitated colored traveling salesman problem with time windows. TASE, 22, 8057--8068. (www)

  @article{ , abstract = {This work proposes a variant of Colored Traveling Salesman Problem (CTSP) called Capacitated Colored-traveling-salesman Problem with Time-windows (CCPT), which comes from time-sensitive logistics applications. CCPT is first formulated via mathematical programming formulations, and an Elite-guided Memetic Algorithm (EMA) is developed to tackle it. EMA is able to preserve an active archive of elites during an evolution process. It comprises three schemes, i.e., neighborhood-list-2-opt, relocation move, and cross-arc exchange. They are organized in a variable neighborhood descent framework to optimize a specific high-quality individual. Ablation studies fully show their importance for EMA’s performance. 28 CCPT cases are designed based on representative traveling salesman problem instances to conduct benchmark tests. The statistical comparison shows that EMA is significantly better than Variable Neighborhood Search (VNS), Delaunay-triangulation-based VNS, local search, and memetic algorithm in over 85% of the cases. It achieves faster convergence to the solutions than its competitors.Note to Practitioners—This work is motivated by practical needs for distribution logistics with time-sensitive requirements. A capacitated colored-traveling-salesman problem with time-windows is modeled and can be applied to time-sensitive transportation tasks with multiple goods. An elite-guided memetic algorithm is developed to tackle this problem. It executes the local optimization on a specific high-quality solution during the search process. Extensive comparisons demonstrate that the proposed algorithm can provide decision-makers with significantly better routes than other state-of-the-art algorithms.}, archive = {J_TASE}, author = {Xiangping Xu and Xinli Shi and Jinde Cao and Wei Huang}, doi = {10.1109/TASE.2024.3476696}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8057--8068}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Capacitated colored traveling salesman problem with time windows}, volume = {22}, year = {2025}, }
• (2025). Predicting vulnerable road user behavior with transformer-based gumbel distribution networks. TASE, 22, 8043--8056. (www)

  @article{ , abstract = {This study introduces the crossing intention and trajectory with the Transformer networks (CITraNet) prediction model to process multimodal input data of vulnerable road users (VRUs), such as pedestrians and bicyclists, whose behavior is inherently unpredictable. Unlike traditional approaches that rely on sequence transduction and Gaussian distribution-based models, CITraNet employs Transformer networks and Gumbel distribution. First, we utilize multi-head attention and feed-forward layers to extract hidden features from historically observed data to allow effective parallelization and handling of long-range dependencies. Second, CITraNet features an innovative Transformer-based Gumbel distribution network that significantly enhances the model’s ability to accurately predict all possible trajectories using extreme value theory, which replaces the conventional Gaussian distribution models that struggle with discrete and non-linear data. The effectiveness and accuracy of CITraNet are validated on the Taiwan pedestrian (TaPed) dataset, as well as the publicly available JAAD and PIE datasets. The model’s deterministic and stochastic trajectory predictions are assessed over short (0.5s), medium (1.0s), and long (1.5s) intervals, crucial for gauging predictive accuracy across varying durations. The results demonstrate that CITraNet outperforms previous benchmarks. Note to Practitioners—This study advances the capabilities of advanced driver assistance systems (ADAS) by introducing a vision-based, risk-aware method to predict the future crossing intentions and trajectories of pedestrians and bicyclists. While current ADAS technologies, such as pedestrian collision warning (PCW) systems, excel at detection, they fall short in predicting trajectories and intentions. This limitation can lead to inadequate braking response times and distances, increasing the risk of accidents. Our approach offers timely predictions that enable automatic braking systems to stop vehicles safely, thereby preventing collisions with an early collision warning. This innovation presents significant business opportunities in the development of hardware and software vision components for vehicle manufacturers, automotive component suppliers, and chip designers. However, a notable limitation of our current model is its exclusive focus on pedestrians as traffic agents. Future enhancements will aim to incorporate social interactions among all traffic participants, including vehicles and motorcyclists, to deliver a more comprehensive safety solution.}, archive = {J_TASE}, author = {Lia Astuti and Yu-Chen Lin and Chui-Hong Chiu and Wen-Hui Chen}, doi = {10.1109/TASE.2024.3476382}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8043--8056}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predicting vulnerable road user behavior with transformer-based gumbel distribution networks}, volume = {22}, year = {2025}, }
• (2025). Modulated deformation of 4D printed constructs with environmental adaptation. TASE, 22, 8032--8042. (www)

  @article{ , abstract = {4D printed constructs (4DPC) can transform into different forms and shapes upon exposure to environmental stimuli. However, most 4DPC can only achieve simple deformation under specific incentives currently. There is no effective strategy to enable the 4DPC to adapt to environmental changes autonomously. In this paper, an adaptive fuzzy controller based on expert knowledge is proposed, which will compensate for the disturbance and uncertainty caused by external stimuli to ensure the adaptability and robustness of the 4DPC in dynamic change and uncertainty. To facilitate the design of the controller, a long short-term memory (LSTM) neural network model is proposed to characterize the nonlinear deformation characteristics of the 4DPC. Finally, the environmental adaptability experiment of a self-made 4D printed color-changing structure(4DPCCS) shows that the 4DPCCS integrating perception and control functions can actively adapt to the dynamic changes of the environment. The deformation error of the regulation is $\le \pm 2.39\%$ , and the 4DPCCS can adjust and maintain it adaptively according to changes in the external environment, achieving a camouflage function similar to that of a natural chameleon. Note to Practitioners—The aim of this study is to provide a new solution to meet the dynamic and real-time requirements of real environments for adaptive deformation of 4D printed constructions in dynamic environments. Most of the existing control methods act on the printing process of the constructions to guide the structural design of the 4DPC to achieve specific deformation to specific stimulus. This paper proposes a new method to realize the adaptive modulation of 4D printed constructions to external stimuli through the introduction of adaptive fuzzy control. First, we developed and designed a 4D printed color-changing structure with actuation, sensor and color-changing functions. Secondly, based on a large number of experimental data, we use the modeling method of the LSTM neural network to characterize the relationship between the complex deformation of 4D printed constructions and external stimuli. Finally, an adaptive fuzzy control system for 4D printed color-changing structure is developed and designed based on the neural network model. The experimental results show that this method is feasible. Under the action of the proposed control strategy, the constructions can actively identify the changes of the external environment and adjust and maintain adaptively, realizing the function of color-changing camouflage. In future studies, we will further research the autonomous adjustment method for variable performance or variable functionality of 4D printed constructions.}, archive = {J_TASE}, author = {Zixiang Zhou and Yuyan Zhang and Yintang Wen and Yue Di and Haiying Yao and Weitian Zhang and Xiaoyuan Luo and Hongmiao Tian and Jinyou Shao}, doi = {10.1109/TASE.2024.3476317}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8032--8042}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modulated deformation of 4D printed constructs with environmental adaptation}, volume = {22}, year = {2025}, }
• (2025). Adaptive security secondary control for microgrid in prescribed performance. TASE, 22, 8018--8031. (www)

  @article{ , abstract = {This paper studies the prescribed performance adaptive secondary control problem of the microgrid (MG) under denial-of-service (DoS) attacks. By considering parameter uncertainties and voltage losses, a more practical MG model is established with an uncertain control gain and an unknown nonlinear function. Then, an ingenious fuzzy estimator without the uncertain control gain term is proposed to estimate the unmeasurable state variables and reduce the impact of DoS attacks simultaneously. At the same time, a new estimation error is defined to simplify the derivation process of the estimate error system. To ensure that the voltage fluctuations caused by DoS attacks and load perturbations still satisfy the asymmetric constraints, a novel asymmetric performance function with consideration of the allowable voltage deviation criterion is constructed. In addition, to reduce the controller design complexity, the voltage restoration synchronization error with boundary constraint is transformed into the unconstrained performance error. Further, combining with the designed fuzzy state estimator and the performance transformation error, an adaptive security controller is proposed by using the backstepping technology to ensure that the voltage restoration synchronization error can enter the prescribed asymmetric convergence region within the predefined finite time. Finally, the effectiveness of the designed security controller is verified by two simulations with comparisons. Note to Practitioners—In this paper, a more practical MG system is modeled by considering the factors of the circuit component aging, faults, source voltage disturbances, and voltage losses. Next, the prescribed performance voltage restoration problem under DoS attacks is solved. Since DoS attacks and load perturbations have a great impact on the voltage control, a prescribed performance security control method is proposed to ensure that the transient voltage restoration synchronization error remains strictly within the prescribed region. Additionally, both the convergence time and the asymmetric voltage fluctuation region are predefined by practitioners. This enables practitioners to obtain more specific stabilization time and satisfactory transient performance in the MG. The rationality of the theoretical analysis results is verified by simulation experiments.}, archive = {J_TASE}, author = {Jia-Lei Fu and Wei-Wei Che}, doi = {10.1109/TASE.2024.3476282}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8018--8031}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive security secondary control for microgrid in prescribed performance}, volume = {22}, year = {2025}, }
• (2025). Sliding mode control for delta operator sampled systems. TASE, 22, 8006--8017. (www)

  @article{ , abstract = {This paper addresses a class of sliding mode control (SMC) strategies for high-frequency sampled systems with uncertainties, utilizing a generalized Delta operator reaching law. First, Delta operator sampled systems are regarded as a bridge in the unified research framework of discrete-time systems and continuous-time systems. With the increase of sampling frequency, Delta operator sampled systems can maintain consistent stability with their corresponding continuous models. Then, a generalized reaching law by Delta operator is presented, which can degenerate into multiple existed reaching laws under certain conditions. And then, the reaching condition, quasi-sliding mode (QSM) band, and finite time reachability of the proposed generalized Delta operator reaching law are analyzed. Next, sliding mode controllers are designed separately for both matched uncertainty and unmatched uncertainty cases. With the effect of these controllers, Delta operator SMC systems exhibit desirable dynamic characteristics. Finally, two examples on truck-trailer system and ball-beam system demonstrate the effectiveness of Delta operator sliding mode controllers, and show efficient performance in stabilizing high-frequency sampled systems with uncertainties. These scenarios illustrate the potential of the proposed strategies for practical applications. Note to Practitioners—This work aims to develop sliding mode control strategy for Delta operator sampled systems, which is of fundamental significance in the fields of high-frequency sampled control. We propose a generalized Delta operator reaching law and provide a unified framework for high-frequency sampled discrete-time systems and their corresponding continuous-time systems, addressing practical challenges in areas, such as high-speed motors, supercomputing, and network transmission. The reaching condition, quasi-sliding mode bandwidth, and finite time reachability of the proposed generalized Delta operator reaching law are fully considered. Given that the effects of unknown matched or unmatched uncertainties, sliding mode controllers based on Delta operator are designed, which are applied to truck-trailer system and ball-beam system.}, archive = {J_TASE}, author = {Yunlong Liu and Yanru Zhang and Yonggui Kao and Zairui Gao}, doi = {10.1109/TASE.2024.3476085}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {8006--8017}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sliding mode control for delta operator sampled systems}, volume = {22}, year = {2025}, }
• (2025). Accurate and rapidly-convergent GNSS/INS/LiDAR tightly-coupled integration via invariant EKF based on two-frame group. TASE, 22, 7992--8005. (www)

  @article{ , abstract = {Nowadays, increasing attention has been directed toward the integration of global navigation satellite system (GNSS), inertial navigation satellite system (INS), and light detection and ranging (LiDAR) for intelligent system navigation. However, the existing systems, which generally adopt estimators of the extended Kalman filter (EKF) or factor graph optimization (FGO), still face challenges regarding consistency and convergence. Such methods could provide optimal navigation solutions only if the initial guess of the state is sufficiently close to the true trajectory; otherwise, the systems might undergo accuracy loss or even worse, divergence. To address this issue, we derive an invariant extended Kalman filter (IEKF) based on the two-frame group (TFG) in the left-invariant form, and integrate raw GNSS double-differenced observations, inertial measurements, and LiDAR plane features within this framework. By designing a unified group structure that simultaneously maintains both the navigation states and inertial measurement unit (IMU) biases, TFG contributes to the approximate log-linearity and invariance of the system dynamics model, expected to effectively resolve the convergence issue. A set of real-world experiments was conducted to evaluate the system, with results indicating its potential to achieve submeter to centimeter-level positioning accuracy, surpassing state-of-the-art methods in terms of accuracy, availability, and convergence. Note to Practitioners—This work aims to extend the usability and applicability of GNSS/INS/LiDAR integrated systems on highly maneuverable platforms, such as automobiles and drones. In such cases, the existing approaches based on EKF or FGO face the challenges of degradation or divergence, since their optimality is not guaranteed with erroneous initial state guesses due to severe linearization errors. To tackle these issues induced by model nonlinearities, we propose a tightly-coupled GNSS/INS/LiDAR integration based on the derived TFG-LIEKF, which effectively mitigates the linearization errors by making the dynamics model approximately log-linear via a unique group structure that considers both navigation states and IMU biases. On this basis, the proposed system ensures continuity through multi-sensor fusion, improves accuracy through the tightly-coupled scheme and continuous tracking of LiDAR features, and promotes convergence through the estimator based on TFG.}, archive = {J_TASE}, author = {Chunxi Xia and Xingxing Li and Feiyang He and Shengyu Li and Yuxuan Zhou}, doi = {10.1109/TASE.2024.3476069}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7992--8005}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Accurate and rapidly-convergent GNSS/INS/LiDAR tightly-coupled integration via invariant EKF based on two-frame group}, volume = {22}, year = {2025}, }
• (2025). Rare event detection by acquisition-guided sampling. TASE, 22, 7979--7991. (www)

  @article{ , abstract = {Motivated by the challenges in detecting extremely rare failures for sophisticated specifications in circuit design, we consider the problem of detecting regions of interest (ROIs) that consist of specifications with the value of a complex target function for the system performance being below or above a certain pre-specified threshold. Though Bayesian optimization (BO) has been applied to this problem, it is not effective in identifying multiple ROIs as it was originally designed for global optimization and tends to focus on searching the area where the global optimum is most likely to be. In this work, we propose a sampling strategy for fast ROI detection within a limited number of target function evaluations. The sampling distribution is designed so that the probability of a specification being sampled is proportional to the corresponding value of the acquisition function. Such an acquisition-guided sampling algorithm promotes a wider search of the sample space and a simpler incorporation of different criteria to determine the specifications to be evaluated next. To further improve the performance, we propose a new design of the acquisition function and two modifications of existing acquisition functions. Numerical studies on synthetic functions and a real-world circuit design application demonstrate that the proposed method can enjoy a stronger exploration ability provided by sampling and achieve faster ROI detection with higher coverage. Note to Practitioners—This study considers the extremely rare failure detection in automated circuit design, fabrication, packaging, and verification. Obtaining enough observations of interest within a given budget of evaluations is challenging due to the scarcity of extremely rare failures. Bayesian optimization (BO) has been adopted to tackle this problem, but it may not achieve satisfactory coverage of multiple failure regions, since its goal is to find the global optimum of the target function. In this paper, we propose a sampling-based rare event detection strategy tailored to efficiently detect regions of interest (ROIs) with high coverage, along with newly designed acquisition functions incorporating the pre-specified threshold and ideas of experimental design. This sampling strategy has greater robustness to the choice of acquisition function. Also, since multiple queries can be easily obtained through sampling, various criteria can be easily incorporated for determining the next batch of evaluation specifications without much increase in computational complexity.}, archive = {J_TASE}, author = {Huiling Liao and Xiaoning Qian and Jianhua Z. Huang and Peng Li}, doi = {10.1109/TASE.2024.3475951}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7979--7991}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rare event detection by acquisition-guided sampling}, volume = {22}, year = {2025}, }
• (2025). Approximate inference particle filtering for mobile robot SLAM. TASE, 22, 7967--7978. (www)

  @article{ , abstract = {This paper proposes approximate inference particle filtering for mobile robot simultaneous localization and mapping (SLAM) with landmarks. Range-bearing measurements are obtained by detecting landmarks using an onboard laser range finder and the maximum likelihood approach is used to handle unknown data associations. The system model is created depending on the robot motion and range-bearing measurements. The new particle filter is developed to estimate the robot pose and landmark locations separately based on approximate inferences, where the intractable distributions of the robot pose and landmark locations are approximated by optimal Gaussian distributions minimizing Kullback-Leibler divergences. Simulations and experiments are provided to show the performance of the proposed particle filter. Note to Practitioners—The SLAM technique is crucial to an autonomous mobile robot. The laser range finder perceives surroundings and is used for mobile robot SLAM in global positioning system denied environments. This paper proposes a mobile robot SLAM method based on the robot motion model and the onboard laser range finder. Range-bearing measurements are extracted from raw data of the laser range finder as the inputs of the method. The simulations and experiments indicate the proposed method has better accuracy and more robust to unknown data associations. The proposed method can be applied to many practical applications such as transport operations and mowing with a mobile robot. In the future, we will address the SLAM technique for multiple mobile robots.}, archive = {J_TASE}, author = {Shuo Zhang and Jinjun Shan and Yibo Liu}, doi = {10.1109/TASE.2024.3475735}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7967--7978}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Approximate inference particle filtering for mobile robot SLAM}, volume = {22}, year = {2025}, }
• (2025). H∞ security control for singular jump systems with actuator faults and DoS attacks under switching-like adaptive trigger and application to OCCP. TASE, 22, 7955--7966. (www)

  @article{ , abstract = {This research is concerned with $H_{\infty } $ security feedback control for networked singular Markovian jump systems (NSMJSs) with actuator faults and denial of service (DoS) attacks under mode-dependent switching-like adaptive event-triggered scheme. To handle energy-limited denial of service attacks and further strengthen communication efficiency, a mode-dependent switching-like adaptive event-triggered scheme (MSAETS) adopting state-dependent adaptive threshold function is introduced. By exploiting singular value decomposition (SVD) technique and establishing a mode-related discontinuous Lyapunov-Krasovskii (L-K) functional, novel $H_{\infty } $ stochastic admissibility criteria for NSMJSs with DoS attacks and actuator faults are acquired, and the relationship between event-triggered parameters and DoS attacks energy is developed through the clever use of matrices/vectors norms and iterative method. Collaborative design method of event-triggered scheme and security feedback controller is proposed, which can make sure the $H_{\infty } $ stochastic admissibility of NSMJSs while alleviating the communication burden and improving communication efficiency. The validity of the suggested cooperative control strategy is demonstrated via using an oil catalytic cracking process. Note to Practitioners—In actual engineering, numerous dynamic systems are usually modeled as singular systems, such as oil catalytic cracking processes, power systems, etc. Noteworthy, the investigation of singular systems is considerably complicated and more challenging than normal ones since numerous factors have to be contemplated in parallel, including stability, regularity, non-impulsiveness or causality. Most existing control results are poorly suited to these types of practical systems. Meanwhile, practical systems inevitably endure sudden changes in system structure or/and parameters, actuators may fail completely or partially during operation, and open communication environment and limited bandwidth of communication networks pose a range of challenges including packet losses, network-induced transmission delays and malicious network attacks, all of which may lead to reducing system performance, compromising system stability and eventually result in catastrophic accidents. To address the above challenges, in this work, the mode-dependent switching-like adaptive event-triggered $H_{\infty } $ security feedback control scheme is designed for NSMJSs with DoS attacks and actuator faults.}, archive = {J_TASE}, author = {Yujing Pang and Guangming Zhuang and Jianwei Xia and Xiang-Peng Xie}, doi = {10.1109/TASE.2024.3475374}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7955--7966}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {H∞ security control for singular jump systems with actuator faults and DoS attacks under switching-like adaptive trigger and application to OCCP}, volume = {22}, year = {2025}, }
• (2025). Memory-based event-triggered fault-tolerant consensus control of nonlinear multi-agent systems and its applications. TASE, 22, 7941--7954. (www)

  @article{ , abstract = {This article is concerned with the memory-based event-triggered leader-following dissipative fault-tolerant consensus (LFDFTC) problem for the nonlinear multi-agent systems (NMASs) with semi-Markov switching topologies subject to the generally uncertain semi-Markov (GUSM) jumping process. Unlike the existing event-triggered (ET) consensus results, the dynamic memory event-triggered mechanism (DMETM) and memory-based distributed fault-tolerant (FT) controllers are designed to reduce the ET times. By constructing a general mode-dependent Lyapunov-Krasovskii functional (LKF) and strictly $(\bf {\mathcal {R,Q,T}})-\boldsymbol {\gamma }$ dissipative analysis, the dissipative FT consensus conditions of NMASs are derived in this paper. Finally, three actual physical systems are utilized to verify the validity of the proposed method. Note to Practitioners—Owing to the complexity of engineering environment, the consensus control issue of NMASs has attracted widespread attention. Nowadays, the consensus control of NMASs is generally utilized in diverse fields, such as multi-vehicle coordination, smart grids, and unmanned aerial vehicle formation. However, for the electronic device in practical applications, the channel bandwidth is limited due to power and energy constraints, and it is difficult for the fixed communication topologies and traditional periodic sampled-data control method to cope with these unexpected situations. Therefore, the LFDFTC issue for the NMASs with GUSM switching topologies is investigated by adopting DMETM and memory-based distributed FT controllers in this paper. In addition, the proposed FT consensus control methods with prescribed dissipative performance are applied to multiple vehicles time-invariant formation, Chua’s circuits synchronization, and multiple manipulators consensus.}, archive = {J_TASE}, author = {Junyi Wang and Ying Zheng and Jinliang Ding and Huaguang Zhang and Jiayue Sun}, doi = {10.1109/TASE.2024.3474968}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7941--7954}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Memory-based event-triggered fault-tolerant consensus control of nonlinear multi-agent systems and its applications}, volume = {22}, year = {2025}, }
• (2025). Fixture-free 2D sewing using a dual-arm manipulator system. TASE, 22, 7927--7940. (www)

  @article{ , abstract = {This paper proposes a fixture-free 2D sewing system using a dual-arm manipulator, i.e., the seam lines of the top and bottom fabric parts are the same. The proposed 2D sewing system sews two stacked fabric parts together along a desired seam line printed on the top fabric part without the use of a fixture. In the proposed system, the set of aligned and stacked fabric parts is held by the end-effectors of the dual-arm manipulator in coordination. The dual-arm manipulator controls the motion of the fabric parts on the flat sewing table stitch by stitch in coordination, while keeping the manipulated fabric parts flat using the internal force applied to the set of fabric parts. A novel vision-based seam line tracking control is proposed to control the motion of the set of fabric parts along the printed seam line on the top fabric part. The convergence of the tracking error is analyzed for sewing along both straight and curved seam lines and is shown to be specified by the control parameters. Sewing experiments show that the tracking error converges to zero as analyzed. The sewing experiments also show that the newly proposed trajectory generation method, which synchronizes the coordinated motion of the manipulators and the motion of the sewing needle, is essential for achieving accurate sewing.Note to Practitioners—Most semi-automatic sewing machines and pattern sewers on the market use fixtures to handle the stacked fabric parts. They require the user to customize the fixture depending on the shape, size, and material of the fabric parts to be sewn together. Users are required to redesign/reconfigure the fixture to sew different fabric parts. Our robotic sewing system is based on the concept of fixture-free sewing, i.e., the pose of the set of stacked fabric parts is controlled by the end-effectors without using the fixture. The internal force applied to the fabric parts by the end-effectors is used to keep the fabric parts flat, and the position of the fabric parts is controlled by the motions of the end-effectors in coordination with the proposed vision-based seam line tracking control. The proposed robotic sewing system provides practitioners with a new approach to fixture-free automatic sewing of fabric parts.}, archive = {J_TASE}, author = {Fuyuki Tokuda and Ryo Murakami and Akira Seino and Akinari Kobayashi and Mitsuhiro Hayashibe and Kazuhiro Kosuge}, doi = {10.1109/TASE.2024.3474914}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7927--7940}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixture-free 2D sewing using a dual-arm manipulator system}, volume = {22}, year = {2025}, }
• (2025). Toward futuristic autonomous Experimentation—A surprise-reacting sequential experiment policy. TASE, 22, 7912--7926. (www)

  @article{ , abstract = {An autonomous experimentation platform in manufacturing is supposedly capable of conducting a sequential search for finding suitable manufacturing conditions by itself or even for discovering new materials with minimal human intervention. The core of the intelligent control of such platforms is a policy to decide where to conduct the next experiment based on what has been done thus far. Such policy inevitably trades off between exploitation and exploration. Currently, the prevailing approach is to use various acquisition functions in the Bayesian optimization framework. We discuss whether it is beneficial to trade off exploitation versus exploration by measuring the element and degree of surprise associated with the immediate past observation. We devise a surprise-reacting policy using two existing surprise metrics, known as the Shannon surprise and Bayesian surprise. Our analysis shows that the surprise-reacting policy appears to be better suited for quickly characterizing the overall landscape of a response surface under resource constraints. We do not claim that we have a fully autonomous experimentation system but believe that the surprise-reacting capability benefits the automation of sequential decisions in autonomous experimentation.Note to Practitioners Autonomous systems should be able to go beyond repetitive automatic actions that are generally pre-programmed through a recipe. To decide what to do next on the fly differentiates autonomy from automation. Arguably, autonomy is the highest form of automation. To endow a manufacturing with autonomy, one necessary capability is for it to react properly to the “unexpected,” which are those observations disagreeing with its model’s anticipation. Are these bad measurements, an anomaly, or an indicator of model inadequacy? Should the observations be discarded or should the model be updated using the new observation? If latter, should model be updated gradually overtime or radically altered? Figuratively, upon observing the unexpected, we say that a manufacturing control system is “surprised” and ask the question of how it should react. Our investigation shares our current insights on this question.}, archive = {J_TASE}, author = {Imtiaz Ahmed and Satish T. S. Bukkapatnam and Bhaskar Botcha and Yu Ding}, doi = {10.1109/TASE.2024.3474549}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7912--7926}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward futuristic autonomous Experimentation—A surprise-reacting sequential experiment policy}, volume = {22}, year = {2025}, }
• (2025). Self-starting monitoring and dynamic sampling of high-dimensional data streams. TASE, 22, 7897--7911. (www)

  @article{ , abstract = {In today’s manufacturing industries, the development of sensor technology and Internet of Things has made real-time process monitoring of high-dimensional data increasingly vital. However, resource constraints, such as limited power, budget, and transmission capacity, often prevent access to full data streams in real time. This means that practitioners need to effectively monitor the process based on only partially observed data by dynamically deciding the sampling layout in real time. Another common challenge of process monitoring in practice is the lack of historical reference data, which can occur due to process/system upgrades or equipment replacements. To address these critical challenges, this paper proposes MASS (Monitoring with Adaptive Sampling under Self-starting scheme), a novel self-starting monitoring approach tailored to monitor high-dimensional data streams when only limited resources and historical reference data are available. Our monitoring framework is based on a quantile-based nonparametric CUSUM procedure with likelihood ratio-based statistics, and then the Thompson Sampling (TS) algorithm is adopted to handle partially observed data in the self-starting scenario. A key feature of our proposed method is its adaptive estimation of the out-of-control distribution and data quantiles, which ensures robust detection for various shifts in data streams with arbitrary and heterogeneous distributions, even in cases with limited reference data. The outperformance of the proposed method is demonstrated through simulation experiments and a real-world case study. Note to Practitioners—This paper is motivated by the critical challenges of online process monitoring when only limited resources (e.g., limited power availability, limited number of sensors, and limited transmission capacity) and limited historical in-control reference data are available. For example, consider a scenario where a newly established production system requires online monitoring across multiple data streams. In such cases, there is often a deficiency of reference data crucial for constructing a reliable control chart. Additionally, due to resource constraints, it is frequently infeasible to gather information from all streams associated with the production line at each epoch in real time. Unlike previous methods which require either a sufficient amount of reference data, or fully observable data streams, this paper proposes a novel monitoring and dynamic sampling scheme to effectively monitor partially observable data streams with only a small amount of reference data. To implement the methodology, it requires: (i) to initiate the process by estimating quantiles with a small amount of reference data, (ii) to determine which data streams to observe at each time epoch, (iii) to adaptively update the estimation of process parameters during online monitoring, and (iv) to construct a set of local and global statistics that can be used to quickly detect the system anomaly in real time. Numerical experiments and a real-world case study suggest that our proposed method efficiently leverages available data to reduce detection delays and enhance effectiveness against various shifts, in comparison to the benchmark methods.}, archive = {J_TASE}, author = {Jiahui Zhang and Ziqian Zheng and Jun Li and Kaibo Liu}, doi = {10.1109/TASE.2024.3474296}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7897--7911}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-starting monitoring and dynamic sampling of high-dimensional data streams}, volume = {22}, year = {2025}, }
• (2025). Event-based finite time stabilizability and formation control of multi-agent systems. TASE, 22, 7887--7896. (www)

  @article{ , abstract = {The paper considers the event-triggered stabilizability of multi-agent systems (MAS). To reduce the frequency of control input update and information transmission, a novel distributed event-triggered control strategy with state estimation feedback is designed to achieve stabilizability. Event-trigger rules involving dynamic threshold are constructed, which ensure the convergence of systems states in finite time. By utilizing iSCC (independent strongly connected component) partition and Lyapunov stability theory, sufficient criteria for achieving finite time stabilizability have been obtained. Moreover, formation control under event-trigger scheme is addressed based on the obtained stabilizability results. Besides, it has been proven that the event-trigger interval has a positive lower bound, which can avoid Zeno behavior. Finally, the effectiveness of the theoretical results is verified by simulation.Note to Practitioners—In the fields of control and engineering, system stability has always been a research hotspot. Traditional stability analysis often focuses on the asymptotic stability. However, many practical scenarios require the system to reach a stable state within a finite time, i.e., finite time stabilizability, such as rapid response systems, emergency braking systems, and responding to emergencies. Hence, the research on finite time stabilizability has important practical value. In order to reduce the frequency of control updates, this paper proposes the finite time stabilizability under event-trigger scheme and designs a novel feedback controller based on state estimation, which has not been discussed before. At the same time, to explore the impact of network topology on stabilizability, the sufficient conditions for achieving finite time stabilizability of the system are obtained by iSCC graph partition. Dynamic threshold in event-triggering conditions to guarantee the implementation of finite time stabilizability is constructed. Finally, the theoretical results are applied to formation control. The research is of great significance for improving system performance and meeting real-time requirements.}, archive = {J_TASE}, author = {Yinshuang Sun and Zhijian Ji and Yang Shi and Yungang Liu}, doi = {10.1109/TASE.2024.3473972}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7887--7896}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based finite time stabilizability and formation control of multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Energy efficiency improvement of hydraulic manipulator through flow-optimal redundancy resolution with path following. TASE, 22, 7870--7886. (www)

  @article{ , abstract = {For hydraulic manipulators, utilizing the redundant degrees of freedom to optimize the allocation of pump flow is beneficial for the reduction of energy consumption. However, a higher priority task should be considered to follow the reference path under the constraints of actuators and pump. A hierarchical redundancy resolution method is proposed in this study for real-time flow optimization of redundant hydraulic manipulator, which takes into account the path following task under constraints of flow, position, velocity, and acceleration. A constraint reconstruction method is proposed to determine the hard constraints while eliminating redundant ones, such that the computational efficiency is improved. The tasks of path following and flow optimization are addressed hierarchically. Moreover, a feasible region partitioned solution algorithm is designed to simplify the flow optimization problem with a non-smooth cost function into linear programming sub-problems. Four sets of comparative tests under different paths and speeds are carried out on a redundant hydraulic manipulator. The results show that the proposed method enables the hydraulic manipulator to execute tasks accurately and efficiently in real-time. Note to Practitioners—This work was motivated by the need to develop an efficient computational method for reducing the energy consumption of redundant hydraulic manipulators while following a reference path under the constraints of the actuators and pump. The objective was to operate the manipulator effectively and accurately in real-time. In this article, a hierarchical redundancy resolution method is proposed to reduce energy consumption by optimizing pump flow allocation, while considering the path following task and hard constraints. It is different from the off-line methods that pre-planning the trajectories before execution in a structured environment. A constraint reconstruction method and a feasible region partitioned solution algorithm are designed to accelerate the solution procedure. The proposed method enables the hydraulic manipulator to execute the given reference path accurately and energy efficiently in real-time. Energy consumption optimization considering dynamic constraint boundaries and hydraulic system pressure needs to be further investigated.}, archive = {J_TASE}, author = {Linan Li and Min Cheng and Ruqi Ding and Bing Xu}, doi = {10.1109/TASE.2024.3472671}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7870--7886}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy efficiency improvement of hydraulic manipulator through flow-optimal redundancy resolution with path following}, volume = {22}, year = {2025}, }
• (2025). Solving stochastic orienteering problems with chance constraints using monte carlo tree search. TASE, 22, 7855--7869. (www)

  @article{ , abstract = {We present a new Monte Carlo Tree Search (MCTS) algorithm to solve the stochastic orienteering problem with chance constraints, i.e., a version of the problem where travel costs are random, and one is assigned a bound on the tolerable probability of exceeding the budget. The algorithm we present is online and anytime, i.e., it alternates planning and execution, and the quality of the solution it produces increases as the allowed computational time increases. Differently from most former MCTS algorithms, for each action available in a state the algorithm maintains estimates of both its value and the probability that its execution will eventually result in a violation of the chance constraint. Then, at action selection time, our proposed solution prunes away trajectories that are estimated to violate the failure probability. Extensive simulation results show that this approach can quickly produce high-quality solutions and is competitive with the optimal but time-consuming solution. Note to Practitioners—In many practical scenarios one is faced with multiobjective sequential decision making problems that can be solved through constrained optimization. If some of the parameters are known with uncertainty, the event “violating one of the constraints” becomes a random variable whose probability should be bound. As an application of this general problem formulation, in this paper we consider stochastic orienteering, a problem that finds applications when a robot is tasked with performing multiple tasks of varying utility while being subject to a bound on the traveled distance. Many problems in logistics, precision agriculture, and environmental monitoring, just to name a few, can be cast as instances of this optimization problem.}, archive = {J_TASE}, author = {Stefano Carpin}, doi = {10.1109/TASE.2024.3472453}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7855--7869}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Solving stochastic orienteering problems with chance constraints using monte carlo tree search}, volume = {22}, year = {2025}, }
• (2025). Cross-modality manifold adaptive network for industrial multimode processes and its applications. TASE, 22, 7845--7854. (www)

  @article{ , abstract = {In actual industrial scenarios, different operating modes and workloads can lead to multiple modes of working conditions, resulting in significantly diverse feature spaces. However, the heterogeneity and complexity among these modes pose a challenge to traditional data processing methods. Therefore, this paper proposes the cross-modality manifold adaptive Network (CMAN) to facilitate cross-modal information transmission for addressing multi-modal prediction issues. Specifically, CMAN divides the prediction process into two steps. Firstly, the manifold discriminative autoencoder (MDAE) is proposed to extract both local and global manifold geometric structures. The loss function of the designed MDAE in mode recognition is formulated to minimize the ratio between within-modal and between-modal features. In this way, the autoencoder not only learns data representations but also learns to differentiate between data from different classes. This lays the foundation for determining fusion strategies between modes in subsequent steps. Secondly, in the process of multimode prediction, to assist the model in learning and understanding the mutual influences and dependencies between different modes, CMAN shares features between modes through cross connections. It can adaptively preserve task specificity while also utilizing between-task correlations. The effectiveness of the proposed method is validated in the Tennessee Eastman (TE) case and an actual power plant case. Note to Practitioners—The use of soft sensors to monitor key variables of multimode processes is essential for optimizing and controlling chemical processes. However, it is difficult for conventional methods to accurately and comprehensively utilize within- and between-modal information of multimode processes to build robust and powerful soft sensors. In addition, it is difficult to obtain mode-indicating variables in real-world processes. To address these issues, CMAN is proposed in this paper. Firstly, the historical data of each mode in a multimode industrial process are collected, and the CMAN utilizes the manifold discrimination idea to build a mode recognition model. Then, when modeling the specific modes, CMAN utilizes cross-connections to migrate knowledge between modes, which not only considers the information of the modes themselves, but also makes the features between modes cross-transferred. The gating mechanism enables adaptive optimal combination between various types of features. Finally, two sets of cases show that the proposed method has excellent prediction performance.}, archive = {J_TASE}, author = {Xiao-Lu Song and Ning Zhang and Yan-Lin He and Yuan Xu and Qun-Xiong Zhu}, doi = {10.1109/TASE.2024.3472051}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7845--7854}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cross-modality manifold adaptive network for industrial multimode processes and its applications}, volume = {22}, year = {2025}, }
• (2025). Multi-fault diagnosis for series-connected lithium-ion battery packs based on improved sensor topology and correlation coefficient method. TASE, 22, 7833--7844. (www)

  @article{ , abstract = {In this paper, the multi-fault diagnosis problem is investigated for series-connected lithium-ion battery packs based on an improved correlation coefficient method. Different from existing correlation-based fault diagnosis methods having difficulties in distinguishing between sensor faults and connection faults, a novel sensor topology is proposed to separate sensor faults from connection faults. Furthermore, to solve the problem of inconsistent correlation coefficients in different working states, an adaptive threshold mechanism is proposed. By combining the improved sensor topology and adaptive threshold mechanism, multi-fault detection and diagnosis can be realized. Subsequently, through different characteristics of the correlation coefficient and voltages, the types of sensor faults, such as the bias fault, scaling fault, drift fault, sticking state and noise, can be determined. Experiments prove the effectiveness and superiority of the proposed scheme. Note to Practitioners—This paper focuses on the multi-fault diagnosis of lithium-ion battery packs, aiming to enhance system robustness and longevity. The string-level redundancy topology structure greatly simplifies the fault diagnosis process without increasing the hardware cost and system complexity, which greatly reduces the calculation amount. With the adaptive threshold mechanism, the consistency of correlation coefficients in different working states is improved, so the accuracy of fault diagnosis and the robustness to different working states are increased. This paper uses four batteries in series to form a battery pack. In theory, the proposed method is equally effective when the number of cells in the pack rises. Thus, our multi-fault diagnosis method has great potential to be applied to electric vehicles and energy storage systems and adeptly addresses complex and dynamic operational environments. Our research has been validated in practical battery packs, yielding significant successes. By effectively identifying and addressing various faults, we have improved the overall performance and reliability of battery packs.}, archive = {J_TASE}, author = {Yue Cao and Engang Tian and Hui Chen and Huwei Chen}, doi = {10.1109/TASE.2024.3471253}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7833--7844}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-fault diagnosis for series-connected lithium-ion battery packs based on improved sensor topology and correlation coefficient method}, volume = {22}, year = {2025}, }
• (2025). Asynchronous h∞ control of switched systems and its applications: An admissible chain-dependent average dwell time switching strategy. TASE, 22, 7821--7832. (www)

  @article{ , abstract = {This work is interested in the stability analysis and ${\mathcal {H}}_{\infty }$ control for a class of continuous-time switched systems with external disturbances and asynchronous switching. Based on the attributes of the closed-loop switched systems with edge-dependent state-feedback controllers, a novel admissible chain-dependent average dwell time switching strategy is proposed. The new switching is characterized by admissible transition chains derived from combining two admissible transition edges associated with two consecutive switching events, and it is expected to relax the dwell time constraints in the existing admissible edge-dependent average dwell time switching. Moreover, to reflect the nature of the discussed asynchronously switched systems, a hybrid chain- and edge-dependent Lyapunov function is introduced. Through the suggested switching and Lyapunov function methods, the stability and $\mathcal {L}_{2}$ -gain analysis results are obtained, based on which the intended edge-dependent ${\mathcal {H}}_{\infty }$ control scheme is designed. Finally, the provided control scheme is validated through switched DC motor and Boost models as well as a numerical example. Note to Practitioners—Switched systems are powerful tools for modeling engineering systems with switching events. In practical switched applications, recognizing subsystem modes and applying matched controllers can take some time, which leads to the asynchronous phenomenon. Furthermore, external disturbances are unavoidable in real systems. The more relaxed dwell time constraints of switching signals mean better applicability of the obtained results, especially for switched systems subject to asynchronous switching and external disturbances. The disturbance attenuation performance of an asynchronously switched system can be improved by reducing the running time of the subsystem performing poorly on this performance. Therefore, it is meaningful to devise a control scheme with relaxed dwell time constraints, which can be resolved by our proposed edge-dependent control method with admissible chain-dependent average dwell time.}, archive = {J_TASE}, author = {Qian Shen and Shengyuan Xu}, doi = {10.1109/TASE.2024.3470816}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7821--7832}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous h∞ control of switched systems and its applications: An admissible chain-dependent average dwell time switching strategy}, volume = {22}, year = {2025}, }
• (2025). Deep reinforcement learning-based collision-free navigation for magnetic helical microrobots in dynamic environments. TASE, 22, 7810--7820. (www)

  @article{ , abstract = {Magnetic helical microrobots have great potential in biomedical applications due to their ability to access confined and enclosed environments via remote manipulation by magnetic fields. However, achieving collision-free navigation for microrobots in complex and unstructured environments, particularly in highly dynamic settings, remains a challenge. In this paper, we present a novel deep reinforcement learning-based control framework for magnetic helical microrobots, focusing on the tasks of goal-reaching and dynamic obstacle avoidance. To streamline data collection, a specialized training environment capturing essential aspects of navigation for magnetic helical microrobots is devised. The robustness and adaptability of the trained policy are supported using a randomization technique within the training environment. To facilitate seamless integration with real-world magnetic actuation systems, a visual processing algorithm based on OpenCV is devised and incorporated to collect policy observations. Simulations and experiments in various scenarios validate the high robustness and adaptability of the method. The performance assessment revealed a success rate of 99% in navigating the microrobot around 4 dynamic obstacles of comparable speeds and a success rate of 90% in environments with 14 dynamic obstacles. The results indicate the potential for future applications of our method in unstructured, confined, and dynamic living environments.Note to Practitioners—The motivation of this work is to develop a robust and effective control scheme for collision-free navigation of magnetic helical microrobots in dynamic environments. The conventional navigation strategies in dynamic environments mainly include global path planning and local path replanning; thus, highly dynamic environments require frequent updates to the planned path, making it difficult to apply in highly dynamic environments. In this work, a deep reinforcement learning-based control framework is proposed that can guide microrobots through many dynamic obstacles to a series of locations without collisions. The simulation and experimental results validate the efficacy of the proposed control framework and the robustness and adaptability of the trained policy. The proposed control scheme enables better understanding of advanced motion control methods for magnetic microrobots.}, archive = {J_TASE}, author = {Huaping Wang and Yukang Qiu and Yaozhen Hou and Qing Shi and Hen-Wei Huang and Qiang Huang and Toshio Fukuda}, doi = {10.1109/TASE.2024.3470810}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7810--7820}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep reinforcement learning-based collision-free navigation for magnetic helical microrobots in dynamic environments}, volume = {22}, year = {2025}, }
• (2025). A novel contouring control method based on optimal vector-referenced moving frame for 3D trajectory with zero curvature. TASE, 22, 7798--7809. (www)

  @article{ , abstract = {Contouring control of 3D trajectory is critical in multi-axial machine tool, scanning stage and other precision automation systems. Currently, most contouring controllers are based on Frenet frames, thus their limited applicability to nonzero-curvature 3D trajectories rather than zero-curvature ones which exist ubiquitously in multi-axial motion systems. This paper proposes an optimal vector-referenced moving frame based contouring controller (OVRMFCC) suitable for contouring control of 3D zero-curvature trajectories. Firstly, the optimal vector-referenced moving frame (OVRMF) capable of framing arbitrary finite-length smooth 3D trajectory regardless of its curvature was proposed. Then, a contouring controller (OVRMFCC) based on OVRMF was designed, followed by derivation of its analytical form and proof of its convergence. Finally, this controller was deployed to an FPGA-based controller target with comprehensive comparison experiments on a triaxial system. Experimental results indicates that OVRMFCC reduces at least 46.6% maximum contour error, and 25.0% root-mean-square contour error compared to cross-coupled controller. Besides, OVRMFCC achieves almost the same precision on trajectories with nonzero curvature or curvature singularities compared to TCF. It still maintains high-precision contour tracking on trajectories with continuous zero-curvature segments or planned discrete trajectories with sharp curvature changes, while TCF crashes or leads to several times larger contour error. Note to Practitioners—This work is motivated by the increasing need of contouring control of 3D trajectory in precision automation systems. The mainstream 3D contouring controllers, like task coordinated frame method and model predicted contouring controller, are invalid for zero-curvature 3D trajectory which ubiquitously exists in motion system since they are based on Frenet frame which fails to be defined where curvature is zero. Although cross-coupled controller can tackle zero-curvature 3D trajectory, it proves inefficient in reducing contour error as it is commonly model-free and not specifically designed. To solve this problem, we proposed an optimal vector referenced moving frame (OVRMF) for arbitrary infinite-length smooth 3D trajectory framing and proves its existence strictly. The main advantage of OVRMF lies in its independence of curvature, thus its existence everywhere as long as the curve is $C^{1}$ continuous. And based on OVRMF a contouring controller (OVRMFCC) is designed which combines the benefit of both TCF and OVRMF. With this, OVRMFCC can track any infinite-length $C^{3}$ -continuous 3D trajectory, which fill the gap of traditional 3D contouring controller. Experimental results reveal that OVRMFCC maintains high precision regardless of trajectory curvature, which outperforms both cross-coupled controller and TCF.}, archive = {J_TASE}, author = {Qianjun Zhang and Yongzhuo Gao and Siyuan Meng and Hui Dong and Changhai Ru and Wei Dong}, doi = {10.1109/TASE.2024.3470799}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7798--7809}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel contouring control method based on optimal vector-referenced moving frame for 3D trajectory with zero curvature}, volume = {22}, year = {2025}, }
• (2025). Output feedback SOSM control of constrained systems with unmatched uncertainties. TASE, 22, 7786--7797. (www)

  @article{ , abstract = {In this work, we devise an output feedback second-order sliding mode (SOSM) control protocol for a category of nonlinear systems with unmatched uncertainties and asymmetric output constraints. The procedure of the design and analysis for this control tactic primarily encompasses the following steps. To begin with, a nonlinear mapping is introduced to convert the output-constrained SOSM system into a new SOSM dynamics without output constraints. Next, through adopting a scaling transformation with properly scheduled gains, a SOSM controller via output feedback is explicitly constructed. Third, it is theoretically certified that the presented control strategy insures both the finite-time convergence of the whole system and the compliance with the preset asymmetric output constraint. Eventually, the simulation studies on a practical application example confirm the feasibility and benefits of the proposed algorithm. Distinct from the prevailing outcomes in the literature, our control frame reveals two compelling features. One is that the output constraint issue is tackled via the system conversion based control. The other is that the proposed output feedback control design is independent of the separation principle, i.e., the finite-time convergence for the total closed-loop SOSM system embracing a state feedback SOSM controller and a discontinuous observer is together analyzed. Note to Practitioners—The paper considers a class of asymmetric output-constrained nonlinear systems subject to unmatched uncertainties, whose models are pervasively adopted in the engineering fields, e.g., manufacturing sector, transportation, etc. For such systems, the output feedback SOSM control issue hitherto has not been tackled, since under the circumstance that merely the output information is obtainable, the unmatched uncertainties and the output constraints are simultaneously considered. To this end, we successfully establish a novel output feedback SOSM control scheme by introducing the nonlinear mapping, the scaling transformation and the backstepping-like technique. Moreover, the finite-time convergence of the entire system is analyzed and corroborated via the Lyapunov method, and the fulfillment of the pre-specified asymmetric output constraint is assured, which make the proposed strategy more engineering-oriented. In future work, the efforts will be directed towards the validation of the proposed approach on an experimental platform for electric vehicles.}, archive = {J_TASE}, author = {Keqi Mei and Jingkai Liu and Shihong Ding}, doi = {10.1109/TASE.2024.3470779}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7786--7797}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Output feedback SOSM control of constrained systems with unmatched uncertainties}, volume = {22}, year = {2025}, }
• (2025). Simultaneously learning of motion, stiffness, and force from human demonstration based on riemannian DMP and QP optimization. TASE, 22, 7773--7785. (www)

  @article{ , abstract = {In this paper, we propose a motion, stiffness, and force learning framework based on an extended dynamic movement primitive (DMP) and quadratic programming (QP) optimization. The objective is to learn kinematic and dynamic operational parameters from a one-shot human demonstration, through measurement and estimation of the motion, 3-dimensional (3-D) endpoint stiffness, and applied forces of the human arm during manipulation tasks. To this end, first, the framework features an extended DMP to model the motion, stiffness, and force variations in Cartesian space and 2-D sphere manifold. Second, to account for collected errors and human-robot operation gaps, a QP optimization is applied to fine-tune the desired position of the controller. Finally, we validate the framework through two experiments in real scenarios on the Franka Emika Panda robot. Experimental results show that the robot can not only inherit the variation laws of motion, stiffness, and force in the human demonstration, but also exhibit certain generalization capabilities to other situations. The framework provides a reference for robots learning multiple skills via a one-shot human demonstration, which finds great potential application in human-robot cooperation, contact-rich scenarios, and skillful operations, where the motion, stiffness, and applied forces need to be considered simultaneously. Note to Practitioners—Fast programming in robotics through skill transfer plays a critical role in next-generation robots entering ordinary people’s lives. Existing research focuses more on skill learning at the kinematic level and lacks on the dynamic level, such as stiffness and contact force. The goal of this paper is to propose a novel framework for robots learning of motion, stiffness, and force variations from a one-shot human demonstration, simultaneously. To this end, a Riemannian-based DMP method is employed to model the variation laws of motion, stiffness, and force in Cartesian space and 2-D sphere manifold, respectively. In this way, the learning module needs to be run only once, and the patterns can also be generalized to other targets without repeated robot teaching and additional time-consuming processes. To accurately reproduce the learned skills, a human-like motion/stiffness/force controller combined with QP optimization is investigated. In this paper, rather than identifying real environmental parameters, we directly use interacted forces during the human demonstration to represent environmental effects and employ QP to update the desired position in a limited range to account for collected errors and human-robot operation gaps. Experiments on button pressing and polishing tasks by the Panda robot have achieved very good results. The work of this paper lays a foundation for multiple skills learning from human demonstration (LfHD).}, archive = {J_TASE}, author = {Zhiwei Liao and Francesco Tassi and Chenwei Gong and Mattia Leonori and Fei Zhao and Gedong Jiang and Arash Ajoudani}, doi = {10.1109/TASE.2024.3469961}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7773--7785}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Simultaneously learning of motion, stiffness, and force from human demonstration based on riemannian DMP and QP optimization}, volume = {22}, year = {2025}, }
• (2025). Data-driven containment control for unknown MIMO nonlinear MASs under aperiodic DoS attacks. TASE, 22, 7762--7772. (www)

  @article{ , abstract = {This paper considers the problem of containment control for multi-input multi-output (MIMO) nonlinear multi-agent systems (MASs) with output saturation and aperiodic denial-of-service (DoS) attacks. By applying the dynamic linearization method, the nonlinear systems are modified to an equivalent linear data model. Based on the inputs and saturated outputs data, a new data-driven containment control algorithm in the presence of aperiodic DoS attacks is developed, which mainly includes the following perspectives: 1) based on the historical information of the MASs, a switched error system is designed to decrease the impact of aperiodic DoS attacks, where the time duration was constrained due to the limitation of attackers’ energy; 2) the relation between measured error and tracking error is constructed to deal with the difficulties of incomplete data suffered from output saturation; 3) with the help of contraction mapping method and mathematical induction approach, the boundedness of containment error is guaranteed by the convergence of the control strategy in such an insecure environment. Numerical simulations are given to validate the effectiveness of the framework. Note to Practitioners—This paper studies the containment control problem for MIMO nonlinear MASs, which has application value in robotic systems, intelligent transportation, multiple spacecraft systems, and other fields. Note that these systems are notably susceptible to network attacks and saturation nonlinearity, which may decrease the system performance or even cause instability. Since the agent dynamics generally are not accurate even unavailable, the existing model-based technologies are inappropriate for MIMO nonlinear MASs. Therefore, in light of the challenge aroused by aperiodic DoS attacks and output saturation, a novel switched error system-based model-free adaptive containment control approach is developed to preserve data integrity while the stability of containment error is guaranteed. Numerical simulations indicate the validity of the obtained results.}, archive = {J_TASE}, author = {Wen Tan and Zhongsheng Hou and Yuan-Xin Li}, doi = {10.1109/TASE.2024.3469153}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7762--7772}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven containment control for unknown MIMO nonlinear MASs under aperiodic DoS attacks}, volume = {22}, year = {2025}, }
• (2025). MgCNL: A sample separation approach via multi-granularity balls for fault diagnosis with the interference of noisy labels. TASE, 22, 7748--7761. (www)

  @article{ , abstract = {The fault diagnosis based on supervised learning has achieved remarkable results in the intelligent manufacturing, making it an important guarantee for long-term safe and stable operation in modern industry. However, the accuracy heavily relies on high-quality annotation labels, which are expensive to obtain, limiting the diagnosis models applicability in many scenarios. Although obtaining automatically annotated samples from annotators is a promising solution, the generated dataset is always containing incorrect labels (noisy labels), due to perceptual limitations, resulting in low or even invalid the accuracy of model. With the goal of handling this challenge, a diagnostic approach based on multi-granularity information fusion to combat noisy labels, called MgCNL, is proposed, to train the model with high-accuracy, without knowing the specific noise ratio. Specifically, inspired by granular-ball computing, a confidence evaluation method of labels is designed, so that samples with high confidence labels can be selected from dataset with noisy labels for supervised learning, thus avoiding the negative impact of incorrect labels on model performance. Finally, the efficacy was demonstrated on three datasets using different backbones: MgCNL successfully reduced the adverse impact of noisy labels, achieving significantly better results than other advanced methods in various noisy scenarios, which offers a competitive model training strategy for practitioners in intelligent manufacturing or industrial fault diagnosis who are hampered by the costs associated with sample labeling. Note to Practitioners—In modern industry, the cost of manual/expert annotation for high-quality data is is prohibitively expensive, and the data annotated by automatic annotators often contains noisy labels that seriously damages the accuracy of models, which makes many data-driven diagnosis models constrained by training data and difficult to put into practice, posing an urgent challenge to the automation and intelligence of the manufacturing industry. To address this challenge, this article proposed a robust training strategy called MgCNL, aimed at offsetting the negative impact of noisy labels, in the hope that automatic annotation strategy with lower cost can be more widely applied in model training tasks for industrial practice. MgCNL, based on multi-granularity information, can effectively select high-confidence samples from datasets for supervised learning, even under unknown proportions of noise labels, thus reducing the misleading impact of noisy labels on diagnostic models. As a result, MgCNL possesses the ability to robustly train high-accuracy diagnostic models in data with noisy labels, thus enabling automatic annotators to replace experts in dataset construction as a more economical and efficient potential technical approach. Meanwhile, MgCNL also brings value to datasets with uncertain labels, making them applicable without the need to invest significant human resources to verify label reliability.}, archive = {J_TASE}, author = {Fir Dunkin and Xinde Li and Heqing Li and Guoliang Wu and Chuanfei Hu and Shuzhi Sam Ge}, doi = {10.1109/TASE.2024.3469000}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7748--7761}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MgCNL: A sample separation approach via multi-granularity balls for fault diagnosis with the interference of noisy labels}, volume = {22}, year = {2025}, }
• (2025). Fuzzy-boosted event-triggered tracking control of unknown nonlinear networked systems: A PSO-driven RL approach. TASE, 22, 7736--7747. (www)

  @article{ , abstract = {This article is centered on an optimal tracking control problem for nonlinear networked systems with limited network bandwidth and unavailable dynamics. Foremost, considering the absence of system dynamics, a state identifier grounded in generalized fuzzy hyperbolic model (GFHM) is estabilshed to eliminate the reliance on known system states. Meanwhile, in order to efficiently utilize the limited communication resources within the network, a dynamic event-triggered mechanism (DETM) with an adaptive threshold is adopted. Subsequently, the DETM-based Hamilton-Jacobi-Isaacs (HJI) equation is constructed following the Bellman optimality theory. Furthermore, we develop a particle swarm optimization (PSO)-driven reinforcement learning (RL) algorithm to achieve the desired optimal tracking control strategy. It is demonstrated that the tracking error system can be promised to be uniformly ultimately bounded (UUB) stability. Eventually, a numerical experiment is provided to illustrate the validity of the developed RL algorithm.Note to Practitioners—This study targets the optimal tracking control issue of unknown networked systems, which holds practical significance in various fields such as robotics, autonomous vehicles and industrial control. In real-world environments, the dynamics of networked systems are not always accessible, thus this paper employs GFHM to approximate unknown dynamic information. Additionally, due to the involvement of the network, the restriction of communication resources is an unavoidable concern. To address this, DETM is adopted to improve network transmission efficiency. Subsequently, to achieve the optimal tracking control, the HJI equation based on DETM is constructed. The solution to the HJI equation is obtained through the RL algorithm supported by critic neural network, with its weights trained using the PSO algorithm. Compared to the commonly applied gradient descent method, under the PSO algorithm, the initial weights can be randomly assigned, making implementation easier and increasing the success rate of system operation. This is of significant importance in practical applications. Besides, a simulation example illustrates the feasibility of the developed method.}, archive = {J_TASE}, author = {Jian Liu and Wenqiang Wang and Bing Hu and Jinliang Liu and Xiang-Peng Xie and Engang Tian}, doi = {10.1109/TASE.2024.3468614}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7736--7747}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fuzzy-boosted event-triggered tracking control of unknown nonlinear networked systems: A PSO-driven RL approach}, volume = {22}, year = {2025}, }
• (2025). Optimal containment control for T-S fuzzy MASs using improved transform matching membership functions. TASE, 22, 7724--7735. (www)

  @article{ , abstract = {This paper investigates the containment performance optimization control issue of nonlinear multi-agent systems (MASs) with packet loss in the framework of Takagi-Sugeno (T-S) fuzzy approach. Unlike the existing parallel distributed compensation (PDC) and non-PDC control methods that the membership functions (MFs) of the designed containment controllers should keep consistent with those of MASs or satisfy inequality restrictions, an online transform matching MFs control scheme, which can enhance the MFs design and remove the inequality constraints, is proposed for T-S fuzzy MASs. By virtue of the polynomial disassembling strategy and slack matrix inequalities technique, on the basis of Lyapunov stability theory, sufficient conditions are derived for T-S fuzzy MASs such that the followers can achieve states containment regardless of data loss. After that, in the premise of ensuring containment ability of considered MASs, different from the existing works with fixed transform coefficients, a novel optimization policy utilizing differential evolution (DE) algorithm is first proposed to seek the optimal values of scaling and bias coefficients in real time. With the aid of the proposed optimization control, a superior containment performance can be derived, i.e., the followers can faster access the convex hull formed by the trajectories of leaders compared with traditional method. Finally, the effectiveness and superiority of the proposed transform matching MFs online optimization strategy are testified via a contrastive example. Note to Practitioners—In actual engineering, due to complicated working scenario or their own inherent properties, the current systems are hardly to be analyzed and controlled. On the other hand, in T-S fuzzy model-based control method, which is an effective approach to describe and control nonlinear systems, the design of membership functions of controllers is limited by inequality constraints or constructed by designers’ intuition, which will seriously impact the optimal performance of systems. Motivated by above analysis, this paper proposes a new membership functions online optimization approach for nonlinear MASs to automatically seek the optimal values of membership functions thereby obtaining a superior containment ability compared with existing traditional technique. The proposed control approach is significative in real-world applications such as the improvement of formation performance or rapid obstacle avoidance of unmanned aerial vehicles. The practicability of the designed online optimization strategy is illustrated via a contrastive example.}, archive = {J_TASE}, author = {Zhenxing Zhang and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3468474}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7724--7735}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal containment control for T-S fuzzy MASs using improved transform matching membership functions}, volume = {22}, year = {2025}, }
• (2025). LogiCode: An LLM-driven framework for logical anomaly detection. TASE, 22, 7712--7723. (www)

  @article{ , abstract = {This paper presents LogiCode, a novel framework that leverages Large Language Models (LLMs) for identifying logical anomalies in industrial settings, moving beyond the traditional focus on structural inconsistencies. By harnessing LLMs for logical reasoning, LogiCode autonomously generates Python codes to pinpoint anomalies such as incorrect component quantities or missing elements, marking a significant leap forward in anomaly detection technologies. A custom dataset “LOCO-Annotations” and a benchmark “LogiBench” are introduced to evaluate the LogiCode’s performance across various metrics including binary classification accuracy, code generation success rate, and precision in reasoning. Findings demonstrate LogiCode’s enhanced interpretability, significantly improving the accuracy of logical anomaly detection and offering detailed explanations for identified anomalies. This represents a notable shift towards more intelligent, LLM-driven approaches in industrial anomaly detection, promising substantial impacts on industry-specific applications. Our code are available at https://github.com/22strongestme/LOCO-Annotations. Note to Practitioners—This work introduces LogiCode, an innovative system leveraging Large Language Models (LLMs) for logical anomaly detection in industrial settings, shifting the paradigm from traditional visual inspection methods. LogiCode autonomously generates Python codes for logical anomaly detection, enhancing interpretability and accuracy. Our novel approach, validated through the “LOCO-Annotations” dataset and LogiBench benchmark, demonstrates superior performance in identifying logical anomalies, a challenge often encountered in complex industrial components like assembly and packaging. LogiCode provides a significant advancement in addressing the nuanced requirements of detecting logical anomalies, offering a robust and interpretable solution to practitioners seeking to enhance quality control and reduce manual inspection efforts.}, archive = {J_TASE}, author = {Yiheng Zhang and Yunkang Cao and Xiaohao Xu and Weiming Shen}, doi = {10.1109/TASE.2024.3468464}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7712--7723}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {LogiCode: An LLM-driven framework for logical anomaly detection}, volume = {22}, year = {2025}, }
• (2025). Event-triggered model predictive control of spacecraft formation. TASE, 22, 7696--7711. (www)

  @article{ , abstract = {This paper proposes a model predictive control method based on a dynamic event-triggered strategy for high-precision formation control of spacecraft formations using continuous low thrust. An optimal formation control problem is formulated based on the J2 linear time-varying relative dynamics using the Legendre pseudospectral method, and the model predictive controller is designed based on this optimization problem. Thereafter, a dynamic event-triggered strategy is introduced in the model predictive controller, in which the optimization problem is solved only when the event-triggered condition is violated. As a result, the computational pressure of the model predictive control algorithm is greatly reduced without increasing fuel consumption or degrading control accuracy. Meanwhile, the dynamic event-triggered condition allows for relatively even event-triggered intervals, which is more conducive to engineering applications. The feasibility and stability of this control strategy are demonstrated, and the simulation results under two different formation control scenarios show that the proposed controller achieves equivalent control performance to the conventional model predictive controller, and the computational pressure is greatly reduced by designing the dynamic event-triggered condition. Note to Practitioners—This paper presents a study on an event-triggered model predictive control (MPC) strategy for spacecraft formation control. The objective is to achieve high-precision control for spacecraft formations, including formation maintenance and reconfiguration, in the presence of bounded external disturbances. The proposed control strategy significantly reduces the computational burden of the model predictive control by introducing a dynamic event-triggered policy while preserving the control performance of the MPC. Moreover, the design of event-triggered conditions effectively prevents the occurrence of the Zeno phenomenon. The feasibility, system stability, and convergence of the control strategy are rigorously analyzed, and simulation-based verification is conducted in a binary-star formation control scenario. It is important to note that the event-triggered model predictive control strategy presented in this paper is applicable to various leader-follower type formation systems, extending beyond the scope of spacecraft formation control.}, archive = {J_TASE}, author = {Zhaobo Sun and Baolin Wu and Danwei Wang and Junyu Chen}, doi = {10.1109/TASE.2024.3468459}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7696--7711}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered model predictive control of spacecraft formation}, volume = {22}, year = {2025}, }
• (2025). Bi-objective optimization of a flow shop scheduling problem under time-of-use tariffs. TASE, 22, 7679--7695. (www)

  @article{ , abstract = {Time-of-use (ToU) tariffs flexibly offer markedly cheap electricity prices to industrial and residential users during off-peak periods, encouraging them to shift their peak electricity demands in valley periods. Although ToU tariffs play a crucial role in balancing electricity supply and demand, especially in energy-intensive industries, the best trade-off between industrial performance and energy costs has not been well explored. Manufacturing consumes a substantial amount of energy, primarily in the form of electricity, leading to imbalances in power consumption. The flow shop scheduling (FSS) model is one of the most prevalent models in manufacturing. To explore the significant role of ToU tariffs in manufacturing, this study addresses a bi-objective FSS problem under ToU tariffs. The objective is to find the optimal balance between customer satisfaction and total electricity cost. A tight mixed integer programming model is developed to solve this NP-hard problem using business optimizers. On the bases of the problem properties demonstrated in this study, valid inequalities are designed to reduce the solution space of the problem. For small-scale instances, an improved $\varepsilon $ -constraint method is presented to find the Pareto front. For medium and large-scale instances, a two-stage fruit fly optimization (TFFO) algorithm is developed to obtain the near Pareto front. Experimental results demonstrate the efficiency and effectiveness of the proposed model and algorithms.Note to Practitioners—Scheduling for complex systems remains a formidable challenge in manufacturing. Energy cost saving is a major objective for all energy-intensive industries. Effective scheduling is crucial for businesses achieving eco-friendly performance, especially under ToU tariffs. This study aims to provide efficient scheduling model and methods that can guide decision-makers in fostering ecological transitions. The $\varepsilon $ -constraint method can find globally optimal solutions within given constraints. This situation is particularly beneficial for small-scale production systems requiring high accuracy. The TFFO algorithm can handle complex industrial environments and enhance production efficiency. Additionally, the TFFO algorithm is flexible and extensible, enabling it to be generalized to other production scenarios. Overall, the proposed model and algorithms lay a solid foundation for achieving efficient scheduling under ToU tariffs.}, archive = {J_TASE}, author = {Xinyue Wang and Feng Chu and Tao Ren and Danyu Bai}, doi = {10.1109/TASE.2024.3468448}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7679--7695}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bi-objective optimization of a flow shop scheduling problem under time-of-use tariffs}, volume = {22}, year = {2025}, }
• (2025). The input-mapping-based online learning sliding mode control strategy with low computational complexity. TASE, 22, 7670--7678. (www)

  @article{ , abstract = {The data-driven sliding mode control (SMC) method proves to be highly effective in addressing uncertainties and enhancing system performance. In our previous work, we implemented a co-design approach based on an input-mapping data-driven technique, which effectively improves the convergence rate through historical data compensation. However, this approach increases computational complexity in multi-input and multi-output (MIMO) systems due to the dependency of the number of online optimization variables on system dimensions. To improve applicability, this paper introduces a novel input-mapping-based online learning SMC strategy with low computational complexity. First, a new sliding mode surface is established through online convex combination of pre-designed offline surfaces. Then, an input-mapping-based online learning sliding mode control (IML-SMC) strategy is designed, utilizing a reaching law with adaptively adjusted convergence and switching coefficients to minimize chattering. The input-mapping technique employs the mapping relationship between historical input and output data for predicting future system dynamics. Accordingly, an optimization problem is formulated to learn from the past dynamics of the uncertain system online, thereby enhancing system performance. The optimization problem in this paper features fewer variables and is independent of system dimension. Additionally, the stability of the proposed method is theoretically validated, and the advantages are demonstrated through a MIMO system. Note to Practitioners—The design of control strategies that reduce the impact of mismatches between practical systems and models on system performance, while also ensuring applicability, is crucial. To address this issue, this paper proposes a low-complexity IML-SMC strategy. This strategy uses historical real input-output information and the mapping relationship with future dynamics to compensate for the impact of unknown dynamics and improve the system’s convergence rate. Notably, the control strategy introduced in this paper significantly reduces online computational complexity, ensuring applicability, and stability is proven. When there is a deviation between the model and the actual system, practitioners can implement the low-complexity IMC-SMC strategy proposed in this paper to more quickly achieve the control objectives in the actual system.}, archive = {J_TASE}, author = {Yaru Yu and Aoyun Ma and Dewei Li and Yugeng Xi and Furong Gao}, doi = {10.1109/TASE.2024.3467383}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7670--7678}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The input-mapping-based online learning sliding mode control strategy with low computational complexity}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-driven ADP for N-player nonzero-sum games of constrained nonlinear systems. TASE, 22, 7657--7669. (www)

  @article{ , abstract = {In this paper, the dynamic event-driven optimal control problem is investigated for a class of continuous-time nonlinear systems subject to asymmetric input constraints in the framework of nonzero-sum (NZS) games. Initially, by constructing a modified value function, the respective asymmetric input constraint requirements of the controllers involved in the NZS games are successfully satisfied. Then, based on the Bellman’s optimality principle, the N-coupled Hamilton-Jacobi equations are derived for the N-player NZS games. After that, the adaptive dynamic programming (ADP) method is employed to seek for the optimal control policies, in which the simpler single critic neural network structure, instead of the dual network structure of actor-critic in the typical ADP algorithm, is applied. Furthermore, an improved critic network weight updating law is proposed to ensure the stability of the closed-loop system without a hard-to-find initial admissible control scheme. In addition, in order to reduce the update frequency of the controllers to a greater extent, a dynamic event-driven mechanism with adjustable threshold is developed. Finally, a simulation example is given to demonstrate the validity of the developed event-driven control scheme. Note to Practitioners—This paper aims to address the NZS games problem for a category of multi-player continuous-time nonlinear systems featuring multiple input constraints. The applicability of this approach can be widely extended to practical domains, including control applications for reconfigurable robot systems, networked communication systems, etc. The majority of researches on multi-player NZS games problem are focused on the impact of symmetric input constraints. Especially under the premise of ensuring controller optimality, the challenge lies in how to ensure effective control functionality while subjecting the controller to asymmetric constraints. Furthermore, the existing ADP algorithms often depend on an initial admissible control, significantly elevating the implementation difficulty of control solutions in practical applications. To address these challenges, an improved ADP algorithm is developed for input-constrained nonlinear systems within a NZS game framework. This method not only guarantees that the optimal controllers under asymmetric constraints can stabilize all signals, but also avoids the search for challenging-to-find initial admissible controls, thus streamlining the control implementation process.}, archive = {J_TASE}, author = {Siyu Guo and Yingnan Pan and Hongyi Li and Liang Cao}, doi = {10.1109/TASE.2024.3467382}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7657--7669}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-driven ADP for N-player nonzero-sum games of constrained nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Optimizing fixture layout for compliant part assembly: A kriging-based metaheuristic. TASE, 22, 7642--7656. (www)

  @article{ , abstract = {This paper studies a fixture layout optimization problem for the ship assembly. Current studies focus on the positioning problem, i.e., optimizing the positions of a constant number of fixtures, while less considers the sizing problem, i.e., deciding the number of used fixtures. In most of the cases, it is difficult to assure that the adopted number of fixtures can lead to a layout that satisfies all the engineering requirements. The objective of this paper is to minimize the number of fixtures required while optimizing their positions, ensuring adherence to a set of engineering constraints. These include the maximum allowed nodal displacement and the assembly gap between two parts. We propose an efficient two-step metaheuristic approach. In the first step, fixtures are added iteratively to the parts until a feasible layout is achieved. In the second step, a large neighborhood search is employed to remove fixtures while maintaining the feasibility of the layout. To mitigate the computational cost required by finite element simulation, a Kriging model is proposed to evaluate the neighborhood solutions. Numerical experiments show that the proposed method can efficiently solve the fixture sizing and positioning problem in a case study of ship panel assembly. The efficacy of the Kriging model in accelerating layout optimization has been validated. Furthermore, the efficiency of the proposed method is demonstrated through comparisons with two existing approaches in the literature. Note to Practitioners—This paper was motivated by the problem of holding large compliant parts in the ship assembly process. Sheet metals with a low thickness-to-length ratio are susceptible to deformation during assembly when an inappropriate fixture layout is adopted. Existing approaches for layout design focus on optimizing the positions of the fixtures, while the number of fixtures needs to be determined based on engineering experience. This paper suggests a new approach to determine the minimum number of fixtures required, as well as their positions, to guarantee a set of quality requirements provided by the user. This approach employs the direct stiffness method to evaluate part deformation and an efficient metaheuristic to optimize the layout. During the optimization procedure, we use the Kriging technique to predict the layout’s performance, allowing us to obtain high-quality solutions with lower simulation efforts. We demonstrate that the proposed method can successfully generate a layout that satisfies the quality specifications in a case study of ship panel assembly. In comparison to two existing approaches, the proposed method yields a feasible layout with fewer fixtures, thereby reducing fixture setup time and building costs.}, archive = {J_TASE}, author = {Changhui Liu and Chunlong Yu and Jing Wang and Jianfeng Liu and Jianfeng Lu}, doi = {10.1109/TASE.2024.3467177}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7642--7656}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimizing fixture layout for compliant part assembly: A kriging-based metaheuristic}, volume = {22}, year = {2025}, }
• (2025). A multi-observer based optimal control method for nonlinear systems under sensor attacks. TASE, 22, 7632--7641. (www)

  @article{ , abstract = {A multi-observer based optimal control method is presented for nonlinear continuous-time systems under sensor attacks. On the basis of the ideas of sensor redundancy and multi-observers, a neural network (NN) based multi-observer method is introduced to find the attack-free set of sensors in order to deal with the influences of sensor attacks. Then, the adaptive dynamic programming (ADP) method with critic NN approximation is developed by utilizing the correct state estimate. The properties of the multi-observer based ADP method are analyzed by utilizing the Lyapunov theory. Numerical analysis is conducted to show the efficiency of the presented method. Note to Practitioners—In practical engineering, the communication networks may provide new access points for cyberattacks which try to degrade the system control performance because of the physical constraints. These cyberattacks can affect the quality of the system data transmission, which brings great challenges to the effective optimal control of systems. Aiming at the above problems, this paper designs a multi-observer based optimal control method for nonlinear continuous-time systems under sensor attacks. A neural network (NN) based multi-observer method is provided to find the attack-free set of sensors. Then, the adaptive dynamic programming (ADP) method is presented to realize the optimal control of systems based on the state estimate. Numerical analysis is given to show the correctness of the presented method.}, archive = {J_TASE}, author = {Hongyang Li and Qinglai Wei}, doi = {10.1109/TASE.2024.3466894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7632--7641}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A multi-observer based optimal control method for nonlinear systems under sensor attacks}, volume = {22}, year = {2025}, }
• (2025). Safe control framework of multi-agent systems from a performance enhancement perspective. TASE, 22, 7622--7631. (www)

  @article{ , abstract = {In the control problems of multi-agent systems, collision avoidance is a fundamental safety requirement. One effective approach to ensure safety involves combining control barrier functions (CBFs) with quadratic programming (QP), where nominal control inputs are incorporated into QPs to achieve desired control objectives. Additionally, it is crucial to study the control performance of multi-agent systems to balance these objectives with control efforts. This work demonstrates that the performance index is closely related to the hyperparameters in the CBF-based QP controller, and the Bayesian optimization algorithm is used to optimize and improve performance. Firstly, a safe control approach is developed and a unified performance enhancement framework is established to optimize the performance index. Hyperparameters are then explored and categorized, introducing the concept of feasible hyperparameters to describe the attainability of control objectives. Subsequently, the constrained Bayesian optimization algorithm is employed to identify a set of feasible and optimal hyperparameters in a data-driven manner, even when the functional expressions of performance and constraints are unknown. Finally, experiments are conducted to demonstrate the feasibility of the proposed methods in multi-agent systems. Note to Practitioners—Both safety and optimality are of great importance in the control systems. The design and development of controllers for multi-agent systems are currently undergoing significant evolution. Academic researchers and industrial practitioners are actively refining controller designs to perform better in a variety of collaborative tasks. With practical applications in mind, there is a growing demand for control techniques capable of ensuring a safe operating environment while maintaining efficiency in energy consumption. Therefore, this paper aims to furnish practitioners and researchers with a safe control framework, facilitating the refinement of optimal control solutions for efficient collaboration among multiple agents.}, archive = {J_TASE}, author = {Boqian Li and Shengbo Wang and Zhenyuan Guo and Song Zhu and Junjian Huang and Junwei Sun and Guanghui Wen and Shiping Wen}, doi = {10.1109/TASE.2024.3466791}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7622--7631}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe control framework of multi-agent systems from a performance enhancement perspective}, volume = {22}, year = {2025}, }
• (2025). Fast and accurate multi-agent trajectory prediction for crowded unknown scenes. TASE, 22, 7606--7621. (www)

  @article{ , abstract = {This paper studies the problem of multi-agent trajectory prediction in crowded unknown environments. A novel energy function optimization-based framework is proposed to generate prediction trajectories. Firstly, a new energy function is designed for easier optimization. Secondly, an online optimization pipeline for calculating parameters and agents’ velocities is developed. In this pipeline, we first design an efficient group division method based on Frechet distance to classify agents online. Then the strategy on decoupling the optimization of velocities and critical parameters in the energy function is developed, where the slap swarm algorithm and gradient descent algorithms are integrated to solve the optimization problems more efficiently. Thirdly, we propose a similarity-based resample evaluation algorithm to predict agents’ optimal goals, defined as the target-moving headings of agents, which effectively extracts hidden information in observed states and avoids learning agents’ destinations via the training dataset in advance. Experiments and comparison studies verify the advantages of the proposed method in terms of prediction accuracy and speed. Note to Practitioners—Autonomous robots and vehicles are rapidly integrated into social life and industry, and the scenarios that robots work with multiple people or other moving objects in a crowded environment such as streets and factories will be quite common. One of the most important problems for the robot to solve is the real-time and accurate trajectory prediction of multiple agents around itself to ensure safe navigation. However, existing methods either require prior information to train models or critical parameters in advance or have insufficient prediction accuracy, which are not suitable for robot safe navigation in real applications. In this paper, we investigate the real-time multi-agent trajectory prediction problem for a robot in crowded unknown environments. To obtain the accurate predicted trajectories in real-time, we propose a new energy function optimization-based framework to forecast multi-agent trajectories in crowded unknown scenarios. This framework utilizes the observed data to infer unknown information without the dataset and optimizes the trajectories very efficiently, which can be adopted for robot motion planning and navigation in real-world environments.}, archive = {J_TASE}, author = {Xiuye Tao and Huiping Li and Bin Liang and Yang Shi and Demin Xu}, doi = {10.1109/TASE.2024.3466272}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7606--7621}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast and accurate multi-agent trajectory prediction for crowded unknown scenes}, volume = {22}, year = {2025}, }
• (2025). A bounded sub-optimal approach for multi-agent combinatorial path finding. TASE, 22, 7590--7605. (www)

  @article{ , abstract = {Multi-Agent Path Finding (MAPF) seeks collision-free paths for multiple agents from start to goal locations. This paper considers a generalization of MAPF called Multi-Agent Combinatorial Path Finding (MCPF) where agents must collectively visit a set of intermediate target locations before reaching their goals. MCPF is challenging as it involves both planning collision-free paths for multiple agents and target sequencing, i.e., assigning targets to and computing the visiting order for each agent. A recent method Conflict-Based Steiner Search (CBSS) is developed to solve MCPF to optimality, which, however, does not scale well when the number of agents or targets is large (e.g. 50 targets). While MAPF research has developed methods to plan bounded sub-optimality paths for many agents, it remains unknown how to find bounded sub-optimal solutions in the presence of many targets. This paper fills this gap by developing a method AK* for target sequencing (A for Approximation and K* for K-best), which leverages approximation algorithms for traveling salesman problems. AK* is motivated by MCPF, but is a standalone method that can solve K-best routing problems in general. We prove that AK* has worst-case polynomial runtime complexity and finds bounded sub-optimal solutions. With AK*, we develop two CBSS variants that find bounded sub-optimal paths for MCPF. Our results verify the fast running speeds of our methods with up to 200 targets. Note to Practitioners—The motivation of this paper originates from the need to plan conflict-free paths for multiple mobile robots in cluttered environment in warehouse logistics, manufacturing and inspection. While existing methods for multi-agent planning typically consider finding paths from starts to goals, this paper investigates the case, where agents must collectively visit a set of intermediate target locations before reaching their goals, for the purpose of inspection, picking or placing parts, etc. To solve the problem, this paper first develops an algorithm to find K-best solutions for traveling salesman problems with bounded sub-optimality, which then leads to two multi-agent planners that can handle hundreds of targets and tens of agents. We provide a Gazebo simulation to showcase the usage of the planner in a warehouse like environment.}, archive = {J_TASE}, author = {Zhongqiang Ren and Sivakumar Rathinam and Howie Choset}, doi = {10.1109/TASE.2024.3466183}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7590--7605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A bounded sub-optimal approach for multi-agent combinatorial path finding}, volume = {22}, year = {2025}, }
• (2025). Finite-time stabilization of nonlinear systems with actuator saturation. TASE, 22, 7582--7589. (www)

  @article{ , abstract = {This paper deals with finite-time stabilization (FTS) of nonlinear systems subject to actuator saturation. We present an analytical method to design the finite-time controller such that the FTS of the system can be guaranteed in the presence of actuator saturation. Moreover, a certain interrelation among the saturation bound, the system structure, and the controller parameters is established. Based on such a relationship, the settling time (ST) and the domain of attraction (DA) can both be estimated in an easily verifiable manner, which are ignored in many existing results of actuator saturation. Moreover, it is shown that the parameter of the designed controller can provide a tradeoff between the estimations of the ST and the DA by adjusting the controller parameters. All the results are illustrated in simulations. Note to Practitioners—Actuator saturation, a phenomenon where the control input exceeds the limits of an actuator, plays a crucial role in control theory and its applications. In control systems, actuator saturation can lead to performance degradation, instability, and even system failure if not properly addressed, such as Chornobyl nuclear power plant accident and the YF-22 crash incident. Note that in the presence of actuator saturation, most existing results focus on Lyapunov stability or finite-time boundedness, which makes it challenging to guarantee that the system converges to equilibrium within a finite time. This paper was motivated to break through this bottleneck and solve the FTS problem. Unlike the existing saturation results, we first decompose the system based on controllability. Then a straightforward controller design is proposed for the controllable part, while constraints on the system structure are provided for the uncontrollable part. Based on such conditions, not only is the estimation of the DA easily obtained, but the corresponding ST estimation is also derived. The obtained results in this paper are of great significance for providing a more general and easily achievable design for saturated control input. And these results are applied to various industries, such as aerospace, automotive, mobile robotics technology, and industrial automation.}, archive = {J_TASE}, author = {Shuchen Wu and Xiaodi Li}, doi = {10.1109/TASE.2024.3466049}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7582--7589}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time stabilization of nonlinear systems with actuator saturation}, volume = {22}, year = {2025}, }
• (2025). Cooperative path tracking-based learning control for unknown multi-agent systems via dynamic event-triggered mechanisms. TASE, 22, 7569--7581. (www)

  @article{ , abstract = {This paper investigates the event-triggered output path-tracking control of networked heterogeneous multi-vehicle (agent) systems with unknown model dynamics. Different from most existing distributed observer methods to estimate the leader vehicle’s state matrix and state, these state-based observer approaches raise the disadvantages of high dimensionality and high frequency of data exchange of state. To address this, in this paper, we propose a novel adaptive distributed output observer (ADOO) that estimates the coefficients of the minimal polynomial instead of requiring knowledge of all the entries of the leader vehicles system matrix. Moreover, our proposed ADOO is model-free without relying on the leader’s accurate system, unlike the model-based way in existing works. Meanwhile, an asynchronous dynamic event-triggered control strategy is developed to reduce the communication load among neighboring vehicles. Then, a decentralized path-tracking controller is learned via a model-free matrix updating learning technique to achieve optimal path-tracking control without requiring an initial stabilizing control policy. By rigorous mathematical analysis shows that our proposed algorithms not only can greatly reduce the dimension of existing observer methods and the frequency of information exchange among neighboring vehicles, but also exclude the Zeno phenomenon. Finally, the numerical simulation is used to validate the efficiency of the theoretical algorithms under investigation. Note to Practitioners—This paper investigates the learning-based distributed optimal path-tracking control for heterogeneous vehicle systems with event-triggered communications. The proposed distributed control algorithms can be applied to multiple ground vehicles, air vehicles, and underwater vehicles. Unlike existing results on distributed state observer methods, the obtained results rely on accurate system dynamics and ignore the transient performance, which makes the designed controller far from optimal in potential applications. Moreover, the optimal learning algorithm strictly relies on the initial stabilizing control policy related to accurate system dynamics, and it may be helpless when the system model is completely unknown. To overcome those issues, we developed a model-free matrix updating learning technique to study distributed output path tracking control by collecting system data instead of using accurate system dynamics. Our algorithm not only overcomes the initial stabilizing assumption but also ensures the algorithm convergence in a model-free fashion. Besides, our proposed distributed observer can greatly lower the dimension and the frequency of information exchange of existing observers. Potential applications of the proposed control algorithms include cooperative formation control of heterogeneous unmanned systems.}, archive = {J_TASE}, author = {Yong Xu and Meng-Ying Wan and Di Mei and Zheng-Guang Wu}, doi = {10.1109/TASE.2024.3465439}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7569--7581}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative path tracking-based learning control for unknown multi-agent systems via dynamic event-triggered mechanisms}, volume = {22}, year = {2025}, }
• (2025). Heuristic scheduling for robotic job shops using petri nets and artificial potential fields. TASE, 22, 7556--7568. (www)

  @article{ , abstract = {Numerous mobile robots are employed for material transporting in today’s manufacturing industry. Therefore, it is crucial to consider the time consumption of both process operations and transportation to minimize the makespan. This paper addresses the joint scheduling optimization problem of automated guided vehicle (AGV) routing and task allocation for robotic job shops (RJSs). A novel artificial potential field (APF) design method is developed based on place-timed Petri nets (PNs) for heuristic scheduling in RJSs. First, a real-life scale RJS is modeled with a place-timed PN, which contains a route subnet and several task ones. The scheduling problem is then transformed into a search one for the transition firing sequence with the minimum makespan, which can be obtained by the A* algorithm. Second, a task-APF design method is proposed for sorting processes. It defines potential energy parameters for places in task subnets by utilizing the topological structure of PNs. The heuristic function $h_{\mathrm {MPD}}$ is obtained, and it is proved to be admissible. Experimental results show that the $h_{\mathrm {MPD}}$ heuristic enables the A* algorithm to generate an optimal schedule for RJSs, and its search efficiency surpasses that of the most advanced admissible heuristic functions available, particularly in the scenario described in this paper. However, its search efficiency decreases significantly when the number of jobs and AGVs increases. Third, a route-APF design method is proposed for AGV routing, and a novel heuristic function, denoted by $h_{\mathrm {TPD}}$ , is developed based on the task-APF and route-APF. Experimental results demonstrate that $h_{\mathrm {TPD}}$ is more adaptable for scheduling RJSs with large-scale jobs and multiple AGVs than $h_{\mathrm {MPD}}$ , which improves the search efficiency by an average of one order of magnitude. It can find an optimal or near-optimal schedule within a reasonable amount of time, even when dealing with hundreds of jobs and multiple AGVs. Note to Practitioners—Scheduling manufacturing systems is typically an essential and complex combinatorial optimization problem, known as NP-hard, involving AGV routing and processing sorting. To address these challenges, we propose a heuristic scheduling method with two heuristic functions, which are constructed by a novel artificial potential field design method based on place-timed Petri nets. This approach can produce a high-quality schedule within a reasonable time frame and demonstrates superiority in scheduling, especially for large-scale jobs in robotic job shops. Therefore, it can be applied to real-life scheduling problems of manufacturing systems.}, archive = {J_TASE}, author = {Sijia Yi and Jiliang Luo}, doi = {10.1109/TASE.2024.3464857}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7556--7568}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Heuristic scheduling for robotic job shops using petri nets and artificial potential fields}, volume = {22}, year = {2025}, }
• (2025). An inverted differential mechanism capable of achieving very large amplification ratio: Design and control. TASE, 22, 7546--7555. (www)

  @article{ , abstract = {A multitude of flexure-based displacement amplifiers have been developed to amplify piezoelectric actuators for achieving both high-precision motion and large output stroke. Single-stage amplifiers are compact but provide only amplification of several times, while multi-stage amplifiers are able to achieve amplification of dozens of times but are generally bulky in structure. In this work, a displacement amplifier with only a single stage displacement amplifier, which shows the capability of obtaining a very large amplification ratio, is proposed. The displacement amplifier contains two semi-bridge mechanisms with a slight geometric difference between them. This slight difference makes the amplifier require a very small input while achieving a large displacement at the output, leading to a very large amplification ratio. A kinetostatic model considering the nonlinearities in the deflections of both the flexure hinges and the links for the amplifier is developed, based on which the parameters of the amplifier are optimized to maximize the amplification ratio, resulting in an amplifier exhibiting an amplification ratio of 107. The optimization results were validated by those of a finite element model, proving the effectiveness and correctness of the proposed amplifier and the kinetostatic model. The finalized design was prototyped and the measured amplification ratios in a bilateral output mode and a unilateral output mode are 98.10 and 88.42, respectively. A neural network PID controller was designed for the displacement amplifier, with a maximum trajectory tracking error less than 4.7% of the displacement amplifier was achieved. Note to Practitioners—This paper was motivated by the need to extend the actuation stroke of piezoelectric actuators, which are crucial in automation systems requiring high-precision displacement and large motion range. Traditional single-stage amplifiers are compact but provide limited extension. Multi-stage amplifiers offer greater extension, but their bulkiness limits the practical use. We have developed an innovative single-stage amplifier with very large amplification ratio that incorporates a unique arrangement of semi-bridge mechanisms. This design significantly extend the actuation stroke of piezoelectric actuators with a compact structure. The integration of a neural network-based PID controller improves the accuracy and efficiency of the positioning control of the amplifiers system. The principles and design approach we used could also be applied to fields where high precision and large displacement are needed, for example, micro/nano manufacturing and aerospace applications. This could open up new avenues for enhancing the efficiency and capability of devices in these sectors.}, archive = {J_TASE}, author = {Houqi Wu and Guimin Chen}, doi = {10.1109/TASE.2024.3464688}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7546--7555}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An inverted differential mechanism capable of achieving very large amplification ratio: Design and control}, volume = {22}, year = {2025}, }
• (2025). BladeView: Toward automatic wind turbine inspection with unmanned aerial vehicle. TASE, 22, 7530--7545. (www)

  @article{ , abstract = {This paper presents a fully automatic method, BladeView, for drone-based wind turbine blade inspection using an Unmanned Aerial Vehicle (UAV). With the need for highly efficient blade inspection coupled with the rapid increase of wind turbines, existing methods provide limited automation on wind turbine parameter estimation, full blade coverage, and safety control. We introduce an Automatic Parameter Calculation (APC) algorithm and an Automatic Flight System (AFS) in BladeView to compute wind turbine parameters and inspection paths, respectively. Leveraging triangulation and linear fitting integration techniques, the APC automatically calculates the wind turbine parameters and estimates the relative angle and position between a drone and the turbine. Furthermore, with dynamic path finding and B-spline optimization, the AFS plans a path covering 3 blades within specified flight corridors, in compliance with the turbine parameters obtained from APC. Thus, the proposed BladeView can properly ensure an inspection’s automation, coverage, safety, and smoothness. The efficiency and usability of BladeView are validated through 100,000 flight simulations in the Gazebo simulation environment and 9,239 field runs at various wind farms, including offshore, near-shore, deserts, mountainous areas, farmlands, and suburbs.Note to Practitioners—The proposed BladeView is distinguished in three aspects: (1) It automatically adapts wind turbines with varying geometric properties and physical locations relative to the take-off point. (2) It dramatically improves the quality of collected data with optimal UAV speed and flight corridors. (3) It thoroughly covers all three blades of a Horizontal-Axis Wind Turbine (HAWT), including regions where defects frequently occur. Thus, BladeView is more efficient and robust than existing UAV-based methods for blade inspection, with only around 25 minutes per HAWT. Moreover, it does not require experienced pilots to fly the UAV and manual interventions are rarely needed. Extensive simulation and real-world experiments demonstrate the efficiency and usability of BladeView in various on- and offshore wind farms.}, archive = {J_TASE}, author = {Cong Yang and Hua Zhou and Xun Liu and Yan Ke and Bo Gao and Marcin Grzegorzek and Zeyd Boukhers and Tao Chen and John See}, doi = {10.1109/TASE.2024.3464640}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7530--7545}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {BladeView: Toward automatic wind turbine inspection with unmanned aerial vehicle}, volume = {22}, year = {2025}, }
• (2025). Priority-based energy allocation in buildings through distributed model predictive control. TASE, 22, 7516--7529. (www)

  @article{ , abstract = {Many countries are facing energy shortages today and most of the global energy is consumed by HVAC systems in buildings. For the scenarios where the energy system is not sufficiently supplied to HVAC systems, a priority-based allocation scheme based on distributed model predictive control is proposed in this paper, which distributes the energy rationally based on priority order. According to the scenarios, two distributed allocation strategies, i.e., one-to-one priority strategy and multi-to-one priority strategy, are developed in this paper and validated by simulation in a building containing three zones and a building containing 36 rooms, respectively. Both priority-based strategies fully exploit the potential of predictive control solutions. The experiment shows that our scheme has good scalability and achieves the performance of the centralized strategy while making the calculation tractable. Note to Practitioners—The motivation of this paper is to develop a priority-based allocation strategy adapted to energy-limited systems. When energy is limited, the strategy can rationally allocate energy and satisfy the urgent need for energy supply in some specific zones. Two priority strategies are proposed for the case that a single subsystem corresponds to a particular priority and multiple subsystems correspond to the same priority, respectively. The developed strategies have been validated by co-simulation with MATLAB and EnergyPlus in a small-scale three-zone building and a large-scale 36-zone building to show their effectiveness.}, archive = {J_TASE}, author = {Hongyi Li and Jun Xu and Qianchuan Zhao}, doi = {10.1109/TASE.2024.3464612}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7516--7529}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Priority-based energy allocation in buildings through distributed model predictive control}, volume = {22}, year = {2025}, }
• (2025). Membership-function-dependent reachable set synthesis of takagi-sugeno fuzzy singular multi-agent systems subject to deception attacks. TASE, 22, 7506--7515. (www)

  @article{ , abstract = {This paper focuses on the membership-functions-dependent reachable set (RS) synthesis of a class of Takagi-Sugeno (T-S) fuzzy singular multi-agent systems subject to deception attacks. To do this, with the help of the T-S fuzzy model, the nonlinear system model is transformed into a quasi-linear system. Then, the control method in the linear system is used to solve this kind of nonlinear problem. Firstly, two suitable non-singular value matrices M and N are obtained by singular value decomposition, and then the singular system is decomposed into algebraic subsystem and differential subsystem by matrices M and N. Secondly, a consensus topology with communication delay under deception attack is designed. In this cases, the sufficient conditions of RS for the simultaneous existence of external disturbances and deception attacks are given, and the specific value of the RS is obtained by solving the linear matrix inequality using the matrix decoupling method. Finally, the effectiveness of the proposed method is verified by taking the actual DC motor built on the physical platform and a class of nonlinear single-species bio-economic system as examples.Note to Practitioners—In the context of contemporary huge industry and engineering, while rapid development, security is also one of the issues that must be considered. However, the modern engineering system has the characteristics of complexity and variability, and a single module cannot operate it completely, so multiple separate modules need to work together through a unique connection. For this reason, security issues will become much more difficult. In addition, in modern network engineering, there are often some criminals who destroy the system, resulting in its normal operation, economic losses, serious situations, or endanger personal safety. Such behavior is usually called network attacks. Deception attacks are part of network attacks, which can harm the system by spreading false information. Therefore, in researching security issues, deception attacks need to be paid attention. Based on the above analysis, the membership-functions-dependent reachable set synthesis of a class of Takagi-Sugeno fuzzy singular multi-agent systems subject to deception attacks is studied in this paper.}, archive = {J_TASE}, author = {Liang Zhang and Xudong Zhang and Xudong Zhao and Ning Zhao}, doi = {10.1109/TASE.2024.3464580}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7506--7515}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Membership-function-dependent reachable set synthesis of takagi-sugeno fuzzy singular multi-agent systems subject to deception attacks}, volume = {22}, year = {2025}, }
• (2025). Facial expression monitoring via fine-grained vision-language alignment. TASE, 22, 7492--7505. (www)

  @article{ , abstract = {In the fields of health care and clinical monitoring, vision-based Facial Expression Recognition (FER) has achieved significantly progress, but it still faces the challenge of poor generalization ability under unconstrained conditions of occlusions and pose variation. Recently, Vision-Language Model (VLM) has greatly advanced the FER task. However, the existing VLM-based FER methods typically leverage a hard-crafted prompt (e.g., “a photo of [class]”) and only focus on the holistic semantic alignment, which may suffer from modal heterogeneity. In this work, we propose a fine-grained vision-language model via Prompt Masking for FER (PMFER). Specifically, for each expression, we first create fine-grained prompts using facial action units to guide the image encoder to learn discriminative representations. Further, to finely align text prompts and visual action units, we randomly drop a phrase description in the prompts and then predict the dropped phrase by conducting modal cross attention, implicitly promoting fine-grained vision-language alignment. In addition, we also design a modal-adversarial strategy to holistically eliminate the modal difference between visual and textual embeddings in a common latent space. Experimental results demonstrate that our PMFER model outperforms the state-of-the-art methods on several FER benchmarks, especially under the conditions of occlusions and pose variations. Note to Practitioners—Facial expression recognition is very important in health care and clinical monitoring, which provides an useful tool to assess the psychological and physiological conditions of patients. Although FER has made significant progress with the development of deep learning technologies, it still faces problems in the complex environments (e.g., occlusions and pose variations). To address the above issues, we propose a novel FER method in this work based on the recent vision-language model. It takes RGB image and text prompts as input and finally predicts the expression classification. Different from the existing methods, the proposed PMFER can enable fine-grained modal alignment for facial key units. Compared with the state-of-the-art methods on the public datasets, it can achieve better results, especially under the conditions of occlusions and pose variations. Also, we evaluate the proposed method on a real-world pain dataset, and the results demonstrate that PMFER has a good generalization and can be applied to health care.}, archive = {J_TASE}, author = {Weihong Ren and Yu Gao and Xi’ai Chen and Zhi Han and Zhiyong Wang and Jiaole Wang and Honghai Liu}, doi = {10.1109/TASE.2024.3464519}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7492--7505}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Facial expression monitoring via fine-grained vision-language alignment}, volume = {22}, year = {2025}, }
• (2025). Multi-mobile-robot transport and production integrated system optimization. TASE, 22, 7480--7491. (www)

  @article{ , abstract = {A production workshop with mobile robots can be considered as a hybrid system consisting of a production system and a transportation one. Mobile robots are responsible for transferring production tasks among the machines of a production system and constitute a multi-robot transport system. It is highly coupled with a production system because of the interdependency that exists between production scheduling and mobile robot assignment. In this work, we study their integrated optimization problem for a mobile robot-based job shop with blocking properties. Its aim is to minimize total completion time as an objective function to improve overall operational efficiency. We consider the speed of a robot that varies according to whether it is loaded or not. We formulate this new problem into a mixed integer linear program to provide an algebraic description. Then, we propose a constraint programming method to solve it with high efficiency. The superiority of constraint programming over mixed integer linear programming in terms of the number of variables and constraints is analyzed. Numerous experiments on benchmark examples show that constraint programming can well handle the concerned problem. Under a one-hour time limit, it can exactly solve its instances while mixed integer linear programming cannot. Under a one-minute time limit, it obtains much better solutions than mixed integer linear programming and heuristic strategies, thus implying its high potential to be put into industrial applications. Note to Practitioners—The integration of mobile robots into production workshops has emerged as a pivotal strategy to enhance the operational efficiency of an advanced manufacturing system. This integration transforms the traditional job shop into a hybrid system, where mobile robots play a crucial role in transferring production tasks among machines. This brings a unique challenge to practitioners due to the intricate interdependencies between production scheduling and mobile robot assignment. The focus of our study is on the optimization of a mobile robot-based job shop to improve its overall operational efficiency. Our approach involves formulating this complex problem as a mixed-integer linear program, thereby providing a concise mathematical representation. We propose a constraint programming method to solve the problem efficiently. Through numerous experiments on benchmark examples, our findings indicate that the proposed constraint programming method can well solve the concerned problem given long or short solution time. This underscores its high potential for practical implementation in industrial scenarios.}, archive = {J_TASE}, author = {Ziyan Zhao and Xingyang Li and Shixin Liu and MengChu Zhou and Xiaochun Yang}, doi = {10.1109/TASE.2024.3421889}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7480--7491}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-mobile-robot transport and production integrated system optimization}, volume = {22}, year = {2025}, }
• (2025). Distributed energy-efficient flexible manufacturing with assembly and transportation: A knowledge-based bi-hierarchical optimization approach. TASE, 22, 7463--7479. (www)

  @article{ , abstract = {As manufacturing processes continue to evolve, distributed and energy-efficient flexible manufacturing systems have become the central paradigm of intelligent manufacturing. To effectively meet the demands of intricate production scheduling requirements, it is crucial to take into account practical factors, such as two-stage assembly production and finite transportation resources. However, this imposes greater demands on the efficiency of scheduling algorithms. In this paper, a knowledge-based bi-hierarchical optimization algorithm (KBOA) is specially designed for solving the distributed energy-efficient assembly flexible job shop problem with transportation constraints (DEAFJSP-T). The objectives of minimizing makespan and total energy consumption are simultaneously taken into account. The search process of the KBOA is structured into two distinct modes. In the first mode, single-objective searches are conducted along several specified directions. To elevate the overall quality of the population, knowledge-based initialization strategies are employed, tailored for different objectives. Furthermore, the knowledge-based local search is designed to strengthen the exploitation of the KBOA by leveraging the interconnections among subproblems. Incorporating the search results from the first mode, the multi-objective search is conducted based on Pareto dominance in the second stage. Considering the characteristics of the problem, individual cooperation is employed to enhance the search efficiency of the KBOA. Extensive experiments are conducted to assess the performances of the KBOA. The statistical comparisons demonstrate that the KBOA is superior in solving the DEAFJSP-T in terms of efficiency and effectiveness. Note to Practitioners—Drawing inspiration from a real-world case in an electrical machinery equipment factory, we have observed the implementation of distributed energy-efficient flexible manufacturing with assembly and transportation (DEAFJSP-T). In this practical scenario, we must concurrently optimize two conflicting objectives of time and energy consumption, taking into account the intricate coupling relationship among processing, assembly and transportation, and satisfying complex constraints. Traditional scheduling methods often fall short in directly resolving such problems. Consequently, this research holds significant practical application value and academic importance. Given the distinctive nature of the problem, we have introduced a bi-hierarchical optimization approach to reduce redundant searches while retaining critical Pareto front information. To enhance the overall quality of the population, we have developed several knowledge-based heuristics for initialization. Additionally, we have devised problem property-driven individual cooperation and knowledge-based local search to augment the exploration for non-dominated solutions and strike a balance between exploration and exploitation. Our experimental results demonstrate the superior performance of our approach, surpassing state-of-the-art algorithms. In essence, this article presents a bi-hierarchical optimization algorithm infused with problem-specific knowledge to tackle an intricate scheduling problem derived from real-world scenarios, namely DEAFJSP-T. Looking ahead, we aim to address scheduling in uncertain scenarios and explore the integration of machine learning into optimization methods.}, archive = {J_TASE}, author = {Zixiao Pan and Ling Wang and Jingjing Wang and Yang Yu and Rui Li}, doi = {10.1109/TASE.2024.3396474}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7463--7479}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed energy-efficient flexible manufacturing with assembly and transportation: A knowledge-based bi-hierarchical optimization approach}, volume = {22}, year = {2025}, }
• (2025). A novel mathematical model for the flexible job-shop scheduling problem with limited automated guided vehicles. TASE, 22, 7449--7462. (www)

  @article{ , abstract = {Automated Guided Vehicles (AGVs) have found widespread application in discrete manufacturing systems. In flexible job-shop environments, the integrated scheduling of machines and AGVs is a significant research direction to improve the productivity. However, the existing mathematical model assigns non-existent transport tasks to the corresponding AGVs, resulting in poor performance. To tackle this weakness, this paper proposes a novel mixed integer linear programming (MILP) model. Firstly, the flexible job-shop scheduling problem with limited AGVs (FJSPLA) is decomposed into four sub-problems, and the interactions and dependencies between the sub-problems are elaborated. Secondly, the existence of transport tasks is explained in detail based on the disjunctive graph model. Subsequently, a more efficient MILP model is proposed, leveraging insights from the four sub-problems and the disjunctive graph model. Finally, comparison experiments are conducted, encompassing two benchmarks (FJSPT and EX), along with a real-world case. The proposed model exhibits a more streamlined formulation with fewer decision variables and constraints in comparison to existing models. It successfully proves optimality for the most challenging instance FJSPT7 as well as 15 instances in EX benchmark. Compared with the existing model, the experimental results not only demonstrate the effectiveness and superior performance of the proposed model but also show the practicality in addressing real workshop problems. Note to Practitioners—Automated guided vehicles (AGVs) have been extensive application in various industries, prompting practitioners to integrate the scheduling of machines and AGVs during production planning. To address this realistic production problem, this study develops a novel MILP model. Through comprehensive analyses, integrated scheduling is decomposed into four sub-problems and the correlations between the four sub-problems are accurately presented. For each sub-problem, we establish the corresponding mathematical formulations. Practitioners can use the work in this paper to clearly understand the integrated scheduling problem, and can easily use the optimization software to solve the model. As in our case study, practitioners collate the production information according to their workshop, and the model can give the optimal solution for integrated scheduling in an acceptable time. The optimal solution obtained from the proposed model can guide the practitioners to maximize the productivity of the workshop.}, archive = {J_TASE}, author = {Youjie Yao and Qihao Liu and Ling Fu and Xinyu Li and Yanbin Yu and Liang Gao and Wei Zhou}, doi = {10.1109/TASE.2024.3356255}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7449--7462}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel mathematical model for the flexible job-shop scheduling problem with limited automated guided vehicles}, volume = {22}, year = {2025}, }
• (2025). A knowledge-driven cooperative coevolutionary algorithm for integrated distributed production and transportation scheduling problem. TASE, 22, 7435--7448. (www)

  @article{ , abstract = {With increasing market competition, integration between production and transportation in supply chain has been valued to improve the operational performance. Meanwhile, the distributed manufacturing has emerged as a modern paradigm, fostering flexible and intelligent development. In this paper, we address an integrated distributed production and transportation scheduling problem (IDPTSP), taking into account the production plan in distributed heterogeneous hybrid flow-shops, as well as the transportation decisions made by third-party logistics provider. First, the objective calculation is presented to minimize total costs and energy consumption during production and transportation process, and problem property is analyzed to obtain an optimal pickup time strategy. Second, a knowledge-driven cooperative coevolutionary algorithm (KCCA) is proposed, incorporating multiple problem-specific heuristics and operators based on the characteristics of the subproblems in IDPTSP. Third, Q learning assisted cooperative coevolutionary search is proposed via analyzing the interconnections among different subproblems to effectively and efficiently explore their search space. Fourth, local intensification search with multiple operators fusing prior knowledge is incorporated for low-density regions in objective space to enhance exploitation ability. Extensive experiments are carried out to test performances of the KCCA and the numerical comparisons demonstrate effectiveness of the specific designs and superiority of the KCCA over state-of-the-art algorithms in solving the IDPTSP. Note to Practitioners—Drawing inspiration from a real-life case in electronics supply, the focus on distributed manufacturing and the integration of manufacturing and logistics has garnered significant attention from managers due to the enterprises’ need to reduce intermediate stocks and enhance operational performance. The primary focus of enterprises is on reducing costs during production and transportation. Additionally, energy consumption has become a crucial consideration due to rising energy costs and limited energy supplies. However, the complexity of this practical scenario has significantly increased due to the presence of multiple subproblems and objectives. Traditional scheduling methods are unable to effectively and efficiently obtain a Pareto front with convergence and diversity. Based on the characterises of the scheduling problem, we provided a knowledge-driven cooperative coevolutionary algorithm to minimize total costs and energy consumption during production and transportation process. At production stage, factory assignment and job sequence are determined in distributed heterogeneous hybrid flow-shops, while job batching and vehicle routing are determined at transportation stage. Through the analysis of problem properties, an optimal pickup time strategy is introduced and multiple rules and search operators are provided to enhance optimization efficiency. Meanwhile, the framework of cooperative coevolutionary search offers an effective approach to addressing such complex integrated scheduling problems. From experimental comparisons, the effectiveness of the proposed algorithm is verified surpassing state-of-the-art algorithms. Therefore, practitioners can benefit from non-dominated schedules characterized by reduced costs and enhanced energy efficiency.}, archive = {J_TASE}, author = {Jingjing Wang and Ling Wang and Honggui Han}, doi = {10.1109/TASE.2024.3422473}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7435--7448}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A knowledge-driven cooperative coevolutionary algorithm for integrated distributed production and transportation scheduling problem}, volume = {22}, year = {2025}, }
• (2025). Learning-aided iterated local search algorithm for integrated order batching, picker assignment, batch sequencing, and picker routing problem. TASE, 22, 7421--7434. (www)

  @article{ , abstract = {This work tackles an integrated order batching, picker assignment, batch sequencing, and picker routing problem in warehouse environments. A Learning-Aided Iterated Local Search (LILS) is proposed to efficiently find its high-quality solutions. The main optimizer is iterated local search. A novel bidirectional long short-term memory network-embedded autoencoder, built through end-to-end unsupervised learning with an encoder and decoder, guides the search direction. To capture the implicit relationships among strongly-coupled subproblems, long short-term memory layers are incorporated in the encoder. A network-aided mutation operator is introduced to enhance global search in a low-dimensional feature space. In the decoder, low-order and high-fit subsolutions are identified and reconstructed to generate mutated offspring solutions using long short-term memory layers and a masking mechanism. To balance the exploration and exploitation of LILS, an information exchange method is developed. Numerical experiments show that LILS outperforms several existing methods in generating high-quality schedules within a reasonable time. Note to Practitioners—In a warehouse environment, an integrated optimization problem is often addressed by using heuristics due to limited computational resources. However, expedient heuristic rules tend to produce subpar results. While meta-heuristics can generate relatively better schedules, they are time-consuming, particularly for population-based algorithms that must iteratively evaluate fitness functions for numerous candidate solutions. In order to strike a balance between computational demands and solution-qualities, our approach combines machine learning techniques with meta-heuristics. Specifically, we integrate a bidirectional long short-term memory-based autoencoder into iterated local search to enhance the latter’s global optimization capability. The integration of machine learning and meta-heuristics enables efficient generation of superior schedules in a limited time. Various experimental results demonstrate that the proposed method significantly outperforms its recently-developed competitive peers, thus greatly facilitating the efficient operation of a smart warehouse.}, archive = {J_TASE}, author = {ZhengCai Cao and XinSai Lv and ChengRan Lin}, doi = {10.1109/TASE.2024.3477982}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7421--7434}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-aided iterated local search algorithm for integrated order batching, picker assignment, batch sequencing, and picker routing problem}, volume = {22}, year = {2025}, }
• (2025). Distributed heterogeneous flow shop scheduling method for dual-carbon goals. TASE, 22, 7409--7420. (www)

  @article{ , abstract = {As an important field leading the rapid development of China’s economy, industry is an important support for building a modern power, and it is also a large carbon emitter in China. Therefore, it is of key significance to promote industry to achieve the peak of carbon emissions for the realization of China’s “dual-carbon goals”. Aiming at the problem of distributed heterogeneous flow shop scheduling problem based on dual-carbon goals (DHFSP-DCGs), a novel distributed heterogeneous flow shop scheduling model was constructed to minimize the maximum completion time and total carbon emissions, and a knowledge-driven multi-objective memetic algorithm was proposed. Firstly, considering the machine characteristics of heterogeneous factories and the conflict between two optimization objectives, the encoding and decoding methods based on double sequences are designed. Secondly, a cooperative initialization strategy is proposed to generate the initial solutions with good diversity and convergence. Thirdly, according to the characteristics of distributed heterogeneous flow shop scheduling problem, a knowledge-based local search strategy is designed to improve the quality of the solution and the performance of the algorithm, and carbon reduction strategy is used to reduce the carbon emission in the production scheduling process. Finally, the effectiveness of the proposed strategy and algorithm is verified by comparative experiments. Note to Practitioners—This paper is to solve the problem of distributed heterogeneous flow shop scheduling problem based on dual-carbon goals (DHFSP-DCGs), and the methods proposed could bring many benefits to practitioners. Firstly, considering the machine characteristics of heterogeneous factories and the conflict between two optimization objectives, the encoding and decoding methods based on double sequences are designed. Secondly, a cooperative initialization strategy is proposed to generate the initial solutions with good diversity and convergence. Thirdly, according to the characteristics of distributed heterogeneous flow shop scheduling problem, a knowledge-based local search strategy is designed to improve the quality of the solution and the performance of the algorithm, and carbon reduction strategy is used to reduce the carbon emission in the production scheduling process.}, archive = {J_TASE}, author = {Xuesong Yan and Hao Zuo and Chengyu Hu and Wenyin Gong and Liang Gao}, doi = {10.1109/TASE.2024.3371940}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7409--7420}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed heterogeneous flow shop scheduling method for dual-carbon goals}, volume = {22}, year = {2025}, }
• (2025). Integrated scheduling of automated rail-mounted gantries and external trucks in U-shaped container terminals. TASE, 22, 7394--7408. (www)

  @article{ , abstract = {In this paper, the scheduling of yard cranes (YCs) and external trucks (ETs) working in a U-shaped automated container terminal yard is proposed as a new problem. This problem arises due to the characteristic of the U-shaped layout, where the ETs enter the yard through the U-shaped lanes and interact with the YCs. To formulate this problem, a three-objective optimization model is established to simultaneously schedule YCs and ETs, considering their efficiencies. Its solution is based on the nondominated sorting genetic algorithm III (NSGA-III), which is improved during initialization, crossover and mutation to make it applicable to the problem. To calculate the objective values in each iteration of the NSGA-III, a method to make the equipment interactions equal at all times is proposed. As there are special requirements for the priority of the objectives, a new method of selecting the final solution is presented, which leads to a more suitable solution. The case study involving sensitivity analysis shows that appropriate equipment quantity arrangement and task assignments have a positive impact on improving the efficiency of yard operations. Experimental results demonstrate that the proposed model and algorithms perform better. In summary, the model and algorithms proposed in this paper can effectively solve the proposed problem, which can offer a variety of options for decision-makers in an actual terminal operation environment. Note to Practitioners—The YC scheduling of a U-shaped layout presents new characteristics, and the influence of ETs cannot be ignored, posing a novel problem. Furthermore, during the operation of container terminals, extended waiting time for ETs is often caused by considerations related to ship schedules, especially in the mixed stacking mode. This can frustrate cargo owners and waste valuable ET resources. Hence, this paper designs a joint scheduling strategy for YCs and ETs in U-shaped yard layouts. The proposed strategy combines optimizing YC operation plans with maximizing ET efficiency, while ensuring efficient yard operations.}, archive = {J_TASE}, author = {Yueyi Han and Huarong Zheng and Weihao Ma and Baicheng Yan and Dongfang Ma}, doi = {10.1109/TASE.2024.3403728}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7394--7408}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated scheduling of automated rail-mounted gantries and external trucks in U-shaped container terminals}, volume = {22}, year = {2025}, }
• (2025). Collaborative optimization of multi-equipment scheduling and intersection point allocation for U-shaped automated sea-rail intermodal container terminals. TASE, 22, 7373--7393. (www)

  @article{ , abstract = {Guided by green and low-carbon policies, the construction of intelligent container terminals is imperative. Based on the emerging U-shaped automated sea-rail intermodal container terminal (SRICT), this paper focuses on the collaborative optimization of multi-equipment scheduling and intersection point allocation considering direct and indirect transshipment modes. With the objectives of maximizing the balanced workload of bilateral-cantilever automated yard cranes (AYCs) and minimizing the invalid operation cost of the yard, a bi-objective programming model is formulated to determine the handling area for bilateral-cantilever AYCs, intersection points for direct transshipment containers, and scheduling schemes for bilateral-cantilever AYCs, automated guided vehicles (AGVs), and internal trucks (ITs). We employ Q-learning (QL) and cumulative prospect theory (CPT) based on multi-objective particle swarm optimization (MOPSO) to solve the model by guiding the search and adaptively adjusting parameters. By comparing evaluation indicators, we have verified that our algorithm is superior to the benchmark algorithms. We conducted comparative experiments on various repair operators and optimization strategies. Results indicate that the proposed model and algorithm can balance the workload of AYCs, reduce energy consumption, and improve handling efficiency. Note to Practitioners—This paper addresses the transshipment connection between ships, the yard, and trains. The U-shaped automated SRICT adopted a side-operation layout with more interaction points. Furthermore, due to constraints related to laying cables and the environment, automation technology has not fully covered all functional areas at the SRICT. Therefore, the new layout and semi-automation present a challenge for transferring sea-rail intermodal containers. This paper proposes a tailored model and algorithm for emerging automated SRICTs with bilateral-cantilever AYCs. The collaborative optimization of multi-equipment scheduling (AYCs, AGVs, and ITs) and intersection point allocation is studied for the first time.}, archive = {J_TASE}, author = {Xiaohan Wang and Zhihong Jin}, doi = {10.1109/TASE.2024.3428392}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7373--7393}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Collaborative optimization of multi-equipment scheduling and intersection point allocation for U-shaped automated sea-rail intermodal container terminals}, volume = {22}, year = {2025}, }
• (2025). Automated guided vehicle scheduling problem in manufacturing workshops: An adaptive parallel evolutionary algorithm. TASE, 22, 7361--7372. (www)

  @article{ , abstract = {In the realm of scheduling problems, metaheuristics have been widely embraced as superior solutions, appreciated for their ability to generate resolutions for non-deterministic polynomial-time hard (NP-hard) problems swiftly. This paper presents a novel parallel evolutionary algorithm (PEA), which marries metaheuristics and parallel computing to amplify computer performance utilization. Four operators and a restart strategy are incorporated into the proposed PEA to bolster both its global and local search capabilities. An accelerated calculation method for two operators is proposed. The algorithm also features an adaptive method that generates sub-threads and parameters based on computer performance, along with rotation for evaluating solutions. A random search sub-thread is established to update the solution. The algorithm is tested on the workshop automated guided vehicle (AGV) scheduling problem and compared against other optimization algorithms to ascertain its efficacy. The test results overwhelmingly highlight the superior performance of the proposed algorithm. Note to Practitioners—The paper introduces a novel parallel evolutionary algorithm (PEA) for scheduling problems, which combines metaheuristics and parallel computing to enhance computer performance utilization. The algorithm incorporates four operators and a restart strategy, along with an accelerated calculation method for two operators. It also includes an adaptive method to generate sub-threads and parameters based on computer performance, as well as rotation for evaluating solutions. A random search sub-thread is established to update the solution. The proposed algorithm is tested on the workshop automated guided vehicle (AGV) scheduling problem, producing superior results compared to other optimization algorithms. Its ability to swiftly generate resolutions for NP-hard problems can greatly benefit industries that rely on efficient scheduling, such as logistics and manufacturing. However, it is important to note that the algorithm has some limitations. Further research is needed to explore its application in different domains and evaluate its performance in more complex scheduling scenarios. Additionally, the algorithm’s scalability and adaptability need to be thoroughly examined to ensure its practicality in real-world settings.}, archive = {J_TASE}, author = {Zhongkai Li and Quanke Pan and Zhonghua Miao and Hongyan Sang and Weimin Li}, doi = {10.1109/TASE.2024.3419848}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7361--7372}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated guided vehicle scheduling problem in manufacturing workshops: An adaptive parallel evolutionary algorithm}, volume = {22}, year = {2025}, }
• (2025). Optimal assignment and scheduling of cranes in slab yard for iron and steel production enterprises. TASE, 22, 7349--7360. (www)

  @article{ , abstract = {Slab yards serve as temporary slab storage between a continuous casting stage and a rolling stage. Considering non-crossing and safe clearance constraints of slab yard cranes, this work studies a multi-crane assignment and scheduling problem in the slab yard. An mixed-integer linear programming (MILP) is formulated to minimize the slab completion time. Due to its NP-hardness, the problem for large-sized instances is computationally intractable. Thus, we develop a logic-based benders decomposition algorithm (LBBD) to solve it. First, we exploit a generalized decomposition of this problem into a relaxed main problem (RMP) and a sub-problem (SP). Solving the former allocates slabs to each crane. Then, the sequence of the assigned slabs can be found by solving its corresponding sub-problem. Finally, to verify the effectiveness of LBBD, we identify a lower bound (LB) of the optimal objective function. The problem instances on real data from an iron and steel plant are created. The result of LBBD is close to such lower bound and can be found efficiently. Note to Practitioners—This work deals with a crane assignment problem with multiple cranes for handling input slabs in a slab yard. This problem is formulated as an MILP model to minimize the completion time. Its time complexity grows exponentially with the problem size. Thus, we develop a LBBD to solve it. The numerical results reveal that LBBD can find the optimal or near-optimal solution for all realistic instances in affordable computational time. Its use can ensure the high utilization of cranes and efficient service in iron and steel plants.}, archive = {J_TASE}, author = {Xu Wang and MengChu Zhou and QiuHong Zhao and ShiXin Liu and XiWang Guo and Liang Qi and Aiiad Albeshri}, doi = {10.1109/TASE.2024.3490039}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7349--7360}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal assignment and scheduling of cranes in slab yard for iron and steel production enterprises}, volume = {22}, year = {2025}, }
• (2025). Conflict-free routing of twin reclaimers in the stockyard based on a time-space network model. TASE, 22, 7335--7348. (www)

  @article{ , abstract = {In the stockyard of dry bulk materials such as coal and iron ore, multiple stacker-reclaimers operate collaboratively to enhance stockyard productivity. However, effectively planning the routes for these interconnected machines presents a significant challenge. We propose a new modeling framework for this conflict-free routing problem in which the reclaiming process is modeled in a so-called time-space network (TSN) framework. The formulated optimization problem is mixed integer programming (MIP), which has been proven NP-hard. To address its computational efficiency, we develop a two-level metaheuristic algorithm to simplify the encoding complexity of the original problem. In the developed two-level algorithm, a task priority ordering candidate is listed at the top level, while the detailed conflict-free routes are constructed at the bottom level using an insertion-based conflict-free reclaiming and customized route improvement. The developed metaheuristic is tested on a large number of instances in comparison with the Gurobi solver and three commonly used methods for such a routing problem. The experimental results show that the proposed algorithm outperforms the other four methods regarding the solution quality and the computation time. Note to Practitioners—This study is motivated by the need for efficient route planning of interconnected reclaimers in the stockyard of dry bulk materials such as coal and iron ore. To address this problem, we present a novel modeling framework for addressing the conflict-free routing problem, wherein the reclaiming process is intricately modeled within a so-called time-space network framework. Solving the resulting mixed-integer programming is of high computational complexity; therefore, we have engineered a two-level metaheuristic framework to reduce its complexity. The developed methodology has been extensively tested on instances featuring real-world stockyard operations and contributes to terminal operators by improving productivity and resulting in more economic benefits. Future research will integrate real-time data and dynamic optimization techniques to facilitate routing decisions that can adapt to evolving operational conditions and stockyard configurations.}, archive = {J_TASE}, author = {Jianbin Xin and Chang Liu and Andrea D’Ariano and Shi Qiang Liu and Jing Liang}, doi = {10.1109/TASE.2024.3454226}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7335--7348}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Conflict-free routing of twin reclaimers in the stockyard based on a time-space network model}, volume = {22}, year = {2025}, }
• (2025). Breaking the limit on the number of robots through the conflict-free scheduling in robotic mobile fulfillment systems. TASE, 22, 7324--7334. (www)

  @article{ , abstract = {The robotic mobile fulfillment system (RMFS) utilizes robots to automate logistics handling. Although robots can replace workers, there are serious limitations and conflicts that make the system inefficient, resulting in a limited number of robots in the system. A conflict-free scheduling approach is studied, demonstrating better system efficiency, breaking the limit on the number of robots, and accommodating more robots in operation compared to the traditional strategy. A mathematical model aiming to minimize the total completion time is presented. An A* algorithm is modified for path planning, and an approach with two avoidance rules and six priority rules is proposed for conflict-free scheduling. Based on a case study involving four layout scales, the conflict-free scheduling is suitable for layouts with a higher number of robots, pillars, and a unidirectional path networks. More robots can be accommodated within the system, and even a threefold increase (45 to 150) will not result in a decrease in system efficiency, reducing the total completion time by 49.3%. In the largest layout system, the number of robots doubled (160 to 320) and the total completion time was reduced by 26.1%. This indicates the great potential for the conflict-free scheduling approach in future logistic applications, which could reduce resource consumption for scheduling and increase system efficiency. Note to Practitioners—This paper was motivated by the problem of multi-robot scheduling in robotic mobile fulfillment systems. It can be scalable to other layouts of multi-robot systems. The maximum permissible floor space for fire zoning in the fire regulations. Conflicts can easily occur if a large number of robots are operating in a limited space. Resulting in a limited number of robots in the system. This paper proposes a novel conflict-free scheduling to break the limit on the number of robots. The approach includes how to plan the path, which avoidance rules and priority rules to utilize. We mathematically model the path planning and conflict-free for multi-robot complete all task. We show the effectiveness of conflict-free scheduling in breaking the limit on the number of robots under different layouts. We define the ratio of the number of QR code squares to the number of robots, which is used to evaluate scheduling methods in various layout scenarios. The largest one has 3,880 QR codes, which is the maximum allowable floor space for single-story e-commerce warehouses. We doubled the number of robots and reduced the total completion time by 26.1%. In future research, we will consider more rules and several expansions of the layout.}, archive = {J_TASE}, author = {Yangjun Sun and Ning Zhao and Longwei Tang and Lei Luo}, doi = {10.1109/TASE.2024.3420107}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7324--7334}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Breaking the limit on the number of robots through the conflict-free scheduling in robotic mobile fulfillment systems}, volume = {22}, year = {2025}, }
• (2025). Designing liveness-enforcing supervisors for manufacturing systems by using maximally good step graphs of petri nets. TASE, 22, 7312--7323. (www)

  @article{ , abstract = {Many deadlock control methods rely on reachability graphs of Petri nets (PN), thus suffering from state-space explosion issues. This paper proposes a novel approach to designing liveness-enforcing supervisors for PN by constructing its maximally good step graphs (MGSG), a class of partial order techniques in mitigating the aforementioned issues. Specifically, we first categorize the MGSG markings into allowed and unallowed ones. Then, we define good and risky transitions at allowed markings. Through the execution of risky transitions, the initial marking cannot be reachable from the allowed ones. Next, we design a maximal number of risky transitions (MNRT) problem to compute control places. In MNRT, all allowed markings and the firing of all good transitions are permitted, while risky transitions are forbidden. The objective is to maximize the prevention of risky transitions by using a single control place, which can be achieved by solving an integer linear programming problem. MNRT problems are recursively solved for unforbidden risky transitions until their resulting markings are prohibited. Finally, a controlled PN is generated and has been demonstrated to retain liveness. The experimental results show that our approach effectively reduces the number of control places and mitigates state-space explosion issues over its state-of-the-art peers. Note to Practitioners—Raw materials are handled in manufacturing systems through a series of well-coordinated processes. However, multiple processes competing for the same resources (like machines or robots) may result in an undesirable deadlock situation, where production slows down or entirely halts. A liveness-enforcing supervisor is designed to counter such a situation. As a mathematical tool, PN can be applied to model manufacturing systems. The supervisor design involves adding control places and directed arcs into PN to ensure its liveness, which implies that the systems can execute their tasks and never enter into deadlocks. In this work, we develop a deadlock control approach, which can generate a live and controlled PN without exploring a complete state space of PN. Experimental results show the effectiveness of our approach in reducing the number of control places and its practicability applied to large-scale systems, thus surpassing state-of-the-art methods.}, archive = {J_TASE}, author = {Hao Dou and Dan You and Shouguang Wang and MengChu Zhou}, doi = {10.1109/TASE.2024.3450656}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7312--7323}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Designing liveness-enforcing supervisors for manufacturing systems by using maximally good step graphs of petri nets}, volume = {22}, year = {2025}, }
• (2025). An intelligent multi-layer control architecture for logistics operations of autonomous vehicles in manufacturing systems. TASE, 22, 7296--7311. (www)

  @article{ , abstract = {In this paper, autonomous vehicles are considered for addressing logistic operations in manufacturing systems. The starting idea consists in organizing a given group of autonomous robots/vehicles in a finite set of platoons in charge to accomplish prescribed job(s) within the manufacturing system. Three aspects are then needed to be formally outlined: task scheduling, routing decisions and command inputs computations. Here, a new distributed multi-layer architecture has been conceived by using three methodologies: timed colored Petri nets, deep reinforcement learning and model predictive control. Roughly speaking, timed colored Petri nets are exploited to formally model the manufacturing system so that an optimal scheduling task complying with the required jobs and the available vehicles is derived; then, run-time routing decisions are obtained by using a distributed reinforcement learning algorithm which exploits the available information provided by the vehicle sensor module; finally, the distributed model predictive control algorithm is built by resorting to a set-theoretic approach where most of the computations are off-line performed. A flexible manufacturing system consisting of four machines and a Load/Unload station is used for simulation purposes. Specifically, five jobs are hypothesized and some scenarios with an increasing number of available vehicles are simulated. In order to evaluate the benefits of the proposed approach, a time criterion based on the completion of all the jobs is considered with the aim to put in light that increasing the number of vehicles improves the control performance until congestion phenomena become unavoidable. Note to Practitioners—This paper proposes an innovative methodology for addressing the logistic operations within flexible manufacturing systems (FMSs) by using a team of autonomous ground vehicles. Unlike existing approaches, the core of this framework consists in combining along a hierarchical structure the capabilities of timed colored Petri nets and the deep reinforcement learning techniques to determine a near-optimal scheduling and run-time routing decisions that are provided to the distributed model predictive units in charge to accomplish the prescribed task. This multi-layer architecture has two main merits: a single platoon, completely disconnected from the others, is devoted to perform its own job; computational burdens are affordable during the on-line operations because most of the computations are moved in the off-line phase.}, archive = {J_TASE}, author = {Domenico Famularo and Giancarlo Fortino and Francesco Pupo and Francesco Giannini and Giuseppe Franzè}, doi = {10.1109/TASE.2024.3435342}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7296--7311}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An intelligent multi-layer control architecture for logistics operations of autonomous vehicles in manufacturing systems}, volume = {22}, year = {2025}, }
• (2025). A matheuristics for the configuration of automated vertical lift modules warehouses. TASE, 22, 7284--7295. (www)

  @article{ , abstract = {The design of the layout of Vertical Lift Module (VLM) warehouses is a non-trivial process that involves selecting dimensions, internal configuration, and allocation of each tray to avoid space loss while satisfying logistic constraints. Our contribution in this context is a two-phase matheuristics –an algorithm that combines exact mathematical methods and heuristics– to simplify the design of VLMs layout. The proposed matheuristics relies on three Mixed-Integer Linear Programming models, addressing the internal configuration of trays and the allocation of trays into columns based on industrial logistic constraints. This approach requires as input parameters the items features, predetermined tray types with different dimensions, matheuristic settings, and a priority rule for tray allocation. The algorithm outputs to the logistics operator types and quantities of trays needed, internal partitioning, item positions in each tray, and tray positions in each column. Extensive testing demonstrates the effectiveness of our approach under realistic scenarios. Additionally, we introduce a comprehensive set of priority rules for allocating trays into columns, providing a comparison to assist logistics operators in selecting the most suitable for specific scenarios. Note to Practitioners—VLMs are closed structures composed of columns that house a variable number of sliding trays and a lift-mounted module that handles the trays. Their operation revolves around the so-called “goods to the man” principle, where goods are automatically brought to the operator using a dedicated access bay. This design ensures that stored items are easily accessible to operators, reducing time for order picking and improving workplace safety and ergonomics. Designing a VLM for logistics companies is a complex and time-consuming task since it requires optimizing a target objective while satisfying a large set of constraints. The current manual approach lacks automated methods and relies on experienced operators and iterative improvement processes. Our two-phase matheuristic algorithm automates VLMs layout configuration, addressing tasks from item placement to tray allocation within columns. The algorithm is versatile, considering various practical and logistic constraints, making it applicable in warehouse design decision support systems or engineering software.}, archive = {J_TASE}, author = {Giulia Tresca and Graziana Cavone and Paolo Scarabaggio and Raffaele Carli and Mariagrazia Dotoli}, doi = {10.1109/TASE.2024.3490986}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7284--7295}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A matheuristics for the configuration of automated vertical lift modules warehouses}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Smart coordination for logistics operational control in manufacturing under the evolution trend of digital economy. TASE, 22, 7281--7283. (www)

  @article{ , archive = {J_TASE}, author = {Wenfeng Li and Mariagrazia Dotoli and Xiaoou Li and Walter Lucia and Jianbin Xin}, doi = {10.1109/TASE.2025.3544002}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7281--7283}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Smart coordination for logistics operational control in manufacturing under the evolution trend of digital economy}, volume = {22}, year = {2025}, }
• (2025). Quality-related spatio-temporal information analytics-based multiunit synergetic monitoring for plant-wide industrial processes. TASE, 22, 7268--7280. (www)

  @article{ , abstract = {Modern industrial plants generally demonstrate the characteristics of large scale, long process, and multiunit collaborative operation, which makes the spatio-temporal distribution an inherent nature, and the quality stability is usually hard to be guaranteed. A quality-related spatio-temporal information analytics based multiunit synergetic monitoring framework is presented in this paper. In this framework, the spatio-temporal properties are analyzed from the unit level and the process level, respectively. Firstly, for each operation unit, the quality supervised spatio-temporal support region is constructed with a concurrent feature extraction strategy. In this strategy, temporal dynamic features are extracted by a long short term memory (LSTM) network with attention mechanism. Concurrently, the spatial feature is extracted with the mutual information-kernel principal component analysis approach. Secondly, for the plant-wide process, a third order multiunit-spatio-temporal feature tensor is constructed for feature fusions. Via tensor decomposition, the interconnected associations among units and the quality inheritance along the process are explored, and the original feature space is decomposed into several subspaces. Finally, a multiunit synergetic monitoring model is developed over subspaces and the comprehensive monitoring results are given by Bayesian fusion. Reasonable interpretations can be provided in the monitoring results. The effectiveness of the proposed framework is verified on a real hot strip mill process.Note to Practitioners—This paper intends to provide a spatio-temporal information analytics and fusion framework for multiunit processes and to develop a quality-related process monitoring method for industrial plants. Different from the existing works, the monitoring model built in this paper is based on the multiunit-spatio-temporal sensitive information, which is extracted by a concurrent strategy and fused by the tensor model. In addition, the quality inheritance among multiunits is considered in this framework and the monitoring results in each subspace can provide helpful instructions for the field technicians. In detail, the fault-relevant unit can be located by the relatively independent subspace monitoring, and the quality-related anomaly propagation tendency can be indicated by the strong associative subspace monitoring.}, archive = {J_TASE}, author = {Chi Zhang and Jie Dong and Kaixiang Peng and Hanwen Zhang}, doi = {10.1109/TASE.2024.3463650}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7268--7280}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quality-related spatio-temporal information analytics-based multiunit synergetic monitoring for plant-wide industrial processes}, volume = {22}, year = {2025}, }
• (2025). Toward improving dynamic resource scheduling in the context of digital twin of emergency department. TASE, 22, 7255--7267. (www)

  @article{ , abstract = {The Emergency Department (ED) is a principal sector in the hospital that has to make crucial decisions effectively under uncertainties to ensure that the patients receive good quality care. Improving decision-making in ED, especially the resource scheduling decision, becomes one research interest. A promising approach to improve decision-making is by utilising Digital Twin (DT). This context of decision-making in DT is both (i) an opportunity given by the DT which provides data close to real-time and (ii) constrained by the DT which forces to make decisions dynamically rather than statically. That is, our research contributes to the decision layer of DT. More precisely, we propose a decentralised organisation to enhance the resource scheduling decision-making process in ED. This is a multi-agent system with a local dynamic resource scheduling based on a Mixed Integer Linear Programming (MILP) model. This proposition is compared to two other organisations, which are a centralised organisation that optimises the dynamic scheduling MILP globally, and the FIFO sequencing currently performed in the paediatric ED of the University Hospital Centre in Saint-Étienne, France. The results show that the proposed decentralised organisation provides good results quickly and seems promising for improving resource scheduling decision-making in ED with DT support. Note to Practitioners—This research provides an initial study on the approach to improve the decision-making process in the Emergency Department (ED) of the hospital by using Digital Twin (DT). We focus more on improving resource scheduling decisions, which is a part of the decision layer of DT. We propose a decentralised organisation based on a multi-agent approach for solving the resource scheduling problem. Each resource agent, i.e. nurse, can run a local dynamic resource scheduling based on a Mixed Integer Linear Programming (MILP) model. Through our proposition with a centralised model and current practice in ED based on the triage-based First in First Out (FIFO) rule, we can see that our decentralised organisation is a favourable decision organisation in the context of decision improvement with DT, as it is capable of providing favourable results in a short time. This research will be further extended by including different resources and tasks in ED to make the model closer to reality in ED.}, archive = {J_TASE}, author = {Jessica Florencia and Thierry Moyaux and Lorraine Trilling and Guillaume Bouleux and Vincent Cheutet}, doi = {10.1109/TASE.2024.3463489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7255--7267}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward improving dynamic resource scheduling in the context of digital twin of emergency department}, volume = {22}, year = {2025}, }
• (2025). Simultaneous position-and-stiffness control of underactuated antagonistic tendon-driven continuum robots. TASE, 22, 7238--7254. (www)

  @article{ , abstract = {Continuum robots have gained widespread popularity due to their inherent compliance and flexibility, particularly their adjustable levels of stiffness for various application scenarios. Despite efforts to dynamic modeling and control synthesis over the past decade, few studies have incorporated stiffness regulation into their feedback control design; however, this is one of the initial motivations to develop continuum robots. This paper addresses the crucial challenge of controlling both the position and stiffness of underactuated continuum robots actuated by antagonistic tendons. We begin by presenting a rigid-link dynamical model that can analyze the open-loop stiffening of tendon-driven continuum robots. Based on this model, we propose a novel passivity-based position-and-stiffness controller that adheres to the non-negative tension constraint. Comprehensive experiments on our continuum robot validate the theoretical results and demonstrate the efficacy and precision of this approach. Note to Practitioners—This paper is motivated by our experience and practical needs in building continuum robotic platforms. Stiffness, flexibility, and accurate configuration regulation are practically important properties for this class of robots. Even though simultaneous position-and-stiffness control is a mature topic for rigid and softly-actuated robots, it remains an open problem for continuum robots with theoretically guaranteed performance. The intention of this paper is to address this situation by proposing a model-based solution for a class of underactuated tendon-driven continuum robots. Hence, we propose an energy-based model and a passivity-based controller to regulate stiffness and configuration concurrently. We believe that this work can be beneficial for academic and industrial research in the context of control algorithms for continuum robots.}, archive = {J_TASE}, author = {Bowen Yi and Yeman Fan and Dikai Liu and Jose Guadalupe Romero}, doi = {10.1109/TASE.2024.3462958}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7238--7254}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Simultaneous position-and-stiffness control of underactuated antagonistic tendon-driven continuum robots}, volume = {22}, year = {2025}, }
• (2025). Disturbance observer-based finite-time control algorithm for robotic bolt-tightening via visual feedback. TASE, 22, 7226--7237. (www)

  @article{ , abstract = {The increasingly mature visual feedback technology has greatly promoted the development of high-altitude unmanned operations, particularly in bolt-tightening operations for the power transmission tower. Nevertheless, traditional control algorithms relying on visual feedback encounter specific difficulties and challenges during bolt-tightening tasks execution. This paper presents a disturbance observer-based finite-time control algorithm to reject disturbances during bolt-tightening operations and mechanical errors inherent to the robotic system. Through theoretical analysis of this algorithm, based on visual feedback technology and the finite-time control method, the finite-time stability of the overall closed-loop system is ensured. Both the simulation and experimental results are given to validate the feasibility of the proposed control method. Compared with traditional PID, LQR, and MPC controllers, the convergence speed is increased by more than 16%, and the accuracy of steady-state error is also increased by more than 22%, moreover, the success rate is as high as 98%. It is highlighted that the proposed control method has superior disturbance rejection capability, faster convergence speed, and better reliability. Note to Practitioners—This study addresses challenges in visual feedback-dependent bolt-tightening operations for power transmission towers. A novel disturbance observer-based finite-time control algorithm is introduced, ensuring the stability of the closed-loop system. The method excels in rejecting disturbances and minimizing mechanical errors during bolt-tightening tasks. The simulation and experimental results confirm its superiority over conventional controllers in terms of disturbance rejection, convergence rate, and operational precision. Practitioners should consider the algorithm’s potential to enhance the efficiency and reliability of bolt-tightening in high-altitude unmanned operations.}, archive = {J_TASE}, author = {Junyi You and Haibo Du and Chih-Chiang Chen and Yansheng Liu and Qing Du}, doi = {10.1109/TASE.2024.3462468}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7226--7237}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Disturbance observer-based finite-time control algorithm for robotic bolt-tightening via visual feedback}, volume = {22}, year = {2025}, }
• (2025). An optimal control scheme for grid-connected voltage source inverter via grid voltage modulated-direct power control. TASE, 22, 7216--7225. (www)

  @article{ , abstract = {In this paper, we propose a linear quadratic regulator (LQR) for a kind of three-phase two-level voltage source inverter on the basis of grid voltage modulated-direct power control (GVM-DPC) principle. The proposed method has a similar control structure to the conventional GVM-DPC scheme, but the algorithm parameters tuning problem therein could be solved by LQR. Considering the inverter system often converges to a nonzero equilibrium point, an auxiliary variable corresponding to the tracking error is necessary. Unfortunately, when utilizing the LQR technology to solve the optimal tracking problem directly, the dynamic response and steady-state performance might be poor in some situations. By developing a novel auxiliary variable, the drawbacks aforementioned can be avoided and the optimal feedback gain matrix only needs to be calculated once during the control process offline. Then, a Lyapunov stability analysis is used to prove the exponential stability of the system. Finally, simulation and experimental results validate the performance of the proposed method.Note to Practitioners—With the high integration of renewable energy in modern power systems, inverters have become the core device for the transition among electricity, wind, and solar energy. This trend makes the research of power converters regarding control problems increasingly important. Although existing studies can solve some basic problems such as controlling the output current or power of inverters, most of them are based on phase-locked loops, and controller parameters are not easy to tune. This motivates us to develop a user-friendly, low computational burden, easily-to-tune parameters control algorithm, which is convenient to implement on embedded devices to control the output power of inverters. Inspired by this, a direct power control method for inverters using linear quadratic regulator is proposed in this study, and a phase-locked loop-less method is developed by utilizing the theory of grid voltage modulation. Furthermore, this method establishes a new auxiliary variable to ensure global exponential stability of the linear quadratic regulator, resulting in satisfactory transient performance. The above advantages have been further validated through experiments and compared with existing algorithms.}, archive = {J_TASE}, author = {Hao Shen and Jiawei Xu and Qianjin Zhang and Ju H. Park and Chen Peng}, doi = {10.1109/TASE.2024.3462464}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7216--7225}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An optimal control scheme for grid-connected voltage source inverter via grid voltage modulated-direct power control}, volume = {22}, year = {2025}, }
• (2025). Force feedback event triggering-based tracking control for wheeled mobile robots. TASE, 22, 7205--7215. (www)

  @article{ , abstract = {In soft deformable terrain environments, the robot slips due to dynamic changes in wheel-ground contact, which poses a great challenge to the design of the driving torque of its motion control system. To solve the trajectory tracking control problem of wheeled mobile robots in soft deformation terrain, an event triggering mechanism based on wheel-ground mechanical parameters was designed, in which wheel-terrain mechanics has an important influence on the driving torque and is included in the control system design process. Aiming at the wheeled mobile robot in the working environment of soft ground, considering the rolling resistance of the wheel during its driving process, a dynamic model based on wheel-ground interaction is established. Estimation of unmodelled dynamic and rolling resistance terms for wheeled mobile robots in soft deformable terrain environments by adaptive neural networks. Based on the static event triggering strategy based on constant threshold, a hybrid threshold dynamic event triggering strategy based on rolling resistance is proposed. By proving that there is a positive lower bound on the inter-event time, which means that Zeno behavior is avoided. Meanwhile, the lower bound of inter-event time will change with the designed dynamic threshold. Finally, the good control performance of the proposed algorithm under different ground environments is verified by simulation. Note to Practitioners—With the advancement of detection tasks, the working environment of wheeled mobile robots has become increasingly complex. In the motion control of a wheeled mobile robot in a soft deformable terrain working environment, the influence of the robot ’s wheel-to-ground contact is crucial to the successful realization of the task. The existing wheeled mobile robot control methods for soft deformable terrain working environment usually ignores the influence between wheels and ground, which cannot meet the application requirements of this complex scene. Aiming at the problem of tracking control of wheeled mobile robots in soft deformable terrain working environment, this paper, the traction force change caused by wheel-ground contact mechanics is taken as the main factor of event-triggered mechanism, and the force feedback event-triggered tracking control method is designed. Theoretical algorithms and simulation results show that a trade-off between robot tracking performance and communication resources in different ground environments is realized.}, archive = {J_TASE}, author = {Shu Li and Tao Ren and Yan-Jun Liu and Lei Liu and Feng Wan}, doi = {10.1109/TASE.2024.3461953}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7205--7215}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Force feedback event triggering-based tracking control for wheeled mobile robots}, volume = {22}, year = {2025}, }
• (2025). An improved trajectory tracking mechanism with adaptive critic for event-based multiplayer zero-sum games. TASE, 22, 7193--7204. (www)

  @article{ , abstract = {In this paper, based on the adaptive critic control method, an improved event-based trajectory tracking mechanism of continuous-time (CT) nonlinear multiplayer zero-sum games (MZSGs) is established. It is worthy of note that previous papers studying the trajectory tracking issue of nonlinear CT MZSGs only apply to the case where the reference trajectory eventually converges to zero. Consequently, this paper develops an improved mechanism to overcome this weakness. Later, an event-triggered framework is brought in to reduce the amount of computation and improve control efficiency. In this process, an innovative triggering condition is provided. At the same time, the infamous Zeno behavior is ruled out through theoretical analysis. Furthermore, the event-based near-optimal controls and event-based near-worst disturbances for tracking error dynamics are gained by building and adjusting a single critic neural network. Immediately after, by utilizing the Lyapunov method, the uniform ultimate boundedness stability of the tracking error and the weight estimation error is ensured. Lastly, an example containing two case studies is offered to validate the validity of the established mechanism. Note to Practitioners—Complex industrial processes often involve multiple control inputs and may also be affected by multiple disturbances at the same time, which can be referred to as a MZSG. Since many industrial processes can be viewed as a tracking question of nonlinear systems and the event-triggered mechanism can decrease the computational cost, the tracking problem for event-based nonlinear MZSGs is studied in this paper, which is significant for control practitioners. Moreover, the Hamilton-Jacobi-Isaacs equation is often difficult to solve when dealing with the game problem. Hence, an adaptive critic technique is presented to acquire the near-optimal controls and the near-worst disturbances, which replaces the traditional actor-critic framework and thus simplifies the theoretical analysis. Note that this paper proposes an innovative triggering condition to relax the restriction on the choice of disturbance rejection level. Compared to previous works dealing with the tracking problem of nonlinear MZSGs, the method presented in this paper makes the choice of the reference trajectory more flexible and thus enhances the applicability in general industrial processes. Finally, stability analysis and simulation results are given. Note that for different practical situations, practitioners can adjust the related parameters to achieve the tracking control of MZSGs and minimize the computational cost.}, archive = {J_TASE}, author = {Menghua Li and Ding Wang and Junfei Qiao}, doi = {10.1109/TASE.2024.3461873}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7193--7204}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An improved trajectory tracking mechanism with adaptive critic for event-based multiplayer zero-sum games}, volume = {22}, year = {2025}, }
• (2025). Quasi-synchronization of heterogeneous fractional-order dynamical networks with time-varying couplings. TASE, 22, 7181--7192. (www)

  @article{ , abstract = {This paper addresses the problem of quasi-synchronization for a class of heterogeneous fractional-order dynamical networks with time-varying couplings. Our proposed approach, called delay-dependent hybrid impulsive control, considers the network’s topology and utilizes Lyapunov functions for synchronization analysis. We employ a graph-theoretic method to construct these Lyapunov functions, including a novel time-varying graph-theoretic Lyapunov function (TGLF) that changes with the network’s structure and state variables. However, applying fractional derivatives to the TGLF poses challenges, as the general Leibniz formula and chain rule do not apply. To overcome this obstacle, we establish a new fractional derivative rule for time-varying-coefficient convex functions. We also introduce a new lemma for a fractional-order delayed impulsive differential inequality to analyze the quasi-stability and instability of impulse-free systems. By combining the TGLF and the fractional-order delayed impulsive differential inequality, we derive criteria for quasi-synchronization and estimate the allowable error bounds. Finally, we provide numerical examples to demonstrate the effectiveness of our proposed approach.Note to Practitioners—Most quasi-synchronization results of complex dynamical networks are mainly supported by the traditional quadratic Lyapunov function or the graph-theoretic Lyapunov function. However, these functions are only effective for networks with fixed couplings, which limits the application scope of the Lyapunov method. Aiming at the case of time-varying couplings, this paper proposes a novel time-varying graph-theoretic Lyapunov function that changes with the network’s structure and state variables. By applying the new fractional derivative rule of this function, we shall obtain quasi-synchronization criteria of heterogeneous fractional-order complex dynamical networks with time-varying couplings by using delay-dependent hybrid impulsive control. To verify the application of theoretical results, a generalization of the fractional-order power system is provided, and we realize its quasi-synchronization theoretically and numerically.}, archive = {J_TASE}, author = {Yao Xu and Haodong Wang and Xinzhi Liu and Wenxue Li}, doi = {10.1109/TASE.2024.3461809}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7181--7192}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quasi-synchronization of heterogeneous fractional-order dynamical networks with time-varying couplings}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered sliding-mode bipartite consensus for multi-agent systems with unknown dynamics. TASE, 22, 7170--7180. (www)

  @article{ , abstract = {This paper addresses a data-driven sliding mode bipartite consensus issue for nonlinear discrete-time multi-agent systems with antagonistic interactions and limited communication resources. Initially, the signed graph theory is employed, and a combined measurement error function is formulated, transforming the bipartite consensus issue into a traditional consensus issue. An enhanced compact form dynamic linearization model is then established based on the input/output data and the formulated combined measurement error function. Moreover, a dynamic event-triggered function and a sliding-mode surface are designed, leading to the development of a fully distributed dynamic event-triggered sliding-mode bipartite consensus (DET-SMBC) approach. The proposed DET-SMBC approach is subsequently extended to a fully distributed dynamic event-triggered robust sliding-mode bipartite consensus (DET-RSMBC) scheme to improve robustness. The convergences of the tracking errors of both methods are rigorously deduced. Finally, simulation studies and hardware experiments are conducted to demonstrate the effectiveness of the proposed methods. Note to Practitioners—In multi-agent systems, the applicability of existing methods can be reduced by some issues, such as uncertain dynamics models, unknown disturbances, and the limitation of communication bandwidth. These issues can influence existing methods’ usefulness and cause instability, so DET-SMBC and DET-RSMBC methods are proposed in this paper. Compared with existing results, identifying a precise dynamics model for each controlled plant is unnecessary, the necessity of high-performance hardware for data transmission is relieved, and the effects of unknown disturbances are reduced. Moreover, the proposed methods are applied to realistic servo motor systems to conduct speed bipartite consensus tasks well. It is noted that most complicated mechanisms are controlled by servo motors, so the proposed methods can be applied to more practical engineering systems.}, archive = {J_TASE}, author = {Huarong Zhao and Li Peng and Linbo Xie and Hongnian Yu}, doi = {10.1109/TASE.2024.3461806}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7170--7180}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered sliding-mode bipartite consensus for multi-agent systems with unknown dynamics}, volume = {22}, year = {2025}, }
• (2025). Data-driven distributed iterative learning control for multiple HSTs under independent weighted communication and input saturation. TASE, 22, 7159--7169. (www)

  @article{ , abstract = {Cooperative control of multiple high-speed trains (MHSTs) is considered as a pivotal technology for enhancing the safety and efficiency of train groups. This study investigates the problem of distributed iterative learning control for MHSTs under the input saturation, employing a data-driven control strategy. Moreover, a novel independent weights-based communication protocol is developed to improve the coordination efficiency of MHSTs. Firstly, a novel linearization method is utilized to transform the dynamic model of MHSTs into an equivalent linearized model associated with saturated inputs. Secondly, based on this model, we develop both a train-to-train (T2T) communication protocol and a distributed model free adaptive iterative learning control (DMFAILC) scheme that are independent of any specific models or structural information about MHSTs. Subsequently, the overall stability is analyzed using the compression mapping method, and the complete tracking of HSTs and the coordination performance of MHSTs are confirmed through Simulink simulations and real-time StarSim hardware-in-the-loop (HIL) semi-physical platforms. Note to Practitioners—The practical challenges that drive this research on cooperative control of MHSTs are the intricate nature of train models and the inherent instability in train operating conditions. Moreover, potential applications include automatic operation of HSTs and coordination of MHSTs. Specifically, the contributions of this work include: (1) eliminating the need for precise models of HSTs; (2) establishing stability conditions for HSTs under saturation conditions; and (3) enhancing coordination efficiency among MHSTs through an independent weight protocol. However, a limitation of this study is that the theoretical findings have not been validated on real HSTs. To address the current lack of practical implementation of the train cooperative control strategy, further investigations will be conducted to incorporate more realistic constraints, such as those related to traction network limitations and optimal scheduling considerations.}, archive = {J_TASE}, author = {Wei Yu and Deqing Huang}, doi = {10.1109/TASE.2024.3461780}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7159--7169}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven distributed iterative learning control for multiple HSTs under independent weighted communication and input saturation}, volume = {22}, year = {2025}, }
• (2025). Cooperative moving target fencing control for two-layer UAVs with relative measurements. TASE, 22, 7145--7158. (www)

  @article{ , abstract = {This paper investigates a two-layer distributed control protocol for multiple unmanned aerial vehicles (UAVs) to fence a moving target cooperatively, using relative measurements. The multi-UAVs are divided into two layers: one is equipped with target detection sensors that can acquire relative bearing information; the other is equipped with relative distance sensors that can only acquire relative positions of neighbors. First, a bearing-based controller is presented for the leader layer to eliminate the singularity problem of observability of a moving target. Then, a distance-based controller is developed to fence the moving target within a convex hull at every moment for the follower layer. The relative spacing within the formation does not need to be given in advance, and the rotation and scaling of the formation can be quickly adjusted according to the design parameters. Third, we prove sufficient conditions for the design of the formation configuration, together with the selection of the fencing scale. The fencing protocol can be directly extended to 3-D space, and the form of the controller for 3-D scenarios is given as well. Finally, two numerical simulations and a hardware-in-loop (HIL) experiment verify the effectiveness of the proposed protocol. Note to Practitioners—This paper was motivated by the problem of cooperative moving target fencing control for multiple fixed-wing UAVs. In many practical missions, for instance, multiple UAVs protect a critical target or entrap an intruder target, it is typical that multiple fixed-wing UAVs encircle a moving target. Additionally, it may occur with communication limitations and the absence of global information in some complex environments, and the moving target may be non-cooperative. This would lead to existing fencing methods failing to be applied, which require communication to share information about the target and UAVs. A two-layer control frame is presented to address this problem. We use onboard sensors (target detection sensors for the leader layer and relative distance sensors for the follower layer) to obtain relative measurements of the moving target and other UAVs, removing communication and global information constraints. Then, a scale factor is introduced to make the proposed protocol widely applicable in narrow areas or places with obstructed views, driving the formation translating, rotating, and scaling with the movement of the target. The proposed protocol guarantees that the moving target is encircled by the convex hull formed by multiple UAVs, and it can be extended to 3-D space directly. In future research, we will perform the proposed protocol in practice.}, archive = {J_TASE}, author = {Shulong Zhao and Jiayi Zheng and Kun Liu and Jun Liu and Xiangke Wang}, doi = {10.1109/TASE.2024.3461727}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7145--7158}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative moving target fencing control for two-layer UAVs with relative measurements}, volume = {22}, year = {2025}, }
• (2025). Safe human-UAS collaboration from high-level planning to low-level tracking. TASE, 22, 7132--7144. (www)

  @article{ , abstract = {This paper studies the problem of safe human-uncrewed aerial system (UAS) collaboration in a shared work environment. By considering human and UAS as co-workers, we use Petri Nets to abstractly model evolution of shared tasks assigned to human and UAS co-workers. Particularly, the Petri Nets’ “places” represent work stations; and therefore, the Petri Nets’ transitions can formally specify displacements between the work stations. The paper’s first objective is to incorporate uncertainty regarding the intentions of human co-workers into motion planning for UAS, when UAS closely interacts with human co-workers. To this end, the proposed Petri Nets model uses “conflict” constructs to represent situations at which UAS deals with incomplete knowledge about human co-worker intention. The paper’s second objective is then to plan the motion of the UAS in a resilient and safe manner, in the presence of non-cooperative human co-workers. In order to achieve this objective, UAS equipped with onboard perception and decision-making capabilities are able to, through real-time processing of in-situ observation, predict human intention, quantify human distraction, and apply a non-stationary Markov Decision Process (MDP) model to safely plan UAS motion in the presence of uncertainty. Given the current and next UAS waypoints, assigned by the MDP planner, the paper applies Potryagin’s minimal principle to plan the desired trajectory of the UAS and uses a feedback linearaztion trajectory control to enable UAS with stable tracking of the desired trajectory. Note to Practitioners—Despite advances in aerial robotics, there are still significant barriers for their integration and application into human-centered jobs. Safety-related concerns, potential hazards to labor, and limited mission duration are some major barriers for using such a helpful technology. The main goal of this paper is to come up with a reliable solution for long-term collaboration between humans and unmanned aerial system (UAS) for safe and efficient accomplishment of a human-centered work. To this end, we propose to use Petri Nets to abstractly model and specify human-robot collaboration and effectively plan the tasks assigned to UAS and human co-workers. In this context, human co-workers can do some tasks that are difficult and somehow unsafe to be carried out by UAS. On the other hand, UAS can be deployed to monitor the work environment, recognize human wellness, or carry some small payloads. UAS co-workers enabled with on-board perception capabilities can also predict human coworker intention and quantify human co-worker distraction in real time through online processing of in-situ observations. Furthermore, UAS can safely plan its motion without human intervention or supervision, by combining Markov Decision Process, optimal control, and trajectory tracking control models.}, archive = {J_TASE}, author = {Hossein Rastgoftar}, doi = {10.1109/TASE.2024.3461726}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7132--7144}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe human-UAS collaboration from high-level planning to low-level tracking}, volume = {22}, year = {2025}, }
• (2025). Approximation-free prescribed performance fault-tolerant control of nonlinear systems: Considering deferred actuator switching. TASE, 22, 7122--7131. (www)

  @article{ , abstract = {This paper addresses the prescribed performance control (PPC) problem of nonlinear systems with unknown control direction. By virtue of composite error transformation and Nussbaum gain identification, a prescribed performance controller free of approximation tool, known initial condition and control direction is proposed. We then extend it to the setting where deferred actuator switching is utilized together with error monitoring to eliminate the effect of actuator fault on PPC. By adding an input auxiliary system, the controller is improved to be capable of tolerating the switching delay of actuator, at the price of temporary performance relaxation rather than the system instability. Using proof of contradiction, it is proved that all closed-loop signals remain globally bounded in the Lyapunov sense. Two simulation examples are given to verify the superiority of the proposed approach over the existing ones. Note to Practitioners—This paper is motivated by designing an algorithm that comprises dynamic redundancy strategy to exclude the impact of actuator faults while achieving the prescribed control performance. The existing methods usually rely on the intelligent approximation, known initial condition and control direction, and require the controller redesign when the backup actuator defers the switching. Due to this consideration, we use the full-error composite transformation and Nussbaum function to derive the PPC mechanism, making the controller not only approximation-free but also applicable to any error initialization and control direction. Moreover, the input auxiliary signal is introduced such that both fault-free PPC and delays-involved redundancy PPC can be uniformly achieved by a set of control framework, which leads to low computation burden and eases the parameter tuning.}, archive = {J_TASE}, author = {Chen-Liang Zhang and Ge Guo}, doi = {10.1109/TASE.2024.3461323}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7122--7131}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Approximation-free prescribed performance fault-tolerant control of nonlinear systems: Considering deferred actuator switching}, volume = {22}, year = {2025}, }
• (2025). A new adaptive robust sliding mode control for high-precision PKMs: Design, stability analysis, and experiments. TASE, 22, 7105--7121. (www)

  @article{ , abstract = {This paper proposes a novel adaptive feedback sliding mode control for parallel kinematic manipulators (PKMs), built on the conventional model-based sliding mode control. This structure was chosen for its robustness towards uncertainties and external disturbances. The contribution of this research is the inclusion of a feedforward term based on the dynamic model of the PKM in the control design. This feedforward term compensates for high nonlinear dynamics, as well as avoids measurement noise in control inputs. Additionally, the fixed feedback gains of the sliding mode controller are redesigned as adaptive gains, which provide better correction actions for larger tracking errors. A stability analysis of the proposed control solution, based on Lyapunov’s method, is provided. The effectiveness of the proposed controller is demonstrated through its application to a Gough-Stewart platform (MISTRAL parallel robot) in various real-time experimental scenarios. The proposed approach is compared with various controllers demonstrating its superiority. It ensures nominal root mean square tracking errors of (i) about $22\,\mu m$ in joint space, and (ii) about $27\,\mu m$ in traveling plate Cartesian position. Note to Practitioners—This paper was motivated by the problem of robustness towards uncertainties and operating conditions for PKMs and especially in industry. The literature contains some control solutions to deal with these issues by designing robust or adaptive nonlinear controllers. Indeed, despite their success, for the first type, large uncertainties may degrade the tracking performance and the robustness is not always guaranteed. Adaptive controllers can be more robust, but their tuning can be difficult and especially in the case of real-time estimation of dynamic parameters. In this paper, a new robust-based sliding mode control with adaptive feedback gains is proposed to ensure the robustness towards uncertainties and external disturbances. The idea is to design adaptive feedback gains for a model-based sliding mode control, resulting in an improved robustness, while ensuring an easy tuning of its parameters. The proposed solution is validated through real-time experiments on an industrial parallel robot, and compared with some model-based controllers. The obtained results show clearly the benefits of the proposed control design and its effectiveness and robustness. In future work, the proposed controller can be endowed with a real-time estimation of the dynamic parameters, as well as, a redesign based on the super-twisting algorithm.}, archive = {J_TASE}, author = {Youcef Fitas and Ahmed Chemori and Johann Lamaury and Thierry Roux}, doi = {10.1109/TASE.2024.3461156}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7105--7121}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A new adaptive robust sliding mode control for high-precision PKMs: Design, stability analysis, and experiments}, volume = {22}, year = {2025}, }
• (2025). Relative localization with non-persistent excitation using UWB-IMU measurements. TASE, 22, 7092--7104. (www)

  @article{ , abstract = {In multi-robot systems, accurate relative localization is indispensable for executing collaborative tasks in GPS-denied environments. This paper focuses on the relative localization problem relying on onboard UWB and IMU sensors. First, we propose a nominal adaptive gradient-based relative position observer for each robot. The estimation of time-varying relative position is transformed into the online constant parameter identification problem using only relative distance and velocity information. Furthermore, in order to relax the standard assumptions of persistently excited relative motions, a finite-time adaptive relative localization scheme is developed using the dynamic regression extension and mixing (DREM) technique. This scheme merely requires filtered relative velocity satisfying interval excited condition, which is milder than the persistent one. Finally, simulations are presented to verify the effectiveness of our theoretical results, followed by flight experiments on a team of three quadcopters. It indicates that the relative localization accuracy can reach centimeter level. Note to Practitioners—This paper is motivated by the relative localization problem without relying on any external infrastructure under GPS-denied environments, especially for situations where the robots’ trajectories cannot be persistently excited. Existing relative localization approaches generally assume that the robots’ velocities or displacements satisfy the persistent excitation condition, which restricts the motion forms of robots. This paper presents a new method that only requires the filtered relative velocity between robots to satisfy the interval excitation condition, so that accurate relative position estimations can be achieved within a finite time. In this paper, we provide a linear regressor equation generation method using the linear filter techniques, which mathematically characterizes the relationship between measurable signals (distance, velocity) and relative position. Then, we design a relative localization scheme based on the DREM method and give the convergence analysis. Both simulations and physical experiments suggest that the proposed method in this paper shows high localization accuracy about 10 cm and fast convergent speed. But it has not yet been applied to the specific control tasks. In future research, we will address the integration of relative localization and formation control in such scenarios.}, archive = {J_TASE}, author = {Yue Wang and Qingkai Yang and Hao Cui and Hao Fang}, doi = {10.1109/TASE.2024.3460811}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7092--7104}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Relative localization with non-persistent excitation using UWB-IMU measurements}, volume = {22}, year = {2025}, }
• (2025). Predefined-time synchronization for FitzHugh-nagumo neural networks with external disturbances and multiple time scales: An adaptive control approach. TASE, 22, 7080--7091. (www)

  @article{ , abstract = {The problem of predefined-time (PdT) synchronization of FitzHugh-Nagumo (FHN) neural networks is investigated. The FHN neural networks have shown proficiency in accurately representing the essential characteristics of neurons and effectively simulating the structural dynamics and functioning of the brain. The investigation of synchronization in FHN neural networks will enhance our comprehension of how neurons behave dynamically and promote their valuable utilization in the field of artificial intelligence. In this study, the successful integration of adaptive control and PdT control techniques not only achieves synchronization within a predetermined time but also eliminates the need to preselect controlling gains. The work develops new PdT stability theorems for singular perturbation systems and synchronizing controllers that exclude linear terms. Consequently, a set of sufficient conditions for achieving this synchronization goal is provided. Furthermore, it mathematically demonstrates the implementability of an upper bound for the settling time by parameterizing the predefined time as a controller parameter, thereby eliminating the need for explicit recalculations. Moreover, the control gains can be adjusted adaptively, enhancing the flexibility of control parameter tuning. Finally, several numerical simulations are presented to validate the effectiveness of the designed adaptive PdT control strategies. Note to Practitioners—The study of neural network-based synchronization in nonlinear systems is rapidly emerging as an important tool in many fields, including automation, telecommunications, computer science, mathematics, physics, and biology. Among the neural network architectures, FitzHugh-Nagumo (FHN) neural networks stand out for their ability to closely model the behaviors and dynamics of real neurons. This capability is not just theoretical; it has significant practical applications in enhancing the understanding of brain functions and improving the performance of artificial intelligence systems. In particular, neural networks with different time scales have better self-learning and generalization ability. Our innovative approach is the integration of adaptive control with predefined-time control techniques. This combination not only guarantees synchronization within a specific time frame but also dynamically adjusts these control parameters, enhancing the system’s adaptability to changes and disturbances without requiring prior knowledge of specific control gains. In practical applications, achieving efficient synchronization can significantly enhance AI systems, making them more efficient and reliable. For instance, in autonomous vehicles, better synchronization can improve decision-making processes, while in robotics, it can lead to more advanced and responsive automations.}, archive = {J_TASE}, author = {Shuting Chen and Ying Wan and Jinde Cao}, doi = {10.1109/TASE.2024.3460389}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7080--7091}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined-time synchronization for FitzHugh-nagumo neural networks with external disturbances and multiple time scales: An adaptive control approach}, volume = {22}, year = {2025}, }
• (2025). Control-oriented modeling, simulation, and predictive control of district heating networks. TASE, 22, 7064--7079. (www)

  @article{ , abstract = {District heating networks (DHNs) are recognized as pivotal infrastructures to reduce the carbon footprint of the heating sector. However, their large-scale dimension and the involved nonlinear dynamics lead to significant issues regarding their modeling and optimal control. Tackling these challenges, two dynamic DHN modeling frameworks are presented: a detailed simulation model and a control-oriented one. The simulation model allows the development of lightweight dynamic DHN simulators, accurately represents the main system dynamics in reduced computational time, and is well-suited for testing online control algorithms. This model is leveraged to develop a simulation library for DHN systems in the Modelica environment, named DHN4Control, freely distributed with this article in Nigro (2023). Then, a methodology for deriving a control-oriented DHN model is presented, suitable for designing a nonlinear model predictive control (NMPC) regulator, ensuring computationally efficient and cost-effective regulation while respecting operative constraints. The developed modeling and control methods are tested in simulation considering as benchmark a Test Facility district heating plant located at Ricerca sul Sistema Energetico - RSE S.p.A. in Milan (Italy), showing promising results in terms of modelling accuracy, computational efficiency and energy savings. Note to Practitioners—The dimension and complexity of multi-producer district heating networks (DHN) make their modeling, simulation, and online optimization challenging tasks. This work proposes a solution for district heating networks efficient simulation and optimal control design. The detailed explanations of the modelling equations and the use of open source software increase results replicability for practitioners. The simulation-oriented model comes with a lightweight simulation library, readily available at Nigro (2023). The control-oriented model is reusable on generic multi-producer district heating networks, for computationally efficient optimization-based control design. The optimal control is able to manage thermal generators and storages to minimize DHN operational costs while respecting operational constraints over the whole thermal network. The proposed solution is tested in simulation considering a real-world DHN plant currently under construction at Ricerca sul Sistema Energetico - RSE S.p.A in Milan (Italy). The simulation results show the cost saving performances of optimal controllers developed with the proposed model. Overall, by using the proposed solution, one can achieve faster simulation times and ease the burden of optimal control modelling for the development of an optimal-based control strategy for cost-effective DHN operation.}, archive = {J_TASE}, author = {Lorenzo Nigro and Alessio La Bella and Francesco Casella and Riccardo Scattolini}, doi = {10.1109/TASE.2024.3460173}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7064--7079}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Control-oriented modeling, simulation, and predictive control of district heating networks}, volume = {22}, year = {2025}, }
• (2025). Robust neural dynamics for depth maintenance tracking control of robot manipulators with uncertainty and perturbation. TASE, 22, 7052--7063. (www)

  @article{ , abstract = {The existence of inner uncertainty and external perturbation usually becomes a hindrance for the effective time-variant control of robot manipulators. Both the robustness and convergence property are regarded as two significant issues to be addressed for preferred solutions to robot manipulators. To handle the time-variant motion control of robot manipulators in the presence of both uncertainty and perturbation, a robust recurrent neural network (RRNN) model with definable convergence time (DCT) property is proposed in this paper. Theoretical analysis based on Lyapunov theory rigorously proves that the proposed RRNN model inherently possesses the global stability, robustness and time efficiency. The solution synthesized via the proposed model with uncertainty and perturbation shows desirable time-variant control performance, i.e., faster convergence and higher accurate. In addition, detailed path-tracking examples, performance comparisons, visual-assisted depth maintenance tracking control demonstrations, and extensive tests by applying both PUMA 560 and INNFOS are presented to validate the effectiveness and superiority of the proposed RRNN model for time-variant control of robot manipulators. Note to Practitioners—This article addresses the issue of uncertainty in robot information, a common occurrence in real-time robot learning and control. This paper presents a precise, efficient, and stable solution that leverages real-time feedback information to resolve real-time control problems for robotic manipulators at the velocity level. Additionally, the paper provides a comprehensive overview of the algorithmic steps and theoretical foundations of the RRNN model to facilitate understanding. To validate the effectiveness and superiority of the proposed approach, the study conducts computer simulations and comparisons using actual parameters and models. Finally, an application to the depth maintenance trecking control of robot mainpulators provides an applicative demo of the porposed neural dynamics for practitioners.}, archive = {J_TASE}, author = {Dechao Chen and Yifan Shao and Zhengwen Chen and Shuai Li}, doi = {10.1109/TASE.2024.3458998}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7052--7063}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust neural dynamics for depth maintenance tracking control of robot manipulators with uncertainty and perturbation}, volume = {22}, year = {2025}, }
• (2025). Segmented dynamic adaptive look-ahead smoothing feedrate scheduling with joint jerk constraints of 6R robot manipulators. TASE, 22, 7033--7051. (www)

  @article{ , abstract = {With the continuous development of robotic technology, there is an increasing demand for efficient, smooth, and precise trajectory planning and interpolation. For the robot manipulators with 6 rotational (6R) joints, smooth task paths are crucial for high-quality task and precise dynamic trajectory tracking. However, nonlinear mapping between joint and cartesian space complicates feedrate scheduling under joint constraints. Existing methods, like S-shape feedrate profiles or time-optimal approaches, are inefficient or compromise trajectory stability. This paper proposes a segment-based dynamically adaptive smooth look-ahead feedrate scheduling method based on local dynamic window and a Maximum Velocity Curve (MVC) for 6R robot manipulators. It balances the efficiency and stability of motion execution while considering spline trajectory and joint constraints. Five types of segmented velocities under the local dynamic window are identified, with adaptive smoothing strategies developed. Feedrate remains constant within segments and transitions smoothly between them, enhancing trajectory quality. The results of the smoothing preprocessing can be directly used for feedrate profile generation, ensuring smooth, non-oscillating motion while meeting performance and constraint requirements, which is better suited for real-time interpolation. Simulation and experimentation confirm the proposed method’s effectiveness. Note to Practitioners—The motivation of this article stems from the need to develop a feedrate scheduling method with joint jerk constraints of 6R robot manipulators for practical tasks like polishing, engraving, welding, and spraying. 6R robots, due to their nonlinear coupling kinematics, often fail to meet joint constraints. Existing methods, inefficient or requiring frequent changes in joint acceleration/jerk for time optimality, adversely affect task quality. To balance smoothness and operational efficiency for 6R robot tasks, we propose the segmented dynamic adaptive look-ahead smoothing feedrate scheduling with joint jerk constraints of 6R robot manipulators, ensuring trajectory execution within joint limits and balancing smoothness and operational efficiency, which is better suited for real-time interpolation. This method can be of great interest to readers working on precision manufacturing and robotics feedrate profile generation method.}, archive = {J_TASE}, author = {Yan Xu and Yaqiu Liu and Xun Liu and Jiabin Cao and Lin Zhang}, doi = {10.1109/TASE.2024.3458984}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7033--7051}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Segmented dynamic adaptive look-ahead smoothing feedrate scheduling with joint jerk constraints of 6R robot manipulators}, volume = {22}, year = {2025}, }
• (2025). Model predictive path-following framework for generalized N-trailer vehicles in the presence of dynamic obstacles modeled as soft constraints. TASE, 22, 7018--7032. (www)

  @article{ , abstract = {Collision avoidance is crucial for autonomous navigation systems. Many studies have addressed obstacle avoidance for single unicycles and car-like vehicles in on-road conditions. In this work, we extend the scope to generalised N-trailer vehicles, comprising a single active segment pulling multiple trailers. Unlike approaches that treat obstacles as hard constraints, we model them as soft constraints using Gaussian functions. This method maintains the convexity of the search space, reducing computational demands. However, the regions occupied by obstacles remain feasible. Thus, the Gaussian function’s amplitudes need to be carefully chosen to discourage navigation through these areas. Moreover, closed-loop stability is guaranteed by generating auxiliary references when the nominal path is occluded. The efficacy of this approach is demonstrated through simulated and field experiments with a tractor pulling two trailers. These experiments show the method’s capability to navigate around obstacles efficiently while maintaining computational efficiency, validating its practical applicability. Videos of the experiments and the implemented algorithms are available at https://usmcl-my.sharepoint.com/:f:/g/personal/nestor_deniz_usm_cl/EtU54g1NeslNhD8V7dAeu20B0umnQa4FKiMlzThkTAXYvg?e=swEXwg. Despite the success in real-time implementation, more research is needed to address the open questions discussed at the end of this article. Note to Practitioners—This work focuses on implementing obstacle avoidance for a kind of vehicles widely used in agriculture, mining, luggage transportation, and industry. A LiDAR Velodyne VLP16, configured with its lowest rotation speed for denser point clouds, is used to scan the environment. Proper attachment of the LiDAR to the tractor’s body minimises vibration and azimuth movements, ensuring accurate obstacle detection. Obstacles are modelled as Gaussian functions to maintain the convexity and optimise computational efficiency. The Gaussian function’s amplitude should be set high enough to effectively avoid collision when density of obstacle is high. The framework uses a control horizon $N_{c}$ and a prediction horizon $N_{p}$ beyond the control to anticipate obstacle’s position. However, a large prediction horizons $N_{p}$ is not advised when the model of the dynamic of the obstacles is not accurate.}, archive = {J_TASE}, author = {Nestor Nahuel Deniz and Fernando Auat Cheein}, doi = {10.1109/TASE.2024.3458809}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7018--7032}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model predictive path-following framework for generalized N-trailer vehicles in the presence of dynamic obstacles modeled as soft constraints}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered adaptive fixed-time practical tracking control for nonlinear systems through funnel function. TASE, 22, 7008--7017. (www)

  @article{ , abstract = {The article studies an adaptive practical fixed-time tracking control problem of nonlinear system using the funnel control method and the dynamic event-triggered control mechanism. First, an adaptive practical fixed-time controller is built using radial basis function neural networks and an improved funnel function. On the one hand, it eliminates the impact of unknown nonlinear functions on system performance and forces the tracking error to evolve within the funnel boundary. On the other hand, the transient performance of the system is enhanced by the preassigned funnel boundary. Second, a nonlinear command filter with high-order nonlinear term is constructed to solve the problem of “explosion of complexity” in the backstepping approach. Further, the filter error compensating input is designed to counteract the detrimental impacts of the filter error on the tracking performance. In addition, to achieve the dynamic modification of threshold parameters and minimize communication resources, a dynamic event-triggered mechanism is devised based on auxiliary dynamic variables and dynamic threshold parameters. It is rigorously proved theoretically that the closed-loop system is practical fixed-time stable and Zeno behavior is ruled out. Ultimately, the single-link robotic arm system validates the efficiency of the acquired control approach. Note to Practitioners—The application of traditional control algorithms is severely limited in real industrial systems due to typical constraints on input, output, and communication resources, particularly in autonomous surface vehicles and robot fields. Nevertheless, the existing methods only take into account asymptotic tracking control problems based on the time-triggered mechanism, which is impracticable for a large number of systems in reality. Therefore, the paper studies the problem of funnel control and communication resource of the nonlinear system, and an adaptive practical tracking control scheme is proposed through the dynamic event-triggered mechanism and funnel function, which not only ensures that the nonlinear system is practical fixed-time stable but also improves the transient and steady-state performances of the nonlinear system.}, archive = {J_TASE}, author = {Yudi Wang and Guangdeng Zong}, doi = {10.1109/TASE.2024.3458176}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {7008--7017}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered adaptive fixed-time practical tracking control for nonlinear systems through funnel function}, volume = {22}, year = {2025}, }
• (2025). Secure state estimation for cyber-physical systems against active eavesdropping attacks: A stochastic encryption-based method. TASE, 22, 6998--7007. (www)

  @article{ , abstract = {This article is concerned with the problem of remote state estimation for cyber-physical systems (CPSs) in the presence of an active eavesdropper. To prevent information leakage, a novel stochastic encryption scheme is proposed to protect the data transmitted via packet loss channel. The scheme can guarantee the user’s privacy and force the estimation error of the eavesdropper to diverge by using a predefined scheduling sequence to transmit raw data or an encrypted version randomly. In this case, two types of data are considered: raw measurements and local estimates. Aiming at these two transmission scenarios, the expected estimation error recursions with respect to encryption mechanisms, channel parameters, and the system dynamics are derived in the minimum mean error estimation sense. Further, we show the conditions that require the expected estimation error covariance of the user to be bounded, while the active eavesdropper’s expected estimation error grows unbounded by mathematical induction and matrix theory methods. Finally, an application example is illustrated to verify the performance of the encryption scheme. Note to Practitioners—This paper addresses the secure state estimation for a class of CPSs in the presence of eavesdropping attacks, which plays a vital role in many critical infrastructures, such as unmanned aerial vehicle (UAV) and smart grids. The system’s confidential information is transmitted over the wireless channel to a remote estimator that is likely to be overheard by a malicious eavesdropper. Most existing privacy protection methods not only constrain the capabilities of eavesdroppers, such as packet reception rate, but also have limited ability to disrupt the estimated performance of eavesdroppers. Besides, few studies have been conducted on the impact of active eavesdroppers who may have a higher packet reception rate. To overcome these difficulties, based on control-theoretic approaches, we propose two alternative encryption schemes for two different transmission scenarios. Note that our scheme can guarantee that the user’s estimation error is bounded, while the estimation error of the eavesdropper goes to unbounded. Accordingly, the privacy and estimation performance of the system can be ensured by our encryption scheme, and the encryption schemes can be conveniently applied to practical systems.}, archive = {J_TASE}, author = {Fei Tao and Dan Ye}, doi = {10.1109/TASE.2024.3457903}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6998--7007}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure state estimation for cyber-physical systems against active eavesdropping attacks: A stochastic encryption-based method}, volume = {22}, year = {2025}, }
• (2025). SDD-DETR: Surface defect detection for no-service aero-engine blades with detection transformer. TASE, 22, 6984--6997. (www)

  @article{ , abstract = {Vision-based surface defect detection (SDD) for no-service aero-engine blades provides a fast and effective way to monitor product quality. Most existing detection algorithms for aero-engine blades are 1) based on CNN, including artificially designed non-maximum suppression (NMS) operations, and 2) focus on improving the detection accuracy rather than improving the inference speed and even ignoring the latter. To solve the above problems, we introduce a novel object detection paradigm, DEtection TRansformer (DETR), to design a novel network (SDD-DETR) with high accuracy for the SDD of aero-engine blades. To our knowledge, the paper is the first to introduce the DETR detector to SDD of aero-engine blades. While providing high accuracy, the inference speed of DETR remained slow due to self-attention operation and feed-forward network (FFN). Therefore, two lightweight modules have been designed for SDD of aero-engine blades: a progressive feature input multi-scale deformable attention module (PFI-MSDA) and a lightweight FFN (LW-FFN). PFI-MSDA hierarchically reduces the number of tokens input to the self-attention module, thereby reducing the time complexity of the self-attention layer. LW-FFN shrinks the complexity of multilayer perceptron. In addition, no parameter sharing of the detection head is utilized to compensate for the accuracy drop caused by the lightweight. Experiments verify that our method has the same AP and F1-score as DINO (a DETR-based detector), but our approach is lighter. Compared with DINO, the FLOPs are reduced by $113.4{G}$ , the inference speed is increased by 42.4%, and the runtime memory usage is reduced by $5.9{G}$ , which allows our method to be trained on low-end GPUs with more batch size, further improving the training efficiency. The code is available at https://github.com/VDT-2048/SDD-DETR.Note to Practitioners—The motivation for this paper is to design a high-precision and high-inference speed visual detection method for the SDD of aero-engine blades. Most high-precision vision methods are based on the transformer framework. However, its high complexity and poor compatibility in deployment environments lead to slower detection speeds. Although the application object in this paper is aero-engine blades, it is also applicable in other fields of industry, such as rail detection, plate and strip steel detection, etc. However, the method proposed cannot be supported by deployment environments such as RKNN because of the deformable attention operator, so it takes a certain amount of time to be deployed and put into practical use. Currently, the frameworks of visual and language large models are based on transformers, consistent with the framework of our method, which makes extending our approach to large visual and multi-modal models more accessible.}, archive = {J_TASE}, author = {Xiangkun Sun and Kechen Song and Xin Wen and Yanyan Wang and Yunhui Yan}, doi = {10.1109/TASE.2024.3457829}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6984--6997}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SDD-DETR: Surface defect detection for no-service aero-engine blades with detection transformer}, volume = {22}, year = {2025}, }
• (2025). Two-stage metaheuristic framework based on irregular contours matching for outsourced aircraft maintenance parking stand allocation problem. TASE, 22, 6967--6983. (www)

  @article{ , abstract = {With the increase in aircraft maintenance orders and heterogeneity of aircraft irregular shapes, outsourced aircraft maintenance companies urgently need a more efficient and tailored intelligent aircraft parking allocation method. However, existing methods could be improved in lightweight handling of non-overlapping constraints and effective use of problem-specific heuristics. To tackle these challenges and achieve a more rational allocation of parking stands, a bi-objective optimization model involving rotation angles is firstly constructed to maximize hangar utilization and safety margin, and is decomposed into two single-objective optimization problems via a lexicographic method. To efficiently solve this model, problem characteristics of “irregular contours matching” are analyzed. Furthermore, a series of mechanisms that fully utilize the problem characteristics are designed, thus integrating a two-stage metaheuristic framework based on irregular contours matching. These mechanisms include an aircraft parking strategy based on geometric fit for rationally locating aircraft parking positions and mitigating the dimensional explosion problem, a metaheuristic optimizer with similar insertion neighborhood operation for enhancing the hangar utilization, and a fast safety margin optimization algorithm based on binary searching iterator. Experimental studies conducted on 18 real-world instances show that the proposed framework outperforms several state-of-the-art algorithms, as well as the dynamic search algorithm automatically selected by the CPLEX optimizer. Note to Practitioners—This paper investigates an aircraft parking stand allocation problem that originated in outsourced aircraft maintenance companies. The goal of the problem is to maximize the hangar utilization and safety margin. Existing aircraft parking stand allocation methods ignore the lightweight handling of non-overlapping constraints, optimal configuration of rotation angles and safety margins, and effective utilization of problem-specific heuristics. Thus the allocation efficiency could be further improved when coping with large-scale order requirements. This paper constructs a mathematical optimization model with limited rotation angles, and introduces a concise geometric tool to address aircraft collision problem. Furthermore, a two-stage metaheuristic framework based on problem characteristics is proposed to solve the model efficiently. The superiority of the present method was verified on 18 real-instances with various hangar sizes and maintained aircraft. It is believed that this method effectively improves maintenance resource utilization, reducing labor and maintenance costs of outsourced aircraft maintenance companies.}, archive = {J_TASE}, author = {Ben Niu and Gaocheng Cai and Tianwei Zhou}, doi = {10.1109/TASE.2024.3457773}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6967--6983}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Two-stage metaheuristic framework based on irregular contours matching for outsourced aircraft maintenance parking stand allocation problem}, volume = {22}, year = {2025}, }
• (2025). Fuzzy adaptive saturation control for nonlinear systems with event-triggered sampling and quantization. TASE, 22, 6954--6966. (www)

  @article{ , abstract = {Based on event-triggering and quantization mechanisms, a fuzzy adaptive quantization saturation control problem has been investigated for a class of nonlinear systems. The unknown components of the controlled system are approximated using fuzzy logic systems. Sensors are able to measure the states of the system. The state signals and some associated variables are sequentially event-triggered, quantized, and transmitted to the controller via network communication. To overcome asymmetric input saturation, an auxiliary subsystem is constructed to compensate for the nonlinearity caused by the saturation. By introducing dynamic surface technology and using the backstepping control design method, a quantization saturation control scheme is proposed, and the problem of virtual controllers being non-differentiable in quantization control is solved. Theoretical proof shows that all error signals of the closed-loop system are semi-globally uniformly ultimately bounded, and the Zeno behavior can be avoided. The control scheme is applied to a quarter-car active suspension system, and the simulation results verify the effectiveness of the proposed control method. Note to Practitioners—Data sampling and quantization are critical processes in both computer control systems and network control systems. Event-triggered sampling can reduce network congestion and energy consumption. However, most existing methods rarely consider both event-triggering and quantization of system states and parameters simultaneously. Additionally, both the unknown components of the model and the saturation constraints of the actuator are common in control systems. Taking into account the above issues, this paper proposes a fuzzy adaptive quantization saturation control scheme based on event-triggering and quantization mechanisms. This scheme considers multiple nonlinear links that are more practical in computer control. It has been successfully validated for its applicability in a quarter-car active suspension system.}, archive = {J_TASE}, author = {Tiechao Wang and Hongyang Zhang and Huayang Zhang}, doi = {10.1109/TASE.2024.3457537}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6954--6966}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fuzzy adaptive saturation control for nonlinear systems with event-triggered sampling and quantization}, volume = {22}, year = {2025}, }
• (2025). Adaptive prescribed time fuzzy control of interconnected nonlinear systems and its applications: A compensation-based approach. TASE, 22, 6944--6953. (www)

  @article{ , abstract = {This paper investigates the trajectory tracking control problem of the interconnected nonlinear system subjected to dead zone inputs. A compensation-based prescribed-time fuzzy control framework is designed and successfully implemented on an interconnected system’s real experimental device. A prescribed-time convergence performance boundary function and a tan-type transformation function are introduced to constrain the tracking errors of each subsystem within the interconnected system. To cope with the dynamic changes in the design process caused by error transformations and the effects of disturbances, an adaptive compensation strategy for offset compensation is developed. The unknown system interconnection dynamics are treated as uncertain dynamics, and an adaptive fuzzy logic estimation strategy is designed to offset the effects of the unknown interconnection dynamics and system nonlinear dynamics. In addition, a parameter estimation law is proposed to reduce the effect of dead zone inputs on the control system. The control input event triggering mechanism is developed to reduce the update frequency of the control signals. The stability analysis of the proposed control framework proves that all signals are ultimately bounded. Compared with traditional control strategies for interconnected systems, the designed control framework does not require assumptions about the interconnection dynamics of the system, and the proposed compensation strategy enhances the robustness of the control system. Crucially, this research work verifies the effectiveness and superiority of the designed control framework in practical applications. Note to Practitioners—With the development of modern technology, practical engineering systems have become more and more large, and these large systems contain many subsystems with interconnected influences, so it is necessary to further study the control problems of interconnected nonlinear systems in depth. In the application of real interconnected systems, the existence of interconnection dynamics is the main reason to affect the control effect, so it is a hot issue to study how to eliminate and reduce the interconnection effect and realize fast tracking. In addition, most of the interconnected systems in real engineering applications have dead zone control problems, so it is necessary to design a compensation control strategy to improve control accuracy. In this paper, we discuss the adaptive prescribed-time fuzzy control problem for interconnected nonlinear systems, design a fuzzy compensating control mechanism, and develop a prescribed-time tracking controller to realize fast-tracking of the system and improve tracking accuracy.}, archive = {J_TASE}, author = {Guangjun Chen and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3457028}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6944--6953}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive prescribed time fuzzy control of interconnected nonlinear systems and its applications: A compensation-based approach}, volume = {22}, year = {2025}, }
• (2025). Teacher-student collaboration: Effective semi-supervised model for defect instance segmentation. TASE, 22, 6932--6943. (www)

  @article{ , abstract = {Recent defect instance segmentation methods heavily rely on pixel-level annotated images. However, acquiring labeled defect data from modern manufacturing industries takes significant time and effort. In this paper, we propose a novel semi-supervised approach for defect instance segmentation via Teacher-Student model Collaboration (TSC) to address the challenges of small defect dataset sizes and the blurring boundaries of defects. Specifically, we propose a generalized distribution fusion module (GDFM) to improve the quality of pseudo-labels. This module constructs a Gaussian mixture model to estimate the feature distributions from the student model. Leveraging Bayes’ theorem, we calculate the posterior probability, which significantly enhances the accuracy of classification pseudo-labels and refines the ambiguous regions in segmentation pseudo-labels produced by the teacher model. To manage the blurring boundaries of defects, we propose a cross-supervision contrastive learning module (CSCL). By combining the idea of online hard example mining with contrastive learning, we propose a simple yet effective method to distinguish the easy/hard and positive/negative areas of defect instances of unlabeled and labeled images. Extensive experiments demonstrate that our TSC model achieves state-of-the-art performance across three semi-supervised defect instance segmentation datasets with low annotation ratios. Note to Practitioners—The defect instance segmentation task aims to accurately locate each defect with a corresponding mask. Recent CNN models for defect instance segmentation heavily rely on pixel-level annotations, which demand significant time and effort within modern manufacturing industries. Therefore, we expand the semi-supervised framework to encompass defect instance segmentation and propose a semi-supervised approach for defect instance segmentation. We use labeled images to train the model and utilize the highly confident output of unlabeled images as pseudo-labels to improve the model instance segmentation performance. Our approach is tailored to address two prominent characteristics in detection inspection: small dataset sizes and blurring boundaries, thereby undergoing corresponding improvements. Meanwhile, our method demonstrates promising defect detection performance in real-world industrial settings with minimal annotation requirements. Extensive experiments demonstrate that our TSC model achieves state-of-the-art performance on three defect instance datasets.}, archive = {J_TASE}, author = {Biaohua Ye and Jianhuang Lai and Xiaohua Xie}, doi = {10.1109/TASE.2024.3456910}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6932--6943}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Teacher-student collaboration: Effective semi-supervised model for defect instance segmentation}, volume = {22}, year = {2025}, }
• (2025). Event-triggered h∞ tracking control for dynamic artificial neural network models with time-varying delays. TASE, 22, 6922--6931. (www)

  @article{ , abstract = {Dynamic artificial neural network models refer to the dynamic model structures that include artificial neural networks such as multi-layer perceptrons, which are used to model nonlinear system dynamics. $H_{\infty }$ tracking control for dynamic artificial neural network models with communication delays and external disturbance is investigated in this article. First, we formulate the dynamic artificial neural network model as a sampled-state error dependent model for networked control systems with the event-triggered mechanism, which is designed to reduce the consumption of network resources. The network-induced delay considered in this paper is time varying with a known upper bound. Furthermore, by Lyapunov-based techniques, we present sufficient conditions such that the closed-loop system satisfies the $H_{\infty }$ tracking performance. In addition, we propose a method to co-design the tracking controller and event-triggering parameters in the form of linear matrix inequalities. The effectiveness of our proposed methods is validated through a simulation example and a turbofan engine hardware-in-the-loop experiment. Note to Practitioners—When mathematical models of the plant dynamics are not available, neural network modeling can serve as a useful method for controller design, provided we have numerical information about the system behavior. The novel stability conditions and the established controller design method can be adapted for different classes of dynamic artificial neural network models, including neural state space models, global input-output models and dynamic recurrent neural networks. This characteristic reduces the constraints on model selection, thus expanding the application scenarios. Instead of the stabilization problem, $H_{\infty }$ tracking control problem is considered in this paper, which has a wider range of applications. In view of the growing need to reduce unnecessary consumption of communication resources in some digital control systems such as smart power grid, aircraft, industrial automatic production and so on, different from existing results, a discrete-time version of periodic event-triggered mechanism is adopted in analysis and control of dynamic artificial neural networks. In addition, disturbances and time delays which widely exist in engineering are also considered. Therefore, the proposed method is more suitable for practical applications. Finally, the established method is applied to the turbofan engine multivariable trajectory tracking control problem, and the effectiveness is illustrated by experiments. In future research, we will address the design problem of dynamic artificial neural network observer for the situation where the system states are unmeasurable.}, archive = {J_TASE}, author = {Zi-Jie Wei and Kun-Zhi Liu and Peng-Fei Liu and Xi-Ming Sun}, doi = {10.1109/TASE.2024.3456782}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6922--6931}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered h∞ tracking control for dynamic artificial neural network models with time-varying delays}, volume = {22}, year = {2025}, }
• (2025). A budget-constrained workflow scheduling approach with priority adjustment and critical task optimizing in clouds. TASE, 22, 6907--6921. (www)

  @article{ , abstract = {In the rapidly evolving landscape of cloud computing, scheduling complex scientific workflows poses significant challenges, particularly with constraints like budget considerations. While recent years have seen considerable research focus on budget-constrained cloud workflow scheduling, existing studies primarily concentrate on constructing solutions without delving into potential enhancements. To address this gap, our paper introduces PACP-HEFT, a modified HEFT algorithm integrating priority adjustment and task optimization to minimize makespan. Leveraging the Heterogeneous Earliest-Finish-Time (HEFT) algorithm as its foundation, PACP-HEFT incorporates two priority adjusters informed by task characteristics gleaned from task dependency topology and real-time scheduling data. Additionally, a critical task optimizer conducts thorough analyses and optimizations to reduce overall makespan. Extensive experimentation with real-world workflows underscores PACP-HEFT’s superior performance compared to contemporary algorithms. Note to Practitioners—This paper presents a deterministic approach to budget-constrained scheduling problem, with the objective of minimizing the makespan. The proposed PACP-HEFT algorithm exhibits robust solution enhancement capabilities through the utilization of two priority adjusters and a critical task optimizer. The two priority adjusters fine-tune task priorities based on characteristics extracted from task dependency topology and real-time scheduling data, while the critical task optimizer performs an in-depth analysis of scheduling solutions and optimizes critical tasks to reduce overall makespan. Real-world workflow experiments clearly highlight the superior performance of PACP-HEFT compared to state-of-the-art algorithms. Validation experiments also underscore the effectiveness of each priority adjuster and the critical task optimizer. Moreover, PACP-HEFT’s ability to rapidly reach solutions further enhances its practicality, ensuring timely scheduling outcomes. Furthermore, we assess the adaptability and transferability of these components, emphasizing their potential applicability in enhancing the performance of other scheduling algorithms. For practical applications, this algorithm can be directly applied to rapidly minimize the makespan of budget-constrained scheduling problems in the cloud. Moreover, it can be integrated into other schedulers as a solution enhancement component, offering versatility and improved scheduling efficiency.}, archive = {J_TASE}, author = {Mingjie Fan and Xinchao Zhao and Xingquan Zuo and Lingjuan Ye}, doi = {10.1109/TASE.2024.3456762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6907--6921}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A budget-constrained workflow scheduling approach with priority adjustment and critical task optimizing in clouds}, volume = {22}, year = {2025}, }
• (2025). Synthesis of opacity-enforcing supervisory strategies using reinforcement learning. TASE, 22, 6896--6906. (www)

  @article{ , abstract = {In the control of discrete-event systems for current-state opacity enforcement, it is difficult to synthesize a supervisor by supervisory control theory (SCT) without explicit formal models of the systems. This study utilizes the reinforcement learning (RL) method to obtain supervisory policies for opacity enforcement in the case when the automaton model of the system is unavailable. The state space of the environment in the RL is dynamically generated through system simulation. Actions are defined according to the control patterns of the SCT. A reward function is proposed to evaluate whether the secret is exposed or not. Then, a sequence of state-action-reward chains are obtained as system simulation goes on. The frameworks of Q-learning and State-Action-Reward-State-Action (SARSA) algorithms are adopted to implement the proposed approach. The goal of the training is to maximize the total accumulative reward by optimizing the action selection in the learning process. Then, an optimal supervisory policy is obtained when the training process converges. Experiments are performed to illustrate the effectiveness of the proposed approach. The contributions are two aspects. Firstly, a supervisor for opacity enforcement is learned by RL training without an explicit formal model of the system. Secondly, the ability of the proposed method in computing supervisory policies without formal models addresses a significant gap in the literature and offers a new direction for research in opacity enforcement in discrete event systems. Note to Practitioners—Supervisory Control Theory (SCT) supplies an effective way to synthesize supervisors, which traditionally handles tasks with explicit system models for current-state opacity enforcement by restricting behavior of systems. However, formal models of systems are often confidential or otherwise unavailable. This paper presents a method for supervisor synthesis via reinforcement learning in the case of lacking formal models of systems. The proposed method leverages the characteristics of control patterns in SCT and optimizes the action selection in the training process through a reward mechanism that evaluates the secrecy of states. The approach can be applied to model-free RL frameworks such as Q-learning and State-Action-Reward-State-Action (SARSA) algorithms. The training is performed as the system simulation goes on. When the training process converges, the optimal policy can be used to enforce opacity for the system. However, Q-table is used to save the Q-value in both Q-learning and SARSA algorithms. In the worst case, the size of the Q-table grows exponentially with the number of states and controllable events increasing. This can lead to memory exhaustion when the system’s scale is large. To make the approach scalable, we will attempt to use DRL to train control policies for opacity enforcement in the future.}, archive = {J_TASE}, author = {Huimin Zhang and Li Huang and Wanling Huang and Lei Feng and Xianxian Li}, doi = {10.1109/TASE.2024.3456239}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6896--6906}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synthesis of opacity-enforcing supervisory strategies using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). An intelligent quantum cyber-security framework for healthcare data management. TASE, 22, 6884--6895. (www)

  @article{ , abstract = {Digital healthcare is essential to facilitate consumers to access and disseminate their medical data easily for enhanced medical care services. However, the significant concern with digitalization across healthcare systems necessitates for a prompt, productive, and secure storage facility along with a vigorous communication strategy, to stimulate sensitive digital healthcare data sharing and proactive estimation of malicious entities. In this context, this paper introduces a comprehensive quantum-based framework to overwhelm the potential security and privacy issues for secure healthcare data management. It equips quantum encryption for the secured storage and dispersal of healthcare data over the shared cloud platform by employing quantum encryption. Also, the framework furnishes a quantum feed-forward neural network unit to examine the intention behind the data request before granting access, for proactive estimation of potential data breach. In this way, the proposed framework delivers overall healthcare data management by coupling the advanced and more competent quantum approach with machine learning to safeguard the data storage, access, and prediction of malicious entities in an automated manner. Thus, the proposed IQ-HDM leads to more cooperative and effective healthcare delivery and empowers individuals with adequate custody of their health data. The experimental evaluation and comparison of the proposed IQ-HDM framework with state-of-the-art methods outline a considerable improvement up to 67.6%, in tackling cyber threats related to healthcare data security. Note to Practitioners—This paper aims to address the issue of digital healthcare data access, which requires both ease and security. Existing research either focuses solely on safe access or on high security, which often comes with high computational challenges. In this paper, we present a comprehensive approach that takes into account various challenges such as secure data storage, efficient data communication, and the prediction of malicious entities. We have developed a mathematical system to portray the overall management of healthcare data. All techniques proposed in this paper have been implemented using quantum computing and have been tested on four healthcare datasets. Initial experimental results suggest that the proposed approach is feasible. Our techniques can be applied to discover malicious entities and understand the behavior of real-life users in healthcare processes.}, archive = {J_TASE}, author = {Kishu Gupta and Deepika Saxena and Pooja Rani and Jitendra Kumar and Aaisha Makkar and Ashutosh Kumar Singh and Chung-Nan Lee}, doi = {10.1109/TASE.2024.3456209}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6884--6895}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An intelligent quantum cyber-security framework for healthcare data management}, volume = {22}, year = {2025}, }
• (2025). Periodic event-triggered impulsive control of linear uncertain systems. TASE, 22, 6876--6883. (www)

  @article{ , abstract = {The paper studies the robust stability of linear uncertain systems under periodic event-triggered impulsive control (ETIC). Based on sliding variables, some sufficient conditions are derived to ensure the robust stability of the addressed impulsive control systems, and the effects of both external disturbance and uncertainty on the stability are revealed. Interestingly, it is verified that the proposed ETIC can ensure the robustness of the addressed systems with any bounded external disturbance. While, it is shown that the uncertainty may bring destabilizing impact to the stability. Further, compared with continuous event-triggered mechanism that often requires high hardware accuracy and sensor complexity in previous results, the proposed periodic event-triggered mechanism is more relaxed and realistic. It is also shown that the interevent time generated by the proposed periodic event-triggered mechanism has a unified positive lower bound, which cannot be derived by previous continuous event-triggered mechanism. Two illustrative examples are given to show the excellence of the main results. Note to Practitioners—It is known that the industrial plant emulator setup serves as a scaled-down representation of various industrial plant mechanisms, such as conveyor belts, drives, servomechanisms, and assembly machines. This paper is motivated by the problem of impulsive control for an industrial plant emulator setup, which has been studied by using continuous-time control or intermittent control in the previous results. Note that impulsive control only needs to implement at some discrete instants, which can reduce control cost and communication pressure compared with the continuous-time control and intermittent control. In this study, periodic ETIC, which contains the advantages of both periodic event-triggered control and impulsive control, is proposed to achieve the robust stability of linear multi-input uncertain systems with external disturbances. Some novel sliding-variables-based conditions have been obtained to guarantee the robust stability of addressed systems and the impacts of uncertainty and external disturbances on the stability are revealed simultaneously. It has been verified that the uncertainty may have destabilizing impact on the stability and under some conditions, the addressed periodic ETIC systems are robust with unknown bounded external disturbances. In addition, there exists a unified minimum time between arbitrary two consecutive triggered instants. In fact, the periodic ETIC can be also applied to some practical systems such as the second-order relay system and buck converter system, as these systems can be represented as linear uncertain systems with external disturbances. Furthermore, it is an intriguing future research topic to extend the derived results to the case of nonlinear systems.}, archive = {J_TASE}, author = {Bangxin Jiang and Jinfei Wei and Yang Liu and Weihua Gui}, doi = {10.1109/TASE.2024.3456125}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6876--6883}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Periodic event-triggered impulsive control of linear uncertain systems}, volume = {22}, year = {2025}, }
• (2025). Tree search reinforcement learning for two-dimensional cutting stock problem with complex constraints. TASE, 22, 6860--6875. (www)

  @article{ , abstract = {Reinforcement learning (RL) has been widely used in recent years to solve combinatorial optimization problems; however, it has some limitations when solving such problems with practical features. It is difficult for RL to ensure the feasibility of solutions when solving optimization problems with complex constraints, which limits its industrial application. Using a two-dimensional cutting stock problem derived from the practical plate design process in the steel industry as an example, we propose a learning and searching framework that enables RL to obtain a feasible solution that obeys complex constraints. We first formulate the two-dimensional cutting stock problem as a Markov decision process (MDP) and then design a tree-search-based reinforcement learning (TSRL) algorithm within the proposed learning and searching framework. In the learning process, we establish the approximate stage-independent Bellman equation (ASIB) of the MDP and obtain the agent decision model by solving the ASIB with approximate linear programming and column generation. In the search process, based on the obtained approximate value function, an efficient parallel two-stage tree search is developed as the agent decision generator of RL to obtain near-optimal plate design schemes. The experimental results show that the proposed TSRL algorithm outperforms baselines in terms of the capacity to obtain a feasible solution, computational time, and solution quality. Note to Practitioners—This article was motivated by the slab design problem in the steel industry but it also applied to other two-dimensional cutting stock problems with complex constraints. This work promotes making decisions in industrial production. Our key contribution is to provide a good-enough solution in a short time, which can improve the resource utilization and reduce the production cost of industrial enterprises. In terms of methodology, we provide new insight into reinforcement learning, which makes reinforcement learning more applicable to practice scenarios.}, archive = {J_TASE}, author = {Fengyuan Shi and Ying Meng and Lixin Tang}, doi = {10.1109/TASE.2024.3456074}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6860--6875}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tree search reinforcement learning for two-dimensional cutting stock problem with complex constraints}, volume = {22}, year = {2025}, }
• (2025). Sliding-mode control for 2-D hidden markov jump roesser systems with partial information and its application in metal rolling process. TASE, 22, 6851--6859. (www)

  @article{ , abstract = {2-D systems which can model many practical engineering systems, have attractive research in the control topic in both practice and theoretical aspects. Due to the complex engineering environment, systems may encounter sudden structural changes and be unable to work properly. Fortunately, the Markov jump process can model this situation. Meanwhile, the information of systems may not be accessed timely or only partial information can be accessed. This work focuses on the hidden Markov model-based sliding mode control for 2-D Markov jump systems with partially unknown information and its application in the metal rolling process. By utilizing the Lyapunov function approach, some criteria are obtained to guarantee that the dynamics state reaches the designed sliding surface within a finite time, and the system is passive. Finally, the practicability of asynchronous sliding mode control law is verified by providing a metal rolling process. Note to Practitioners—In some practical systems, their state depends on two independent variables, and the internal structure or parameters of the system are changed due to sudden environmental interference and device failure, which can be described as 2-D Markov jump systems. Meanwhile, in practical engineering applications, obtaining all statistical probability information of systems requires a great cost or is impossible to accomplish. This paper proposes the sliding mode control method for 2-D Markov jump Roesser systems with partial information, in which the partial information problems include the partially known system operation mode information and the partially known probabilities information. A hidden Markov model is used to address the above two partial information problems. The design approach of a hidden Markov model-based sliding mode controller with partially unknown statistical probability information is proposed. The engineering applicability of the method is verified by the metal rolling technology. This study provides a new sliding mode control method to address the control problem for 2-D systems where the system information is partially obtained or accessed.}, archive = {J_TASE}, author = {Zhenghao Ni and Feng Li and Yudong Wang and Hao Shen}, doi = {10.1109/TASE.2024.3455570}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6851--6859}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sliding-mode control for 2-D hidden markov jump roesser systems with partial information and its application in metal rolling process}, volume = {22}, year = {2025}, }
• (2025). A lightweight integrated positioning system with occlusion-aware region-based pose tracking for oral and maxillofacial surgery. TASE, 22, 6839--6850. (www)

  @article{ , abstract = {The development of an accurate and robust positioning system for oral and maxillofacial surgery (OMS) is a challenging task, primarily due to the oral space limitations and line-of-sight occlusions. This paper presents a novel lightweight integrated positioning system for OMS, which can provide practical guidance utilizing only a micro camera installed on the end of the surgical instrument. An efficient region-based pose tracking method for texture-less teeth is proposed, which can use search lines around the object contour and simple local region partitioning strategy to improve pose accuracy. Besides, to deal with the possible partial occlusions of target during surgery, an occlusion-aware weight function is presented and utilized seamlessly in the pose optimization pipeline. This function calculates the pixel-wise occlusion probability using object contour and distance constraint, helping to improve the tracking robustness. Pivot calibration evaluation reveals that the tracking accuracy of the proposed camera-based handpiece is higher than the marker-based handpieces. Comparative experiments demonstrate that proposed pose tracking method has higher accuracy than existing state-of-the-art methods and ablation study confirms the effectiveness of the occlusion handling strategy. The overall positioning experiment indicates that the proposed system has satisfactory static poses stability and positioning accuracy. Furthermore, the main advantage of our system is that it is lighter and more integrated than other systems, which can reduce the system complexity, decrease the risk of line-of-sight occlusion, and lower the surgery cost.Note to Practitioners—This paper is motivated by the problem of restricted oral space constraints and partial occlusions during positioning for OMS. Compared with traditional OMS navigation systems, the designed system is more lightweight and more integrated without other external cameras and additional fiducial markers. Our system can provide practical guidance utilizing only a micro camera installed on the end of the surgical instrument. In addition, an efficient region-based pose tracking method for texture-less teeth is proposed to increase pose accuracy. Since the target can partially be occluded during the procedure, we present a novel occlusion-aware strategy to improve the tracking performance of partial occlusions. Our proposed system achieves a decent balance between positioning accuracy and hardware cost, and can easily be integrated into various dental surgical tools, thus it has tremendous potential for commercialization.}, archive = {J_TASE}, author = {Dongyue Li and Yaoqing Hu and Shaoan Wang and Mingzhu Zhu and Fusong Yuan and Junzhi Yu}, doi = {10.1109/TASE.2024.3455435}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6839--6850}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A lightweight integrated positioning system with occlusion-aware region-based pose tracking for oral and maxillofacial surgery}, volume = {22}, year = {2025}, }
• (2025). Kanban feedback control for wafer delay regulation of cluster tools. TASE, 22, 6822--6838. (www)

  @article{ , abstract = {In cluster tools widely used for semiconductor manufacturing, a wafer processed in a chamber should wait until a robot arm unloads it. Such wafer delays degrade the wafer quality due to residual chemicals and heat and hence cause quality variability and even failures. To prevent excessive wafer delays and their variability for single-armed and dual-armed cluster tools, we propose a simple, effective, and robust feedback control method called Kanban feedback control(KFC) that postpones a wafer loading task until a completion event of an associated task triggers it. We model the feedback control design problem as a problem of adding a feedback path between a pair of transitions to regulate token delays at a place in a timed event graph model. We develop closed formulae for wafer delays of the tools with KFC. We prove that KFC minimizes the worst-case wafer delay. We also prove that KFC makes the tool have a unique 1-cyclic schedule with constant wafer delays and ensures strong stability that recovers the same 1-cyclic schedule and the constant wafer delay at each chamber in a few cycles after a time disruption. By experimentation, we verify that KFC significantly reduces wafer delays and robustly regulates wafer delays and even cycle times against persistent time variation and significant sporadic time disruptions. Note to Practitioners—As circuit widths shrink to a few nanometers and chip architectural complexity soars up due to FinFET, GAA, and high-rise circuit stack-ups, quality risk in wafer fabrication processes surges. Therefore, wafer delays within a chamber after processing can cause more serious quality variations and failures. We propose a simple, effective, and robust way of regulating wafer delays. It can significantly contribute to yield enhancement.}, archive = {J_TASE}, author = {Dong-Hyun Roh and Tae-Eog Lee and Claude Martinez}, doi = {10.1109/TASE.2024.3455253}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6822--6838}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Kanban feedback control for wafer delay regulation of cluster tools}, volume = {22}, year = {2025}, }
• (2025). Hotspot-aware scheduling of virtual machines with overcommitment for ultimate utilization in cloud datacenters. TASE, 22, 6809--6821. (www)

  @article{ , abstract = {We address the problem of under-utilization of resources in datacenters during cloud operations, specifically focusing on the challenge of online virtual machine (VM) scheduling. Rather than following the traditional approach of scheduling VMs based solely on their static flavors, we take into account their dynamic CPU utilization. We employ $\Gamma $ -robustness theory to manage the dynamic nature and introduce a novel variant of bin packing - Probabilistic k-Bins Packing (PkBP), which theoretically protects the Physical Machines (PMs) from hotspots formation within a specified probability $\alpha $ . We develop a scheduling algroithm named CloseRadiusFit and cold-start AI-based prediction algorithms for the online version of PkBP. To verify the quality of our approach towards the optimal solutions, we solve the Offline PkBP problem by designing a novel Mixed Integer Linear Programming (MILP) model and a combination of numerical upper and lower bounds. Our experimental results demonstrate that CloseRadiusFit achieves narrow gaps of 1.6% and 3.1% when compared to the lower and upper bounds, respectively. Note to Practitioners—A growing trend in the cloud industry involves overcommitting VMs on PMs. While this approach can ease the problem of low utilization of resources in datacenters, it also introduces a higher risk of hotspots due to resource contention and competition among VMs. In this work, we propose a novel method that leverages $\Gamma $ -robustness theory and introduce effective heuristics to achieve ultimate utilization of datacenter resources while ensuring desirable service quality. We validate our approach using real-world production data from Huawei Cloud, improving resource utilization by 125% over traditional flavor-based allocation methods, while maintaining the occurrence of hotspots below 5% ( $\alpha =0.05$ ). Our solution only requires VMs’ real utilization data that is typically already collected in cloud providers’ production environments. Therefore, with minimal modifications to the existing scheduling system, cloud providers can easily implement our solution and reap its benefits. Moreover, in cases of the absence of historical utilization data for VMs (cold-start), we use machine learning to predict VM utilization statistics for our approach.}, archive = {J_TASE}, author = {Jiaxi Wu and Pavel Popov and Wenquan Yang and Andrei Gudkov and Elizaveta Ponomareva and Xinming Han and Yunzhe Qiu and Jie Song and Stepan Romanov}, doi = {10.1109/TASE.2024.3454821}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6809--6821}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hotspot-aware scheduling of virtual machines with overcommitment for ultimate utilization in cloud datacenters}, volume = {22}, year = {2025}, }
• (2025). T-S fuzzy sampled-data LFC scheme for wind power system via improved trapezoidal algorithm. TASE, 22, 6797--6808. (www)

  @article{ , abstract = {This paper studies the stability problem of sampled-data-based load frequency control (LFC) doubly fed induction generator (DFIG)-integrated wind power system (WPS). Firstly, a unified fuzzy DFIG-integrated WPS model is constructed by analyzing the nonlinear aspects of governor and turbine dynamics. Secondly, a sampled-data-based fuzzy proportional-integral control strategy (FPICS) is designed for stabilizing the power system. Thirdly, some new stability criteria are established by Lyapunov theory. Additionally, an improved trapezoidal algorithm is proposed to process the integral term in the controller, which can effectively reduce the consumption of computing resources. Finally, the effectiveness of proposed algorithm and the FPICS are verified through simulations. Note to Practitioners—Due to nonlinearities in the power system arising from the physical limitations of non-reheat governors, this paper aims to design an appropriate control strategy to address the nonlinear issues arising from governor valve position limiting, which may adversely affect the stability of the power system. Specifically, we propose an FPICS to tackle the nonlinearities caused by governor valves. Besides, an improved trapezoidal algorithm, which can dynamically determine and utilize the minimal number of subintervals according to the controller output variation, is designed to fit the numerical values of the controller’s integral term. This ensures control signal accuracy while maximizing computational resource savings. Moreover, to enhance the system’s tolerance to significant delays and disturbances, this paper considers the time-varying delay in the controller and establishes new stability criteria. Finally, through theoretical analysis and simulation cases, the effectiveness and reliability of the proposed fuzzy control strategy and improved trapezoidal algorithm are validated.}, archive = {J_TASE}, author = {Jia Ding and Jun Wang and Kaibo Shi and Xiao Cai}, doi = {10.1109/TASE.2024.3454762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6797--6808}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {T-S fuzzy sampled-data LFC scheme for wind power system via improved trapezoidal algorithm}, volume = {22}, year = {2025}, }
• (2025). Fast probabilistic energy flow calculation for natural gas systems: A convex multiparametric programming approach. TASE, 22, 6786--6796. (www)

  @article{ , abstract = {Probabilistic energy flow (PEF) calculation is a fundamental task for the operation and planning of both natural gas systems (NGSs) and integrated energy systems considering uncertainties. Traditional Monte Carlo simulation (MCS) based PEF approach requires repeated energy flow calculations based on a large number of random samples, leading to a huge computation burden. Hence, this paper proposes a convex multiparametric programming (MPP) based fast PEF calculation method for NGSs. First, we develop an energy function based convex optimization model whose optimal solution is equivalent to the solution of deterministic energy flow equations. Then, we propose a convex MPP model with uncertain boundary conditions (such as the gas injections/loads) as the parameters. A multiparametric quadratic approximation algorithm is further introduced to solve the proposed MPP and obtain the analytical energy flow expression. This analytical expression characterizes the mapping relationship between the energy flow solution and uncertain parameters. Finally, we establish an online PEF calculation framework, in which the repeat energy flow calculations can be efficiently performed by merely substituting the uncertain parameters into the analytical energy flow expression obtained offline. Case studies on multiple NGSs verify the effectiveness of the proposed method. Note to Practitioners—This paper introduces a novel approach for efficient PEF calculation in NGSs that addresses the computational challenges posed by traditional methods. Conventional MCS-based PEF approaches involve numerous energy flow calculations with massive stochastic scenarios, resulting in significant computational burdens. The key contribution of our work is to equivalently reformulate the original PEF problem as an energy function based convex MPP model, which incorporates uncertain boundary conditions (e.g., gas injections/loads) as its parameters. Then, a multiparametric quadratic approximation algorithm is suggested to obtain the parametric solution of this MPP, which is essentially an analytical expression of the energy flow solution as a function of uncertain parameters. This allows for efficient repeated energy flow calculations by simply substituting uncertain parameters into the pre-determined expression obtained offline. Thus, the online PEF computational efficiency can be significantly improved. We conduct case studies on several benchmark NGSs to validate the effectiveness and scalability of the proposed method. Numerical results show that the online PEF computation efficiency of the proposed method is approximately 2-3 orders of magnitude faster than that of the NR-MCS method.}, archive = {J_TASE}, author = {Wenhao Jia and Tao Ding and Yi Yuan and Hongji Zhang}, doi = {10.1109/TASE.2024.3454750}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6786--6796}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast probabilistic energy flow calculation for natural gas systems: A convex multiparametric programming approach}, volume = {22}, year = {2025}, }
• (2025). Mirror adaptive impedance control of multi-mode soft exoskeleton with reinforcement learning. TASE, 22, 6773--6785. (www)

  @article{ , abstract = {Soft exoskeleton robots (exosuits) have exhibited promising potentials in walking assistance with comfortable wearing experience. In this paper, a twisted string actuator (TSA) is developed and equipped with the exosuit to provide powerful driving force and variable assistance intensity for hemiplegic patients, which provides human-domain and robot-domain training modes for subjects with different movement capabilities. Since the human-exosuit coupling dynamics is difficult to be modeled due to the soft structure of the exosuit and incomplete knowledge of the wearer’s performance, accurate control and efficient assistance cannot be guaranteed in current exosuits. By taking advantage of the motion characteristic of hemiplegic patients, a mirror adaptive impedance control is proposed, where the robotic actuation is modulated based on the motion and physiological reference of the healthy limb (HL) as well as the performance of the impaired limb (IL). A linear quadratic regulation (LQR) is formulated to minimize the bilateral trajectory tracking errors and human effort, and the adaptation between the human-domain and robot-domain modes can be realized. A reinforcement learning (RL) algorithm is designed to solve the given LQR problem to optimize the impedance parameters with little information of the human or robot model. The proposed robotic system is validated through experiments to perform its effectiveness and superiority. Note to Practitioners—To assist walking for hemiplegic patients, it is crucial to provide comfortable and compliant driving force that can minimize the patients’ voluntary effort. The development of soft exoskeleton is able to realize comfortable wearing experience, and the proposed TSA can output compliant driving force with different training modes and assistance intensities. To address the problem in accurately building the human-exosuit coupled model, this work achieves the adaptive control strategy for patients with various movement capabilities in two steps. Firstly, a mirror adaptive impedance controller is proposed to make the patient’s HL tightly follow the IL’s motion for high safety guarantee performance and training autonomy. Secondly, a reinforcement learning-based LQR framework is constructed to minimize the patient’s voluntary effort by optimizing the prescribed impedance model parameters, which can significantly facilitate assistance efficiency for different patients. The experiments demonstrate that the proposed robotic system can obtain appropriate training modes and efficient walking assistance for human subjects. In the future study, it will be investigated how to assist patients in walking on different terrains with highly stable and adaptive actuation, which will accelerate the application of the proposed exosuit into activities of daily living.}, archive = {J_TASE}, author = {Jiajun Xu and Kaizhen Huang and Tianyi Zhang and Mengcheng Zhao and Aihong Ji and Youfu Li}, doi = {10.1109/TASE.2024.3454444}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6773--6785}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mirror adaptive impedance control of multi-mode soft exoskeleton with reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Leveraging transfer learning for data augmentation in fault diagnosis of imbalanced time-frequency images. TASE, 22, 6762--6772. (www)

  @article{ , abstract = {The rapid advancement of deep learning and time-frequency analysis techniques have brought about a revolution in fault diagnosis for mechanical systems, offering flexible and efficient solutions. Nevertheless, data imbalance issues continue to pose significant obstacles in fault diagnosis modeling. In this research, we propose the use of a domain adaptation generative adversarial network (DAGAN) that capitalizes on transfer learning to extract valuable information from the majority-class data, while concurrently generating and augmenting minority-class data to expand the training dataset. DAGAN incorporates advanced techniques, including deep domain confusion and parameter forgetting, to enhance knowledge extraction and transfer during the transfer learning process, resulting in more realistic and comprehensive generation outcomes when dealing with small sample training. Furthermore, we have developed an imbalanced fault diagnosis method based on DAGAN, which further incorporates Continuous Wavelet Transform and Deep Residual Networks. Finally, the effectiveness and superiority of proposed method are validated on bearing and gearbox datasets. The experimental results demonstrate the outstanding performance of our method in effectively addressing imbalanced fault diagnosis. Note to Practitioners—In this paper, we present a practical approach to enhance the diagnosis of imbalanced faults. Our approach begins by utilizing Time-frequency images, which offer a comprehensive representation of the temporal and spectral characteristics of mechanical systems’ behavior. These images serve as input features and form the foundation for our fault diagnosis modeling. To address the limitations imposed by imbalanced data, we introduce DAGAN and incorporate advanced techniques such as deep domain confusion and parameter forgetting. These techniques facilitate the extraction and transfer of knowledge during the transfer learning process of DAGAN. Consequently, our approach generates more realistic and comprehensive outcomes, even when confronted with limited training samples. To validate the effectiveness and superiority of our proposed approach, we conducted extensive experiments on bearing and gearbox datasets. The results of these experiments demonstrate that our practical approach, which combines wavelet transform-based time-frequency analysis and the innovative DAGAN framework, offers a reliable and comprehensive solution for overcoming challenges associated with imbalanced fault data.}, archive = {J_TASE}, author = {Xiaoyu Jiang and Junhua Zheng and Ziyi Chen and Zhiqiang Ge and Zhihuan Song and Xiaoguang Ma}, doi = {10.1109/TASE.2024.3454418}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6762--6772}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Leveraging transfer learning for data augmentation in fault diagnosis of imbalanced time-frequency images}, volume = {22}, year = {2025}, }
• (2025). Adaptive critic control with knowledge transfer for uncertain nonlinear dynamical systems: A reinforcement learning approach. TASE, 22, 6752--6761. (www)

  @article{ , abstract = {This paper presents an online transfer heuristic dynamic programming (THDP) control approach for a class of nonlinear discrete systems. The proposed approach integrates transfer learning with adaptive critic control. To design a robust optimal control strategy for the nonlinear discrete systems, we utilize sample data collected from a source task to acquire prior knowledge. This prior knowledge is subsequently used to guide the online control process of nonlinear systems of target tasks. To avoid negative transfer effects and conserve computational resources, we introduce a novel attenuation function with a truncation mechanism. Additionally, we develop a disturbance compensation control mechanism to address uncertainties. Furthermore, we demonstrate that the properties of the uncertain nonlinear systems under robust optimal control, as well as the weight error of neural networks, are ultimately uniformly bounded given certain conditions. Finally, two simulations are conducted to verify the performance of the proposed algorithm. Note to Practitioners—Adaptive dynamic programming (ADP) is one of the main methods to solve the Hamilton-Jacobi-Bellman (HJB) equation. However, when using neural network approximation, it often requires a long time of iteration and a large amount of computational process, wasting a lot of computational resources. For this reason, we propose an ADP control scheme with enhanced detection speed: that is, by learning a class of similar tasks to obtain prior knowledge to assist in the online control of our actual system. At the same time, this paper considers system disturbances, which means that they are more universal and robust. After simulation experiments, it has been proven that this scheme has good performance.}, archive = {J_TASE}, author = {Liangju Zhang and Kun Zhang and Xiang Peng Xie and Mohammed Chadli}, doi = {10.1109/TASE.2024.3453926}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6752--6761}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive critic control with knowledge transfer for uncertain nonlinear dynamical systems: A reinforcement learning approach}, volume = {22}, year = {2025}, }
• (2025). Tractable data-driven model predictive control using one-step neural networks predictors. TASE, 22, 6740--6751. (www)

  @article{ , abstract = {Model Predictive Control (MPC) is a popular control strategy that relies on the availability of a prediction model to estimate future system trajectories over a finite time horizon. Recently, researchers have introduced Neural Networks (NNs) into the MPC framework for the development of data-driven prediction models. In MPC, the control actions are computed by solving iteratively, at each time-step, an optimization problem subject to state and input constraints. Finding the optimal solution to such a problem is a crucial challenge in the data-driven setting, due to the complexity and black-box nature of data-driven models such as NNs. This paper addresses this challenge by proposing a hierarchical deep NN formed by a set of cascading one-step NN predictors whose combination constitutes an interpretable prediction model over the entire prediction horizon. Thanks to the proposed NN architecture, it is shown that the resulting optimal control problem is tractable, as it can be solved by employing efficient iterative algorithms, and interpretable, so that input and state constraints can be enforced seamlessly. The effectiveness of the proposed method is validated through numerical simulations. Note to Practitioners—Model Predictive Control (MPC) is a widely used methodology in the industry which typically relies on the availability of a model in the form of step response, transfer function or state-space models. In some cases, the explicit model might not be available or its accuracy may be not sufficient for the required closed-loop performance. This paper aims to develop a simple and practical framework for deploying a model-free data-driven MPC solution based on deep learning. This objective is pursued by suggesting a novel approach using simple neural networks in a cascading interpretable structure. Such networks are used to predict the one-step evolution of the system, and their cascade represents the MPC prediction model over an arbitrary long prediction horizon. We characterize such a neural model focusing on its interpretability and tractability, deriving the resulting optimal control problem to be solved in a receding horizon strategy. We then show that the MPC optimization can be solved efficiently using highly efficient iterative algorithms that can be implemented in practice. Numerical simulations involving the use of the Alternating Direction Method of Multipliers (ADMM) algorithm show its effectiveness for both linear and nonlinear systems.}, archive = {J_TASE}, author = {Danilo Menegatti and Alessandro Giuseppi and Antonio Pietrabissa}, doi = {10.1109/TASE.2024.3453668}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6740--6751}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Tractable data-driven model predictive control using one-step neural networks predictors}, volume = {22}, year = {2025}, }
• (2025). Multisynchronization of coupled multistable neural networks via event-triggered impulsive control and its application to associative memory. TASE, 22, 6729--6739. (www)

  @article{ , abstract = {This article studies multisynchronization of coupled multistable neural networks (NNs) with directed topology via event-triggered impulsive (ETI) control. At first, the activation function (AF) with ${2p}$ corners is proposed and it is proved that an n-neuron subnetwork can produce ${(p+1)^{n}}$ locally stable equilibrium points (EPs) or periodic orbits (POs) under some criteria. Furthermore, to achieve multisynchronization, an ETI controller is designed. Compared with conventional impulsive control strategy, ETI control strategy proposed in this paper can reduce the communication cost and save the bandwidth. Sufficient conditions are given to ensure both dynamical multisynchronization (DMS) and static multisynchronization (SMS) of coupled neural networks (CNNs) with fixed and switching topologies. Moreover, it is proved that the Zeno behavior can be avoided. Lastly, two examples and the application to associative memory are illustrated to testify the validity of the obtained results. Note to Practitioners—ETI control could reduce the number of packets sent as well as the control cost compared with conventional impulsive control. In addition, combining impulsive control with event-triggered schemes into multisynchronization analysis of CNNs is challenging because of the large number of synchronization manifolds in CNNs. Therefore, this research studies multisynchronization of coupled multistable NNs with directed topology via ETI control. A kind of ETI controller is designed and Zeno behavior is avoided. Moreover, the multisynchronization of CNNs is applied to the associative memory for the first time. Compared with the multistability-based associative memory, the multisynchronization-based associative memory can have superiority in resisting the noise interference.}, archive = {J_TASE}, author = {Yang Liu and Zhen Wang and Xia Huang and Hao Shen}, doi = {10.1109/TASE.2024.3453294}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6729--6739}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multisynchronization of coupled multistable neural networks via event-triggered impulsive control and its application to associative memory}, volume = {22}, year = {2025}, }
• (2025). A sparse bayesian learning for diagnosis of nonstationary and spatially correlated faults with application to multistation assembly systems. TASE, 22, 6716--6728. (www)

  @article{ , abstract = {Sensor technology developments provide a basis for effective fault diagnosis in manufacturing systems. However, the limited number of sensors due to physical constraints or undue costs hinders the accurate diagnosis in the actual process. In addition, time-varying operational conditions that generate nonstationary process faults and the correlation information in the process require to consider for accurate fault diagnosis in the manufacturing systems. This article proposes a novel fault diagnosis method: clustering spatially correlated sparse Bayesian learning (CSSBL), and explicitly demonstrates its applicability in a multistation assembly system that is vulnerable to the above challenges. Specifically, the method is based on a practical assumption that it will likely have a few process faults (sparse). In addition, the hierarchical structure of CSSBL has several parameterized prior distributions to address the above challenges. As posterior distributions of process faults do not have closed form, this paper derives approximate posterior distributions through Variational Bayes inference. The proposed method’s efficacy is provided through numerical and real-world case studies utilizing an actual autobody assembly system. The generalizability of the proposed method allows the technique to be applied in fault diagnosis in other domains, including communication and healthcare systems. Note to Practitioners—This article proposes a new process fault diagnosis method: clustering spatially correlated sparse Bayesian learning. This method effectively diagnoses time-varying defects by leveraging the correlation structures in the process when sensor measurements are insufficient. The actual autobody assembly process is utilized to show the proposed method’s effectiveness. The proposed method performs superior to the benchmark methods in fault detection capability. In addition, the proposed method accurately estimates the severity of the process faults, providing significant information to the practitioners for their decision-making in the maintenance schedule. Specifically, the error between the estimation from the proposed method and the actual severity of the process faults achieves less than 10% of error of all the benchmark methods when there exists a high correlation between the variations of the fixture locators in the autobody assembly system.}, archive = {J_TASE}, author = {Jihoon Chung and Zhenyu Kong}, doi = {10.1109/TASE.2024.3452726}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6716--6728}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A sparse bayesian learning for diagnosis of nonstationary and spatially correlated faults with application to multistation assembly systems}, volume = {22}, year = {2025}, }
• (2025). Performance analysis of bernoulli serial lines with small batch production and machine switch On/Off control. TASE, 22, 6702--6715. (www)

  @article{ , abstract = {Manufacturing industries play a pivotal role in global economies. It is a powerful driver of economic growth, creating a significant number of job opportunities and providing support for innovation and technological development. In the present world, where environmental concerns and resource conservation are paramount, reducing energy consumption while ensuring high processing quality is a significant challenge. During production, controlling the on/off of the machine reasonably can reduce energy consumption. This paper explores a method to analyze the performance of production lines containing buffers of finite capacity and machines that can be switched on and off when the specific conditions are met. We focus on finite production processes where the number of products is limited. In this paper, we establish a mathematical model for a three-machine production line and propose analytical methods for calculating production performance indicators. The analysis of the three-machine line employs the Markov method. Additionally, an aggregation method is introduced to extend the analysis to multi-machine systems, yielding promising results through numerical experiments. This research contributes to the advancement of production systems, improving a sustainable and competitive future for the manufacturing industry. Note to Practitioners—Small-batch production refers to the production of a small quantity of products, which can achieve customized order processing and improve resource utilization. Due to the dynamic nature of such processes, the traditional steady-state analysis method may not be applicable. This paper proposes an analytical method to predict the dynamic behavior of small batch production systems with Bernoulli machines that can be switch-on/off controlled and with finite capacity buffers. The algorithm developed in this paper can be used and support the production managers and engineers to predict the dynamic performance of the system with high accuracy. It can assist in decision-making in production control activities.}, archive = {J_TASE}, author = {Zhiyang Jia and Xiuxuan Tian and Zunjun Wang and Gang Wang}, doi = {10.1109/TASE.2024.3452672}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6702--6715}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Performance analysis of bernoulli serial lines with small batch production and machine switch On/Off control}, volume = {22}, year = {2025}, }
• (2025). Finite-time distributed optimization in unbalanced multiagent networks: Fractional-order dynamics, disturbance rejection, and chatter avoidance. TASE, 22, 6691--6701. (www)

  @article{ , abstract = {This study focuses on solving the finite-time distributed optimization issue of fractional-order multiagent systems (FOMASs) over unbalanced directed graphs (digraphs) that are subject to disturbances. Each agent in the FOMASs has a local cost function that is only available to itself, which may or may not be convex and quadratic. To address this challenge, the study proposes a fully distributed gradient-sum-estimation (GSE) optimization algorithm as a continuous control law, which comprises three parts. In the first part, a disturbance estimator term is proposed for each agent to estimate its disturbance within a finite time. In the second part, a sliding-mode control (SMC) term is presented to ensure that all agents reach the sliding surface within a finite time. In the last part, a novel GSE-based optimization term is constructed to capture the global optimal solution within a finite time. The GSE is fully distributed and used for estimating the sum of all gradients within a finite time. This fully distributed GSE optimization algorithm has zero-error finite-time convergence, disturbance rejection, and chatter avoidance properties. Finally, the study verifies the validity and superiority of the proposed GSE optimization algorithm by comparing some graphical simulation results.Note to Practitioners—Due to the presence of disturbances in actual industrial systems, the distributed optimization problem of disturbed FOMASs is studied in this study, which can be applied to energy consumption optimization, parameter estimation and scheduling, economic dispatch, and distributed energy resources optimal in power systems. Our study proposes a novel completely distributed fractional-order controller that is endowed with the properties of disturbance rejection, zero-error finite-time optimal convergence, and chatter avoidance. The commonly existing limitation that each local cost function is convex or quadratic hinders its implementation in real-world scenarios. We address this limitation and broaden the scope of local cost functions to include nonconvex and nonquadratic functions. In addition, by comparing the simulation results, the fractional-order controller of FOMASs has superior steady-state and transient performance over the conventional first-order MASs controller, such as faster convergence and lower energy consumption. As a result, the proposed fractional-order controller is more in line with practical engineering applications.}, archive = {J_TASE}, author = {Ping Gong and Qing-Guo Wang and Choon Ki Ahn}, doi = {10.1109/TASE.2024.3452472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6691--6701}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time distributed optimization in unbalanced multiagent networks: Fractional-order dynamics, disturbance rejection, and chatter avoidance}, volume = {22}, year = {2025}, }
• (2025). Toward cognitive digital twin system of human-robot collaboration manipulation. TASE, 22, 6677--6690. (www)

  @article{ , abstract = {Multielement decision-making is crucial for the robust deployment of human-robot collaboration (HRC) systems in flexible manufacturing environments with personalized tasks and dynamic scenes. Large Language Models (LLMs) have recently demonstrated remarkable reasoning capabilities in various robotic tasks, potentially offering this capability. However, the application of LLMs to actual HRC systems requires the timely and comprehensive capturing of real-scene information. In this study, we suggest incorporating real scene data into LLMs using digital twin (DT) technology and present a cognitive digital twin prototype system of HRC manipulation, known as HRC-CogiDT. Specifically, we initially construct a scene semantic graph encoding the geometric information of entities, spatial relations between entities, actions of humans and robots, and collaborative activities. Subsequently, we devise a prompt that merges scene semantics with prior knowledge of activities, linking the real scene with LLMs. To evaluate performance, we compile an HRC scene understanding dataset and set up a laboratory-level experimental platform. Empirical results indicate that HRC-CogiDT can swiftly perceive scene changes and make high-level decisions based on varying task requirements, such as task planning, anomaly detection, and schedule reasoning. This study provides promising insights for the future applications of LLMs in robotics.Note to Practitioners—Recently, LLMs have demonstrated significant success in various robotic tasks, suggesting their potential as a powerful tool for robotic decision-making. Motivated by this, to improve the production efficiency of HRC in flexible manufacturing, we innovatively combine LLMs with DT technology, and propose a cognitive DT system for HRC, aiming to integrate LLMs into the decision-making loop of HRC system. Experiments conducted in a laboratory-scale platform indicate that the proposed system can handle different decision-making needs in different HRC activities. This system can provide professional guidance to operators in a comprehensible form and serve as a medium for monitoring the safety and standardization of the manipulation process. Future work will explore the use of virtual space provided by the proposed system to optimize the decision outputs of LLMs to make the proposed system more broadly applicable.}, archive = {J_TASE}, author = {Xin Li and Bin He and Zhipeng Wang and Yanmin Zhou and Gang Li and Xiang Li}, doi = {10.1109/TASE.2024.3452149}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6677--6690}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward cognitive digital twin system of human-robot collaboration manipulation}, volume = {22}, year = {2025}, }
• (2025). Semi-supervised domain adaptation with dual-adversarial learning for lane detection. TASE, 22, 6664--6676. (www)

  @article{ , abstract = {Lane detection approaches relying on abundant annotations have achieved great progress in autonomous driving. These approaches may suffer from performance collapse when directly applied to unseen images due to domain discrepancy. Most researches utilize unsupervised domain adaptation (UDA) to mitigate domain shifts. Despite this, there is still a large performance gap between UDA and supervised methods because lack of target supervision makes improving performance of cross-domain lane detection difficult. In this paper, we introduce semi-supervised domain adaptation (SSDA) to lane detection and develop a simple yet effective dual-adversarial learning scheme to minimize domain discrepancy. Specifically, we conduct global alignment by calculating entropy information of unlabeled target and source data to align entropy distribution and enhance feature confidence of lanes. Domain consistency adversarial learning is proposed to make full use of both source and unlabeled/labeled target data. A triplet of features is used to learn lane-invariant representations by distinguishing whether the feature belongs to same domain. This is conducive to promote consistency of lanes information in cross domain. In two datasets, our method obtains a better performance of 42.5% (54.9%) and 45.8% (64.1%) based on a small amount of target supervision. Experimental results demonstrate effectiveness of our method as well as its superiority to other methods in terms of reducing performance gap with supervised methods (https://github.com/shenhuqiji/LD-SSDA). Note to Practitioners—This paper focuses on the cross domain lane detection task to narrow domain discrepancy in lane detection among different scenarios for autonomous driving. It aims to reduce extra label annotations and retraining costs. Considering unsupervised domain adaptation method is hard to further improve detection performance when a large domain shift occurs in adaptation scenarios, we construct an effective dual-adversarial learning scheme based on semi-supervised domain adaptation for lane detection. This method takes as input a triplet of source data, unlabeled and labeled target data to effectively enhance cross domain detection accuracy, which is contributed to the perception tasks in driving scenarios, such as path and motion planning, lane departure warning, lane changing decisions. Our method has a better generalization ability in face of various scenarios and can be easily integrated in software platforms, which greatly facilitates the applications on intelligent systems in autonomous vehicles.}, archive = {J_TASE}, author = {Xuedong Yao and Yandong Wang and Lei Dai and Shihong Zhang and Mingxuan Dou and Yuejin Deng}, doi = {10.1109/TASE.2024.3451507}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6664--6676}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semi-supervised domain adaptation with dual-adversarial learning for lane detection}, volume = {22}, year = {2025}, }
• (2025). GF-SLAM: A novel hybrid localization method incorporating global and arc features. TASE, 22, 6653--6663. (www)

  @article{ , abstract = {A global and feature-based hybrid algorithm, which integrates global information and feature-based simultaneous localization and mapping (GF-SLAM). This system can operate adaptively when external signals are unstable, thereby avoiding cumulative errors produced by local methods. In agricultural planting bases with abundant circular arc features, the focus is on efficiently exploring the correlations between these features to optimize the robust real-time positioning system for work vehicles. In this process, feature-based SLAM (F-SLAM) is applied to partial positioning using a particle filter. Available global information is then fused using an extended Kalman filter (EKF) for precise positioning and deviation correction in the mapping process, thereby achieving an effective combination of two positioning modes. The proposed model was evaluated using two simulation environments and a comparison with representative techniques. Results showed that GF-SLAM was competitive in normal conditions while requiring fewer computations and significantly reducing the drift in F-SLAM for stable global signals. Switching between these two algorithms eliminated positioning errors to within 1 cm for a test case in which global localization was lost in 54.7% of the route, producing an error within 3 cm. The code will be open source. Note to Practitioners—Our adaptive fusion strategy aims to address challenges in real-world agricultural scenarios where global information may be lacking or unstable. This approach enhances the robustness and reliability of robot positioning in practical applications. We invite practitioners to consider the adaptability of our system to diverse environments, particularly those with limited or fluctuating global information. In this paper, we first introduce the EKF module for global localization, then illustrate the establishment of feature maps and particle filter positioning in F-SLAM. We conduct comparison experiments in two virtual environments and real agricultural planting bases, where approximately half of the route lacks global information. The results demonstrate the benefits of our adaptive fusion strategy in practical positioning applications. We also plan to explore the integration of additional sensors, such as cameras, combined with deep learning, to further improve the efficiency and quality of feature extraction. This extension is aimed at mitigating the issue of positioning failure caused by crop and equipment occlusion in agricultural scenarios.}, archive = {J_TASE}, author = {Yijin Xiong and Xinyu Zhang and Wenju Gao and Yuchao Wang and Jing Liu and Qianxin Qu and Shichun Guo and Yang Shen and Jun Li}, doi = {10.1109/TASE.2024.3451297}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6653--6663}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GF-SLAM: A novel hybrid localization method incorporating global and arc features}, volume = {22}, year = {2025}, }
• (2025). Efficient reinforcement learning method for multi-phase robot manipulation skill acquisition via human knowledge, model-based, and model-free methods. TASE, 22, 6643--6652. (www)

  @article{ , abstract = {A novel efficient reinforcement learning paradigm combining human knowledge, model-based and model-free methods is presented for optimal robot manipulation control during complex multi-phase robot manipulation tasks, e.g., the peg-in-hole tasks with tight fit and nut-and-bolt assembly. Firstly, human demonstration is conducted to collect the data during successful robot manipulation, and manipulation phase estimation method integrating with human knowledge is presented to obtain the higher-level planning of the multi-phase robot manipulation tasks. Typical robot manipulation tasks can usually be decomposed into three types of phases, namely free motion, discontinuous contact, and continuous contact. For phase with free motion, the motion planning method is utilized for generating smooth trajectory. For phase with discontinuous contact in the axes of interest during the pre-manipulation process, the rule-based model-free method, namely the Policy Gradients with Human-Guided Parameter-based Exploration (PGHGPE) method is utilized. For the manipulation phase with continuous contacts, the model-based method is utilized because of its higher sample efficiency. Finally, the simulation and experimental studies verify the effectiveness of the presented algorithm. Note to Practitioners—The important premise for the future robot assistants is that the robots should have certain ability of complex manipulation skill learning. Complex manipulation tasks can be decomposed into multiple stages, and HRL is a suitable method for solving this kind of problems. However, HRL faces the challenge of low computational efficiency. To this end, efficient manipulation skill learning for complex manipulation tasks via human knowledge, model-based and model-free reinforcement learning methods are presented, which improves the efficiency of the skill learning process to a practical level.}, archive = {J_TASE}, author = {Xing Liu and Zihao Liu and Gaozhao Wang and Zhengxiong Liu and Panfeng Huang}, doi = {10.1109/TASE.2024.3451296}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6643--6652}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient reinforcement learning method for multi-phase robot manipulation skill acquisition via human knowledge, model-based, and model-free methods}, volume = {22}, year = {2025}, }
• (2025). Occlusion segmentation: Restore and segment invisible areas for particle objects. TASE, 22, 6631--6642. (www)

  @article{ , abstract = {The occlusion problem has consistently posed a significant challenge in the field of segmentation. Most existing segmentation methods require additional annotations and fail to capture the contour information of occluded regions, thus not truly addressing the occlusion issue. Although segmentation tasks involving particle objects also suffer from occlusion problems, the homogeneity of particle objects offers new possibilities for overcoming this challenge. In this paper, we propose an occlusion segmentation framework for particle objects that does not require additional annotations. This framework only necessitates instance-level segmentation labels to obtain complete contour information of particle objects, including occluded regions. First, we decompose the occlusion segmentation task into a generic instance segmentation task and an occlusion repair task for occluded objects. Then, in order to train the occlusion repair model with only instance segmentation-level labels, we quantitatively analyze the occlusion phenomenon, including the mathematical descriptions of occlusion relationships, degrees, and distributions. Next, we geometrically transform and layer overlay the unobscured samples to construct occlusion samples containing labeling information of the occluded regions. These sample sets are used to train a generative model that predicts the contour information of occluded regions. Finally, we fine-tune or post-process the pre-segmentation model with the particle objects containing restored complete contour information to achieve the final occlusion segmentation. We conducted extensive ablation experiments on both the ore-particle dataset and publicly available cell-particle datasets. The experimental results validate the effectiveness, accuracy, and generalizability of our method. Note to Practitioners—Particle segmentation has been faced with the occlusion problem. In this paper, inspired by the similarity between particle objects, we propose a self-supervised occlusion segmentation framework that does not require additional annotation of occlusion layers. Our approach requires only instance segmentation level annotation without more complex additional manual annotation, which is crucial for practical applications. In addition, we decouple the complex occlusion relation modeling into a binary classification problem without knowing precisely the occlusion hierarchy between particles, which further reduces the difficulty of practical applications. Then, we also propose shading transformations to characterize the inter-particle shading distribution to construct shading sample sets from existing samples. Finally, we use these learned and constructed occlusion sample sets to pre-train the generative model for regenerating the occluded objects to complete the final occlusion segmentation. Although in a generic segmentation task, our approach may have some limitations because the segmented objects may not have an apparent similarity. However, our approach using self-supervision and the objects’ properties provides valuable ideas for solving the occlusion problem. In the future, we will solve the occlusion problem regarding the properties of each class of objects rather than just considering the similarity among particle objects.}, archive = {J_TASE}, author = {Jinshi Liu and Zhaohui Jiang and Weihua Gui and Zhiwen Chen and Chaobo Zhang}, doi = {10.1109/TASE.2024.3450900}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6631--6642}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Occlusion segmentation: Restore and segment invisible areas for particle objects}, volume = {22}, year = {2025}, }
• (2025). Sampled-data connectivity-preserving consensus for multiple heterogeneous euler-lagrange systems. TASE, 22, 6619--6630. (www)

  @article{ , abstract = {This paper aims to establish a sampled-data framework to solve the consensus problem for multiple heterogeneous Euler-Lagrange systems (MHELSs). The systems under consideration have heterogeneous dynamics and limited communication range. Different from the existing works, the common requirement of not allowing edge disconnection has been relaxed. Firstly, a sampled-data virtual system is constructed to provide reference trajectory for the actual system. The virtual systems only exchange data at sampling instants, so that connectivity requirement only needs to be satisfied at these moments. Next, a prescribed performance controller is proposed to track the reference trajectory and ensure the systems satisfy the relaxed connectivity requirement. Furthermore, an event-triggered mechanism is developed to reduce the update frequency of the controller. To illustrate the effectiveness of the proposed framework, two numerical examples are provided. Note to Practitioners—This paper investigates the connectivity-preserving consensus problem for multiple heterogeneous Euler-Lagrange systems. The Euler-Lagrange system can effectively describe various practical systems, such as autonomous vehicles, robotic manipulators, and walking robots. The integration of virtual and physical systems enables the proposed algorithm to adapt well to heterogeneous multiagent systems. The sampling-data interaction mode of the virtual system ensures a reduction of communication pressure among agents in practice. The design of the direction selector and force limiter effectively maintains reliable communication. The introduction of event-triggering mechanisms reduces the update frequency of physical controllers and lowers the performance requirements of the robot actuators. It is worth mentioning that we relax the connectivity requirement of the topology for the first time. Therefore, it is permissible for the distance between two connected robots to exceed the maximum communication range.}, archive = {J_TASE}, author = {Chen Chen and Xingyu Gao and Haiying Zhang and Wencheng Zou and Zhengrong Xiang}, doi = {10.1109/TASE.2024.3450518}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6619--6630}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sampled-data connectivity-preserving consensus for multiple heterogeneous euler-lagrange systems}, volume = {22}, year = {2025}, }
• (2025). Enhancing networked control system resilience to TCP/IP protocol DoS attacks: Performance analysis and intelligent controller design. TASE, 22, 6608--6618. (www)

  @article{ , abstract = {This study examines denial-of-service (DoS) attacks on networked control systems (NCSs) caused by TCP/IP protocol vulnerabilities. It conducts a comprehensive analysis of hacker tactics to uncover vulnerability exploitation techniques. The research reformulates the performance error estimation (PEE) problem, framing it as a search for ellipsoid constraints within a P-dependent set. It introduces the Higher-Order Weight Method (HOWM) to optimize sampling intervals, leveraging the unique properties of zero-order-hold sampling. The study enhances Lyapunov-Krasovskii functions (LKFs) using HOWM by partitioning integral terms and applying quadratic scaling to optimize control algorithms, thereby reducing conservatism and tightening upper-bound criteria. Additionally, it proposes an innovative Integral Event-Triggered Control (IETC) strategy for estimating system performance errors. The performance of the proposed control algorithm is rigorously evaluated through simulations on a complex 2-degree of freedom (DoF) helicopter system (HS). Note to Practitioners—This study delves into DoS attacks targeting NCSs due to TCP/IP protocol vulnerabilities. It conducts an exhaustive analysis of hacker tactics to uncover techniques used to exploit these vulnerabilities. The research reframes the PEE problem, casting it as a search for ellipsoid constraints within a P-dependent ellipsoidal set. It introduces the HOWM to optimize sampling intervals, leveraging the unique characteristics of zero-order-hold sampling. Additionally, the study enhances LKFs using HOWM by partitioning integral terms and applying quadratic scaling to optimize the control algorithm, thereby reducing conservatism and tightening upper-bound criteria. Moreover, it introduces an innovative IETC strategy for estimating system performance errors. The performance of the proposed control algorithm is rigorously evaluated through simulations on a complex 2-DoF HS.}, archive = {J_TASE}, author = {Xiao Cai and Yanbin Sun and Kaibo Shi and Xiangpeng Xie and Yeng Chai Soh and Cheng Qiao and Zhihong Tian}, doi = {10.1109/TASE.2024.3449882}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6608--6618}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing networked control system resilience to TCP/IP protocol DoS attacks: Performance analysis and intelligent controller design}, volume = {22}, year = {2025}, }
• (2025). A self-learning memetic algorithm for human-robot collaboration scheduling in energy-efficient distributed mixed fuzzy welding shop. TASE, 22, 6595--6607. (www)

  @article{ , abstract = {Due to the impact of economic globalization, distributed welding shop has become prevalent in real-world manufacturing systems. Moreover, focusing on human-centric, sustainable and resilient industry, Industry 5.0 puts more emphasis on human-robot collaboration (HRC) for its merit in promoting system flexibility and adaptability. However, owing to the instability of human performance, it becomes necessary to employ fuzzy processing time to simulate practical human production. In the context of Industry 5.0, HRC scheduling in distributed mixed fuzzy welding shop is worth exploring, but no related research on this problem is reported. Thus, to address this research gap, this paper investigates a human-robot collaboration energy-efficient distributed mixed fuzzy welding shop scheduling problem (EDMFWSP-HRC), aiming to minimize makespan and total energy consumption (TEC). To solve this issue, a self-learning memetic algorithm (SLMA) is proposed. In SLMA, a hybrid initialization is designed to yield a high-quality initial population. A genetic operator is proposed to improve the exploration capability. A self-learning variable neighborhood search (SLVNS), which hybridizes Q-learning and VNS, is developed to enhance the exploitation capability. A resource adjustment strategy is presented to further optimize TEC. Additionally, to validate the effectiveness of the proposed SLMA, extensive experimental comparisons with 5 other optimization algorithms are conducted. Experimental results illustrate that SLMA outperforms its competitors. Note to Practitioners—Owing to the widespread presence in manufacturing systems, distributed welding shop has attracted considerable attention in both industry and academia. In the context of Industry 5.0, the incorporation of human-robot collaboration (HRC) scheduling in distributed welding shop can promote system productivity and flexibility. Meanwhile, due to the instability of human performance, employing fuzzy processing time to simulate human production more aligns with the practical manufacturing scenario. Thus, this paper investigates a human-robot collaboration energy-efficient distributed mixed fuzzy welding shop scheduling problem (EDMFWSP-HRC). This problem model can be utilized in many welding manufacturing enterprises with HRC production mode. To solve this problem, we design a self-learning memetic algorithm (SLMA) to minimize both makespan and total energy consumption (TEC). The design of all components in SLMA is based on the characteristics of problem. The SLMA can offer the low-energy and high-efficiency schedules for practitioners. Experimental results verify the effectiveness of the proposed SLMA.}, archive = {J_TASE}, author = {Fei Yu and Chao Lu and Lvjiang Yin and Biao Zhang}, doi = {10.1109/TASE.2024.3448435}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6595--6607}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A self-learning memetic algorithm for human-robot collaboration scheduling in energy-efficient distributed mixed fuzzy welding shop}, volume = {22}, year = {2025}, }
• (2025). Novel adaptive global observer-based sliding mode control of a 2-DOF piezoelectric nanopositioning system. TASE, 22, 6581--6594. (www)

  @article{ , abstract = {This paper proposes a novel adaptive global sliding mode control strategy for a two-degree-of-freedom (2-DOF) piezoelectric nanopositioning system based on the integral extended state observer (IESO) technique. Its uniqueness is that a global robustness property is generated in the whole precision motion control process, which effectively circumvents sensitivity to the perturbations during the reaching phase. First, to generate an accurate disturbance estimation for compensation control, the IESO is constructed by incorporating an integral action into the observer design. Then, a global sliding mode control approach is developed for the nanopositioning system to ensure global robustness against the unknown hysteresis nonlinearity and cross-axis coupling motion. Moreover, an adaptive rule is established for the global sliding mode controller, which does not require a priori knowledge of the estimation error, hysteresis, and cross-coupling nonlinearity in the control design. Both simulation and experimental studies are conducted to demonstrate the effectiveness and superiority of the proposed motion control scheme over existing ones. Note to Practitioners—Nanopositioning systems actuated by piezoelectric actuators impose a great challenge to achieve precision motion tracking control. This work reports a global sliding mode control strategy of 2-DOF piezoelectric nanopositioning systems for the first time. By constructing a global sliding surface and control compensator, the global robustness to unknown hysteresis and cross-axis coupling effect can be improved in the whole process to achieve high-precision motion. To guarantee the estimation performance of unknown nonlinearities, an IESO is designed in this paper to suppress the total disturbance. Additionally, to suppress the effects of estimation errors and achieve high-precision motion tracking, two adaptive rules are established to approximate the upper bounds of estimation errors and the integral term in IESO. Compared to previous works, the proposed control scheme is easy to implement and provides superior results, as verified by simulation and experimental results.}, archive = {J_TASE}, author = {Liheng Chen and Zehao Wu and Qingsong Xu}, doi = {10.1109/TASE.2024.3448249}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6581--6594}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Novel adaptive global observer-based sliding mode control of a 2-DOF piezoelectric nanopositioning system}, volume = {22}, year = {2025}, }
• (2025). Guiding the last centimeter: Novel anatomy-aware probe servoing for standardized imaging plane navigation in robotic lung ultrasound. TASE, 22, 6569--6580. (www)

  @article{ , abstract = {Navigating the ultrasound (US) probe to the standardized imaging plane (SIP) for image acquisition is a critical but operator-dependent task in conventional freehand diagnostic US. Robotic US systems (RUSS) offer the potential to enhance imaging consistency by leveraging real-time US image feedback to optimize the probe pose, thereby reducing reliance on operator expertise. However, determining the proper approach to extracting generalizable features from the US images for probe pose adjustment remains challenging. In this work, we propose a SIP navigation framework for RUSS, exemplified in the context of robotic lung ultrasound (LUS). This framework facilitates automatic probe adjustment when in proximity to the SIP. This is achieved by explicitly extracting multiple anatomical features presented in real-time LUS images and performing non-patient-specific template matching to generate probe motion towards the SIP using image-based visual servoing (IBVS). The framework is further integrated with the active-sensing end-effector (A-SEE), a customized robot end-effector that leverages patient external body geometry to maintain optimal probe alignment with the contact surface, thus preserving US signal quality throughout the navigation. The proposed approach ensures procedural interpretability and inter-patient adaptability. Validation is conducted through anatomy-mimicking phantom and in-vivo evaluations involving five human subjects. The results show the framework’s high navigating precision with the probe correctly located at the SIP for all cases, exhibiting positioning error of under 2 mm in translation and under 2 degrees in rotation. These results demonstrate the navigation process’s capability to accommodate anatomical variations among patients. Note to Practitioners—Compared with traditional freehand ultrasound (US) imaging, robotic ultrasound systems (RUSS) have the potential to largely standardize the US diagnosis outcome caused by varying operator expertise if an inter-patient consistent, automatic standardized imaging plane (SIP) navigation process is available. This paper presents a SIP navigation framework for lung US (LUS) examination, which recognizes anatomical landmarks from the US images and fine-tunes the pose of the US probe so that the landmarks are positioned in accordance with a non-patient-specific template image. The special end-effector, active-sensing end-effector (A-SEE), maintains the probe at an optimal orientation with respect to the body, allowing consistent-quality US images to be acquired throughout the navigation. Unlike previous works, our approach can navigate to complicated SIP containing multiple anatomies with interpretable robot arm motion. We verified our framework’s ability to navigate the probe to the SIP with millimeter-level accuracy under phantom and human experiment settings. While preliminary results demonstrate the framework’s efficacy in guiding the robotic LUS procedure, the performance of the system on other examinations (e.g., liver and thyroid US) involving soft tissues requires further validation. In the future, the framework can be applied in various US examinations by implementing specific anatomical feature detection modules.}, archive = {J_TASE}, author = {Xihan Ma and Mingjie Zeng and Jeffrey C. Hill and Beatrice Hoffmann and Ziming Zhang and Haichong K. Zhang}, doi = {10.1109/TASE.2024.3448241}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6569--6580}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guiding the last centimeter: Novel anatomy-aware probe servoing for standardized imaging plane navigation in robotic lung ultrasound}, volume = {22}, year = {2025}, }
• (2025). Finite-time fault-tolerant consensus of UAVs: A switching event-triggered fuzzy control scheme. TASE, 22, 6554--6568. (www)

  @article{ , abstract = {In finite-time consensus tracking (FCT) missions of unmanned aerial vehicles (UAVs), faults in aircraft actuators are crucial factors causing tracking precision degradation. In order to solve this problem, this paper develops a fuzzy fault-tolerant cooperative control scheme for UAVs to achieve FCT missions with predefined precision. Compared to existing results, the settling time formulation approach of consensus errors is significantly simplified, which facilitates operators to preset the settling time. Then a switching event-triggered mechanism is designed to flexibly change the transmission frequency of control signals, enabling the limited communication resources of UAVs to be reasonably consumed before and after the settling time. Based on the Lyapunov stability theory, it is proved that all signals of systems are bounded, and consensus errors can achieve prescribed precision in a finite time. Moreover, Zeno behavior is excluded. Finally, a simulation study demonstrates the efficiency and feasibility of the designed control scheme under various uncertainties. Note to Practitioners—In the existing FCT control schemes for UAVs, the formulation approaches of settling time depend on the structure of the communication network and the initial states of the aircraft, which hampers these schemes implemented in large-scale UAVs control tasks. How to effectively simplify the formulation approach of settling time has crucial practical significance and research value. In addition, since the communication bandwidth of UAVs is limited, the existing control strategies introduce event-triggered mechanisms to decrease the communication burden. However, adopting a single event-triggered mechanism cannot flexibly change the transmission frequency of control signals, which may make the limited communication resources fail to be reasonably consumed before and after the settling time. In order to solve the above problems, this paper develops a switching event-triggered FCT control scheme for UAVs, where the designed trigger mechanism can flexibly change the transmission frequency of control signals. Meanwhile, the settling time can be directly formulated following the wishes of operators.}, archive = {J_TASE}, author = {Hongjing Liang and Lei Chen and Xinsong Yang and Tingwen Huang}, doi = {10.1109/TASE.2024.3447277}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6554--6568}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time fault-tolerant consensus of UAVs: A switching event-triggered fuzzy control scheme}, volume = {22}, year = {2025}, }
• (2025). Adaptive neural prescribed performance control for non-triangular structural stochastic highly nonlinear systems under hybrid attacks. TASE, 22, 6543--6553. (www)

  @article{ , abstract = {In this article, the adaptive neural network prescribed performance control issue is followed with interest for a class of non-triangular structural stochastic highly nonlinear interconnected systems under hybrid attacks, that is, injection attack and deception attack. Unlike the previous achievements, the original states of the system are unknown under the influence of hybrid attacks, and it is particularly difficult to handle highly nonlinear interconnection functions of non-triangular structure in the system. Hence, an effective variable separation method based on adaptive compensation technology is constructed, such that the complex nonlinear functions can be separated and further simplified, and then approximated by neural network. By utilizing a prescribed performance function, the tracking error of the system can converge in a fixed time based on the Lyapunov stochastic stability theory and the framework of the backstepping technology. Ultimately, the reliability and effectiveness of the proposed control strategy can be verified by a practical example. Note to Practitioners—With the rapid increase of the complexity of the practical engineering systems, such as manipulator system, automated highway systems and so on, it is difficult to be described by the common nonlinear system model. Moreover, these networked practical systems are often plagued by external attacks, which seriously affects the performance of the system and makes it difficult to design the control scheme. Therefore, this paper investigates a class of stochastic highly nonlinear interconnected system model which is more in line with the practical demand. On this basis, in order to improve the control efficiency of the practical engineering systems and solve the influence of external attacks, a fixed-time prescribed performance adaptive control strategy is constructed to satisfies the control requirement of the complex stochastic system. The proposed control scheme is more convenient for physical implementation, and is also applied to the single-link manipulators system to prove its feasibility.}, archive = {J_TASE}, author = {Zhechen Zhu and Quanxin Zhu}, doi = {10.1109/TASE.2024.3447045}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6543--6553}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive neural prescribed performance control for non-triangular structural stochastic highly nonlinear systems under hybrid attacks}, volume = {22}, year = {2025}, }
• (2025). A hierarchical framework-based coordinated optimization of building HVAC systems and EVs. TASE, 22, 6530--6542. (www)

  @article{ , abstract = {The demands of electric vehicles (EVs) and building heating, ventilation, and air conditioning (HVAC) systems have considerable flexibility. Their flexibility is influenced by occupants’ behavior resulting in huge complementary and dispatchable capability. Therefore, the coordination of EVs and HVAC systems holds significant potential for optimizing the building demand profiles and energy cost under time-of-use (TOU) tariffs. However, solving the coordinated problem in practice still faces the challenges in computational complexity and global information requirement due to the spatio-temperal coupling constraints. In this paper, a mixed-integer linear programming model is developed to formulate the multi-building energy system with EVs and the impact of occupants’ behavior on the demand and flexibility of the system. The problem is converted into a three-level structure using Lagrangian relaxation framework. A dynamic programming-based Lagrangian relaxation (DPLR) algorithm is developed to independently solve all sub-problems of the three-level structure in a decomposition and coordination way while avoiding the iterative computation between the middle and lower level. The numerical results show the developed method can obtain a near-optimal solution in an efficient way without perceivable degradation in accuracy, which is 4% worse but at least three times faster, compared to the existing centralized algorithm. Note to Practitioners—The escalating demand for EVs and HVAC systems results in increased energy costs and challenges to existing power systems, such as frequency deviations and higher peak loads. Therefore, this paper focuses on the coordinated optimization of the EV charging and building HVAC system operation, while considering rooftop photovoltaic generation supply within the system. Optimal coordination of the above system can effectively reduce energy costs under TOU tariffs and enhance the ability to utilize renewable energy sources. However, solving the coordinated optimization problem still faces difficulties since the computational complexity will exponentially grow with increasing problem scale due to the spatio-temperal coupling between the demand of EVs and HVAC systems. Therefore, a DPLR algorithm is developed. There is a central coordinator that collects the demand information of EVs and HVAC systems and broadcasts the coordination information calculated according to the demand information. Every single EV and HVAC system can make decisions based on the coordination information independently. The DPLR algorithm decouples EVs and HVAC systems and avoids the curse of dimensionality. It has great improvement on computational efficiency and global information requirement reduction. The computational efficiency and effectiveness of the developed method is verified through multi-scale case studies. The numerical results show that compared with independent optimization of EVs and HVAC systems, coordinated optimization can reduce over 44% of the energy costs.}, archive = {J_TASE}, author = {Haoming Zhao and Zhanbo Xu and Jiang Wu and Fengxia Liu and Xiaohong Guan}, doi = {10.1109/TASE.2024.3446908}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6530--6542}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical framework-based coordinated optimization of building HVAC systems and EVs}, volume = {22}, year = {2025}, }
• (2025). Security performance of MOSMLFC power system under historical-frequency-triggered DoS attacks. TASE, 22, 6518--6529. (www)

  @article{ , abstract = {A memory output sliding mode load frequency control (MOSMLFC) strategy is proposed for multi-area interconnected power systems under historical-frequency-triggered denial-of-service (DoS) attacks. Due to the use of the open network, the multi-area power system is prone to cyber-attacks. Different types of attack models have been built to describe the actual attack behavior, so that effective strategies can be quickly formulated in the event of an attack. Therefore, a historical-frequency-triggered DoS attacks model is presented from the perspective of attackers, with the aim of destroying the stable state of the multi-area power system. It is assumed that attackers determine the timing of DoS attacks by monitoring the operational status of multi-area power systems and designing the triggering condition with historical frequency. A MOSMLFC strategy is investigated to ensure the security performance of multi-area power systems under historical-frequency-triggered DoS attacks, which applies the memory output information of the power system to realize the controller design. The security condition of multi-area power systems under historical-frequency-triggered DoS attacks is obtained by Lyapunov’s theorem and linear matrix inequality (LMI). Numerical examples are tested over the IEEE 10-generator 39-bus system and the results prove the usefulness and superiority of the proposed method. Note to Practitioners—Load frequency control is widely applied in multi-area power systems to achieve a balance between the load demand and generation. Frequent cyber-attacks are a threat to the normal operation of the power system. It is therefore necessary to develop appropriate strategies to defend against cyber-attacks. So far, there have been many different forms of cyber-attacks. This has prompted defenders to build different types of attack models to describe the actual attack behavior in order to preemptively formulate appropriate defensive strategies. Smart attacker may notice that certain characteristics of the target system are important, such as the power system frequency. This motivates us to propose a historical frequency-triggered DoS attack model that contributes to a deep understanding of the impact of cyber-attacks on the power system. We propose a unique sliding mode control approach to ensure the stable performance of power system state and output simultaneously.}, archive = {J_TASE}, author = {Siwei Qiao and Xinghua Liu and Gaoxi Xiao and Meng Zhang and Yu Kang and Shuzhi Sam Ge}, doi = {10.1109/TASE.2024.3446883}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6518--6529}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Security performance of MOSMLFC power system under historical-frequency-triggered DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Activator-based economical distributed fault-tolerant control against possible actuator outages with guaranteed performance. TASE, 22, 6503--6517. (www)

  @article{ , abstract = {This paper concerns a distributed control strategy to deal with possible actuator outages of multi-agent systems (MASs). With the designed strategy, the control process is monitored by a fault detection system (FDS). Once the actuator of one agent in use partially or completely fails, the FDS will detect the fault timely and a controller activator will produce its effect. By this means, the healthy actuator can replace the faulty one for control at the delicately designed instants, and the prescribed performance of the MAS can always be satisfied in the presence of possible actuator outages with the help of prescribed performance functions. Moreover, to achieve more diversified constraints to meet wider requirements in practice, the proposed strategy is further extended to combine with a time-varying barrier Lyapunov function so that the time-varying output constraint can be achieved, which makes the designed fault-tolerant control more flexible and sensitive in dealing with failures. With the proposed distributed control method, uncertain actuator failures, especially the actuator outage, can be addressed in the presence of disturbances and uncertain dynamics. Since the failure is allowed to occur multiple times with only one actuator operating for control at any time, the designed approach can be more economical in energy saving compared with the existing methods. Numerical simulations are provided for multiple agents to verify the effectiveness of the proposed algorithm. Note to Practitioners—In this paper, a control strategy is proposed to handle the actuator failure of a class of MASs. To deal with possible actuator outages, which pose a threat to system safety, actuator redundancy is introduced in the proposed strategy. By developing a controller activator to specify the activation time of healthy actuators with the help of a designed FDS, the prescribed performance of the agents can always be guaranteed and the tracking error can be constrained within specified time-varying function curves as expected. Since the proposed strategy enables the successive triggering of actuators, only one actuator works at any instant, which contributes to its energy-saving efficiency. Since the proposed strategy can accommodate possible actuator outages in an economical way with guaranteed prescribed performance, it can be used in various industrial scenarios where high safety of agents is required, for instance, multiple aircrafts, multiple high-speed trains, and multiple industrial robot systems.}, archive = {J_TASE}, author = {Xueyan Xing and Guoqiang Hu and Yingchong Ma and Rahul Singh}, doi = {10.1109/TASE.2024.3446862}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6503--6517}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Activator-based economical distributed fault-tolerant control against possible actuator outages with guaranteed performance}, volume = {22}, year = {2025}, }
• (2025). Recursive learning based smart energy management with two-level dynamic pricing demand response. TASE, 22, 6492--6502. (www)

  @article{ , abstract = {Due to dynamic characteristic of demand response and stochastic nature of power generation, it brings great challenge to smart energy management. In this paper, a demand response model is created with two-level dynamic pricing transaction among grid operator, service provider and customers, which also involves customers’ active participation with load shifting issue. To effectively control system load on the demand side, an improved deep reinforcement learning approach is proposed with a recursive least square (RLS) technique to deal with the dynamic pricing demand response problem, which accelerates the on-line training and optimization efficiency. On the power generation side, a probabilistic penalty-based boundary intersection (PBI) based multi-objective optimization algorithm is improved to optimize the economic cost, emission rate and statistic voltage stability index (SVSI) simultaneously with generated stochastic scenarios, which can ensure energy conservation and environmental protection, as well as system security. The case results reveal that the proposed two-level optimization strategy successfully deals with energy management with dynamic pricing demand response.Note to Practitioners—This paper is motivated by solving stochastic energy management issue of isolated power system with dynamic pricing demand response. Those existing methods merely focus on the load demand or power generation side, and the methods for demand response issue lacks efficient on-line learning ability, while this work proposes a recursive least square based deep reinforcement learning approach to tackle with the two-level dynamic pricing demand response issue, scenario based PBI multi-objective optimization is proposed to solve the power dispatch issue on power generation side, and the numerical analysis results suggest that the proposed optimization strategy can deal with the whole energy management issue well. The future work will focus on the dynamic power-load coordination in the energy management issue.}, archive = {J_TASE}, author = {Huifeng Zhang and Jiapeng Huang and Dong Yue and Xiangpeng Xie and Zhijun Zhang and Gerhard P. Hancke}, doi = {10.1109/TASE.2024.3446849}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6492--6502}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Recursive learning based smart energy management with two-level dynamic pricing demand response}, volume = {22}, year = {2025}, }
• (2025). Robust task allocations by distributing the risk among agents: Theory and algorithms. TASE, 22, 6475--6491. (www)

  @article{ , abstract = {We address the problem of generating robust solutions for the makespan minimization problem on identical agents (parallel machines), under the assumption that only interval bounds of processing times are known. While there are various concepts of robustness in the literature, we prove using pathological examples that any of these criteria may result in allocations with undesirable characteristics. We identify key properties that must be satisfied for a solution to be considered truly robust. Given a set of jobs with associated loads and uncertainties, it is shown that an allocation that balances loads and uncertainties simultaneously is extremely robust and satisfies multiple other existing criteria of robustness within an acceptable approximation factor. Thus, robustness is achieved by distributing the uncertainty/risk among the agents along with the load. The problem of finding a robust allocation is reduced to a bi-criteria two-dimensional load balancing problem, with the two dimensions being the load and the uncertainty. We prove that for the case with 2 agents, an allocation that satisfies a 1.5-approximation on both dimensions simultaneously always exists and can be found efficiently, and is also the best possible guarantee. For the general case with any number of agents, we prove that an allocation that satisfies a 2-approximation on one dimension and a 2.5-approximation on the other always exists and can be found in pseudo-polynomial time. The approximation algorithms presented in this paper are complemented by interesting existential and structural results and contribute to the vector scheduling literature for two dimensions as well. Finally, an extensive numerical analysis is presented, where we demonstrate our algorithms’ near-optimal performance and ability to generate allocations that satisfy multiple criteria of robustness simultaneously in a short amount of time. Note to Practitioners—This paper introduces a simple and provably effective methodology for generating robust allocations in the context of the makespan minimization problem, a critical challenge in operational management that significantly impacts the efficiency and productivity of various industries. We demonstrate using counter-examples that traditional concepts of robustness such as worst-case makespan and min-max regret can lead to overly conservative and practically inefficient allocations, even when solved optimally. Following this, it is shown that an allocation that balances loads and uncertainties simultaneously is extremely robust and satisfies multiple other existing criteria of robustness within an acceptable approximation factor. This leads to a more attractive and practical concept of robustness. Efficient, fast, and provably good algorithms are presented that solve a 2D Load Balancing problem and generate allocations that are balanced with respect to both the loads as well as uncertainties for a large percentage of the possible scenarios. Numerical results provide practitioners confidence in our approach. The algorithm further classifies the jobs as critical and non-critical based on their completion times and uncertainties in a way that leads to provably good allocations. This classification can be further used to obtain intuition about the problem, thus providing managerial insights.}, archive = {J_TASE}, author = {Raunak Sengupta and Rakesh Nagi and Ramavarapu S. Sreenivas}, doi = {10.1109/TASE.2024.3446456}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6475--6491}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust task allocations by distributing the risk among agents: Theory and algorithms}, volume = {22}, year = {2025}, }
• (2025). Generation & clinical validation of individualized gait trajectory for stroke patients based on lower limb exoskeleton robot. TASE, 22, 6463--6474. (www)

  @article{ , abstract = {Existing research suggests that lower limb exoskeleton robots, when used for rehabilitation training based on the pre-stroke gait trajectories of stroke patients, may be more beneficial for gait rehabilitation. However, it’s challenging to obtain such personalized trajectories for specific patients. Therefore, this hypothesis is difficult to be verified. This paper introduces an Individualized Gait Trajectory Generation (IGTG) method based on Fast Fourier Transform (FFT) to approximate and regress pre-stroke gaits, along with conducting clinical rehabilitation validation trials. Initially, human gait trajectories are described using Fourier coefficients to construct gait features. Subsequently, a probabilistic mapping between these gait features and physical body parameters is established. Then, personalized gait trajectories are obtained by applying the inverse Fourier transform to the predicted gait features. The application of fast Fourier transform can reduce the number of the regression data points needed, decrease dependency on large datasets, and enhance the systematic robustness. This algorithm is trained using body parameters and gait trajectories collected from 128 healthy subjects. The algorithm is further applied to generate specific personalized trajectories for the 9 stroke patients. Clinical trial results indicate that rehabilitation training using these individualized gait trajectories reduces blood oxygen saturation (SpO2) and heart rate (HR) by up to 66.67% and 69.23% respectively compared to training with fixed trajectories. Note to Practitioners—The main purpose of this paper is to solve gait trajectories mismatch problem when different stroke patients use lower limb exoskeleton robot for rehabilitation training. Variations in body factors among individuals lead to different gait trajectories including walking speed, gender, age, and other anthropometric parameters. Therefore, this paper introduces a novel Individualized Gait Trajectory Generation (IGTG) method to generate suitable gait trajectories for stroke patients with different body characteristic parameters when taking gait rehabilitation training with a lower limb exoskeleton robot. The detailed methodology introduction and a full analysis of experimental results are also given. Finally, clinical experiments involving stroke patients were conducted to demonstrate the feasibility and effectiveness of the presented method.}, archive = {J_TASE}, author = {Shisheng Zhang and Yang Zhang and Mengbo Luan and Ansi Peng and Jing Ye and Gong Chen and Chenglong Fu and Yuquan Leng and Xinyu Wu}, doi = {10.1109/TASE.2024.3445886}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6463--6474}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Generation & clinical validation of individualized gait trajectory for stroke patients based on lower limb exoskeleton robot}, volume = {22}, year = {2025}, }
• (2025). Set stabilization and robust set stabilization of periodically time-varying boolean control networks. TASE, 22, 6451--6462. (www)

  @article{ , abstract = {This paper investigates set stabilization and robust set stabilization of periodically time-varying Boolean control networks (PTVBCNs) by combining state-flipping mechanism and state feedback control. Firstly, two algorithms are proposed to check whether systems with or without disturbance inputs can be stabilized to the set under the hybrid control. Next, the state feedback matrices are designed by a constructive method such that systems can be stabilized to the given set for two cases. In addition, some algorithms are proposed to find the corresponding flip sequence with minimum cardinality that drives any initial state to the given set in two cases. Finally, some examples are given to illustrate the good performance of the obtained results. Note to Practitioners—In this paper, we address the problems of set stabilization and robust set stabilization of periodically time-varying Boolean control networks using the semi-tensor product. This provides essential theoretical support for the interventional treatment of gene regulatory networks, aiming to transfer all states in the network to a healthy state, which is significant for gene therapy. In gene regulatory networks, we need to consider whether the system’s evolution can transfer all states to a healthy state and how to apply external effects to states that cannot achieve a healthy condition through the system’s evolution alone. Choosing when and how to exert external influence can achieve our goals, which is the significance of our research. This paper provides effective methods to solve the above problems.}, archive = {J_TASE}, author = {Meiling Su and Peilian Guo}, doi = {10.1109/TASE.2024.3445876}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6451--6462}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Set stabilization and robust set stabilization of periodically time-varying boolean control networks}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy iterative learning control of constrained systems with arbitrary initial state errors and unknown control gain. TASE, 22, 6439--6450. (www)

  @article{ , abstract = {An adaptive fuzzy iterative learning control(AFILC) method is presented to address the state tracking issue of constrained systems with arbitrary initial state errors and unknown control gain. A novel desired error trajectory is systematically developed in the polynomial form to relax the identical initial condition, which allows for arbitrary setting of initial values for all the system state errors. The proposed desired error trajectory can also relax the iteration-invariance restriction on the reference signals due to the independence of the reference trajectories. An asymmetric integral fractional barrier Lyapunov function is developed, keeping the tracking error within the preassigned boundary. Moreover, there is no need to estimate the unknown control gain function in the controller design, reducing computation burden. Numerical simulations and experiments in the permanent magnet synchronous motor experimental platform are provided to illustrate the efficacy of the proposed method. Note to Practitioners—Most practical systems often perform repetitive tasks in industrial processes, such as the repetitive handling process of manipulators, and the rotation process of motors, etc. Iterative learning control method is model independent, and fully utilizes the repetitive characteristics during system operation. However, due to irregular initial state drifts caused by locating operations at different iterations, the identical initial condition is often violated in practical iterative learning control applications. This paper presents an adaptive fuzzy iterative learning control method to address the state tracking issue of constrained systems with arbitrary initial state errors and unknown control gain. The problem of inconsistent initial values is addressed by designing a desired error trajectory in the polynomial form, such that arbitrary setting of initial values for all the system state errors is allowed. For safe operation in practice, an asymmetric integral fractional barrier Lyapunov function is developed to keep the tracking error within the preassigned boundary. The satisfactory experimental results on the permanent magnet synchronous motor experimental platform also demonstrate the practical effectiveness of the proposed method.}, archive = {J_TASE}, author = {Huihui Shi and Qiang Chen and Yihuang Hong and Xianhua Ou and Xiongxiong He}, doi = {10.1109/TASE.2024.3445670}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6439--6450}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy iterative learning control of constrained systems with arbitrary initial state errors and unknown control gain}, volume = {22}, year = {2025}, }
• (2025). Configuration optimization of a dual-arm reconfigurable space robot based on closed-chain inertia matching. TASE, 22, 6421--6438. (www)

  @article{ , abstract = {The dual-arm space robot usually forms a closed-chain constraint system with a target through collaborative operations when performing tasks. Most previous related research has focused on the performance of open-chain robots themselves. Studying the manipulation performance of a closed-chain robot system is of great significance. This article proposes a reconfigurable space robot (RSR) system for on-orbit servicing. A graph theory based framework for the automatic generation of reconfigurable robot models is proposed to address the characteristics of variable topology structures. Meanwhile, a new index – the closed-chain inertia matching index is proposed to evaluate its configuration effectively. Compared with traditional dynamic manipulability ellipsoid (DME) and manipulating force ellipsoid (MFE), the effectiveness of the proposed closed-chain inertia matching ellipsoid (IME) is verified. Compared with the DME and MFE, the IME effectively considers the influence of load and can effectively express the dynamic torque force/torque transmission efficiency from the joint actuator to the load in a closed-chain system. The IME can not only be used to determine the optimal joint configuration under a specific closed-chain configuration, but also to determine the optimal nonisomorphic configuration of a reconfigurable robot. Finally, the results of configuration optimization of the closed-chain dual-arm reconfigurable space robot are given.Note to Practitioners—This paper aims to study the determination of the optimal configuration for the RSR during task execution. Due to its various configurations, the RSR has a complex modeling process. This article proposes a modeling framework for automatically generating RSR models that can effectively achieve automatic modeling of various configurations. In addition, most of the previous related research focused on the performance of open-chain robots themselves, ignoring the impact of load. This article proposes a new closed-chain system performance indicator to evaluate the robot configuration. The proposed evaluation indicator effectively considers the influence of load, which is of practical significance to improve the efficiency of task execution. The effectiveness of the proposed index in determining the optimal configuration is demonstrated by simulation verification. In the future, we will focus on multi-arm collaborative operations under a specific configuration.}, archive = {J_TASE}, author = {Zhihui Xue and Jinguo Liu and Yangmin Li and Junsen Liu}, doi = {10.1109/TASE.2024.3445453}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6421--6438}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Configuration optimization of a dual-arm reconfigurable space robot based on closed-chain inertia matching}, volume = {22}, year = {2025}, }
• (2025). A time-delay modeling approach for data-driven predictive control of continuous-time systems. TASE, 22, 6408--6420. (www)

  @article{ , abstract = {This paper aims to predict future optimal control inputs for unknown continuous-time linear systems. In contrast to most existing approaches that generate control input without observation loss, the proposed control scheme analyzes causality between observation and sequence feedback from the time-delay series model, allowing for the existence of observation loss. These time-delay series can be viewed as multiplayer games over a temporal scale, then a temporal game-theoretic approach ensures system stability and performance. By integrating the Bellman principle, a data-driven adaptive dynamic programming algorithm is proposed to avoid system knowledge. Furthermore, the designed parallel data-driven predictive algorithm reduces the computational complexity. Finally, the applicability and effectiveness of the methodology are demonstrated through numerical simulations and practical experiments. Note to Practitioners—This paper mainly concerns the predictive control of unknown continuous systems, which suffer from unknown dynamics and state observation loss. The proposed methods are suitable for weak information feedback and dynamically changing scenarios, such as autonomous driving in low visibility scenarios and endoscopic surgical robots. Most of the current processing methods are model-driven, which makes them unsuitable for unknown system dynamics changes. To address this issue, we propose a time-delay switched strategy for control prediction with stability and optimality guarantees. The practical application can be divided into three parts: i) Data collection: Collect historical multi-intervals accumulated inputs and outputs data, the amount of data should reach the requirement of the full rank of the data matrix; ii) Iterative learning: the optimal control strategy is learned from the historical data through an adaptive dynamic programming method; iii) Deployment: Control intervals are extended through temporal time-delay feedback on historical state trajectory, and length trigger conditions will switch these feedbacks logically for actuators. Finally, the Quanser QBot 2e robot is used as a demonstration example.}, archive = {J_TASE}, author = {Juan Liu and Xindi Yang and Hao Zhang and Zhuping Wang and Huaicheng Yan}, doi = {10.1109/TASE.2024.3445335}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6408--6420}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A time-delay modeling approach for data-driven predictive control of continuous-time systems}, volume = {22}, year = {2025}, }
• (2025). A fixed-time consensus control with prescribed performance for multi-agent systems under full-state constraints. TASE, 22, 6398--6407. (www)

  @article{ , abstract = {This paper investigates a fixed-time consensus control problem of nonlinear multi-agent systems under full-state constraints. First, by designing corresponding constraint functions for system transformation, state-dependent asymmetric time-varying constraints are realized. The feasibility conditions of the system are eliminated, and the requirements on the constraint boundary are relaxed. Meanwhile, a prescribed performance function is designed for the constraints on synchronization deviation, which helps to improve the transient and steady-state performance of the system and ensure rapid consensus convergence on a fixed-time framework. Additionally, considering the frequent communication between the controller and actuator and to decrease the controller update frequency to save system bandwidth, an adaptive threshold event-triggered mechanism is developed. A dynamic parameter is introduced into the triggered mechanism to adjust the triggered threshold, thereby overcoming the issue that static parameters might cause excessive or insufficient event-triggered and avoiding Zeno behavior. Finally, the effectiveness of the proposed strategy is verified through simulation. Note to Practitioners—In complex modern engineering systems, the consensus control of multi-agent systems has become a research hotspot. In practical industrial applications, given the requirements for safety and production efficiency, it is crucial for systems to effectively constrain states and ensure performance. This study employs the prescribed performance strategy within the fixed-time framework to construct the control method. The constraints of the system’s full states are achieved by transforming a constrained system into an unconstrained one. Meanwhile, the use of event-triggered mechanisms saves communication resources. The proposed method not only ensures rapid consensus convergence under full-state constraints but also enhances the control performance of multi-agent systems, closely connected to the needs of practical applications. Future research will continue to investigate how to apply it to practical engineering applications.}, archive = {J_TASE}, author = {Shangbin Long and Weicong Huang and Jianhui Wang and Jiarui Liu and Yixiang Gu and Zian Wang}, doi = {10.1109/TASE.2024.3445135}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6398--6407}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A fixed-time consensus control with prescribed performance for multi-agent systems under full-state constraints}, volume = {22}, year = {2025}, }
• (2025). H∞ control for interconnected systems with unknown system dynamics: A two-stage reinforcement learning method. TASE, 22, 6388--6397. (www)

  @article{ , abstract = {This paper investigates the optimal control problem for interconnected systems with unknown system dynamics through a two-stage reinforcement learning method. First, to address the impact of interconnection term, the optimal control problem for interconnected systems is transformed into obtaining the solution of game algebraic Riccati equations within the framework of $H_{\infty }$ control method. Furthermore, existing optimal control approaches for interconnected systems necessitate precise knowledge of system dynamics, which is difficult to obtain accurately or involves high costs. Thus, we introduce a two-stage reinforcement learning method. The admissible control policies are obtained using the homotopy-based iteration method in the first stage. Then, the optimal control policies are obtained through the policy iteration method in the second stage. The two-stage method not only eliminates the requirement for system dynamics and initial admissible control policies but also ensures convergence speed and accuracy, significantly enhancing its practicality. Finally, a two-machine power system example is provided to validate the feasibility of the two-stage method. Note to Practitioners—Interconnected systems, a class of systems composed of multiple local subsystems, find wide applications in various fields such as power systems, transportation networks, and spatially interconnected systems. Particularly, the optimal control problem of interconnected systems has gradually become a focal point of current research. However, the current research on the optimal control problem of interconnected systems is still constrained by the system dynamics and the initial stability of the system. To relax these limitations, this paper introduces a two-stage method. A homotopy-based iteration approach is employed to obtain control policy and interconnection policy that make the system closed-loop stable, thus achieving the optimal solution. Furthermore, the data-driven approach overcomes the limitations imposed by system dynamics. The feasibility of the two-stage method is illustrated by a two-machine power system model.}, archive = {J_TASE}, author = {Jinxu Liu and Hao Shen and Jing Wang and Jinde Cao and Leszek Rutkowski}, doi = {10.1109/TASE.2024.3444463}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6388--6397}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {H∞ control for interconnected systems with unknown system dynamics: A two-stage reinforcement learning method}, volume = {22}, year = {2025}, }
• (2025). Nonlinear MPC on parallel parking for autonomous vehicles under state-dependent switching. TASE, 22, 6377--6387. (www)

  @article{ , abstract = {In this paper, a framework integrating planning and control is designed to deal with the parallel parking problem of autonomous vehicle. In the planning section, efficient planning is achieved by directly setting target points. In the control part, it is mainly based on nonlinear model predictive control (NMPC) scheme for implementation. Firstly, the parking task is divided into four stages by setting different target points to solve the non-convex state constraint problem in parallel parking. Then, a state-dependent switching law is designed to describe the switching process in different stages based on the different constraints present. In addition, the terminal constraints and terminal cost functions of NMPC scheme are removed to increase the solvability of optimization problems and improve the smoothness of parallel parking for autonomous vehicle. Asymptotic stability and recursive feasibility of autonomous vehicle are ensured. The experimental results show that the NMPC scheme is effective in parallel parking of autonomous vehicle. Note to Practitioners—The motivation of this paper is to solve the problem of parallel parking for autonomous vehicles, with a focus on the design of planning and control framework. In recent years, automatic parallel parking technology has received widespread attention. However, there are still two issues that need further consideration: 1. Most existing planning and control schemes rely on nonlinear optimization processes, but multi-layer nonlinear optimization processes inevitably affect the efficiency and real-time performance of automatic parallel parking schemes. 2. The special construction of parallel parking lots makes it difficult to solve nonlinear optimization problems because the constraint domain is non-convex. To overcome these issues, we first divide the parallel parking process into several stages and directly plan the target points for each stage based on the characteristics of the parallel parking lots, rather than planning the complete parallel parking path. Then, with the help of the prediction and rolling optimization characteristics of the NMPC scheme, the optimal approach to each target point is achieved. In addition, the classic NMPC scheme has been improved (by removing terminal constraints and terminal cost functions) to further enhance its solvability and real-time performance. It is worth noting that the framework of this paper is applicable to most automatic parallel parking needs, as the characteristics of standard parallel parking slots (including size, exit and entrance requirements, safety distance requirements, etc.) are consistent with the planning requirements of the target points in this paper. Moreover, most vehicle models are known, which enables the NMPC scheme in this paper to adapt to different types of autonomous vehicles. In future work, we will further explore the possibility of extending our algorithm to more complex and chaotic environments (including but not limited to the presence of obstacles and external interference) for automatic parking systems.}, archive = {J_TASE}, author = {Hai Zhao and Hongjiu Yang and Zhengyu Wang and Yuanqing Xia}, doi = {10.1109/TASE.2024.3443848}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6377--6387}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Nonlinear MPC on parallel parking for autonomous vehicles under state-dependent switching}, volume = {22}, year = {2025}, }
• (2025). Output feedback bounded control for unidirectional input constrained hysteretic systems with application to piezoelectric-driven micropositioning stage. TASE, 22, 6365--6376. (www)

  @article{ , abstract = {In this study, an observer-based output feedback bounded control strategy is investigated for states unmeasurable nonlinear systems subjected to actuator hysteresis and unidirectional input constraints. The inevitable complex actuators restrictions and hysteresis that exist in practice are considered and transformed by applying a new model transformation operations. To cope with the challenge introduced by the coupling of control inputs and an actuator nonlinear term, a new fuzzy state observer is constructed by introducing a signal filter so that the unmeasurable states and generalized uncertainties, including the hysteresis term, disturbances, and uncertainties, can be estimated synchronously. An output feedback bounded control strategy is proposed that achieves high-precision positioning using a performance-prespecified dynamic surface controller and guarantees that the control signal is bounded within the input constraints by constructing the logarithm Lyapunov functions. The stability of the closed-loop system is proven. Furthermore, the investigated scheme is validated on the piezoelectric micropositioning stage, which is a mechanical system with unknown actuator hysteresis and unidirectional input constraints. Note to Practitioners—This study is motivated by the problem of bounded motion control for smart material-driven micropositioning stages with hysteresis and unidirectional input constraints. The development of micro/nano technology places higher requirements on micro/nanopositioning control. As one of the most commonly used smart material-driven micropositioning stages, piezoelectric micropositioning stage is widely used in mechanical engineering, aerospace technology, bioengineering and other fields. However, influenced by the inherent properties of materials, piezoelectric actuators have complex hysteresis nonlinearity and are subject to unidirectional input constraints, which impair the positioning accuracy and the equipment operation safety. Existing control methods for hysteretic systems are either limited in control performance due to the hysteresis modeling accuracy or difficult to be applied to practical systems with only measurable outputs, and all of existing methods ignore the effect of unidirectional input constraints on the control performance. This study proposes an output feedback bounded control strategy that eliminates the negative effects of hysteresis without constructing an inverse hysteresis model, and realizes high-precision positioning control with prespecified performance, where the bounds of control input is known a priori and satisfy the input constraints.}, archive = {J_TASE}, author = {Linlin Nie and Miaolei Zhou and Xiuyu Zhang and Chun-Yi Su}, doi = {10.1109/TASE.2024.3443457}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6365--6376}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Output feedback bounded control for unidirectional input constrained hysteretic systems with application to piezoelectric-driven micropositioning stage}, volume = {22}, year = {2025}, }
• (2025). Controllers for multiagent systems with input amplitude and rate constraints and their application to quadrotor rendezvous. TASE, 22, 6354--6364. (www)

  @article{ , abstract = {This paper addresses the consensus issue of multiagent systems with both input amplitude and rate constraints. We propose simple yet effective distributed control algorithms that integrate a velocity damping term with nonlinear saturated functions for both undirected and directed graphs. Leveraging the interplay between Barbalat’s lemma and graph theory, we show that all agents can achieve consensus without violating predefined input amplitude and rate constraints through the presented control algorithms. Moreover, we employ the developed framework to solve the rendezvous control problem of quadrotor unmanned aerial vehicles (UAVs) with motion limits. To illustrate and validate our proposed approach, we conduct extensive simulations and comparative experiments. Note to Practitioners—Most existing control methods for multiagent systems achieve consensus but neglect the constraints on the amplitude and rate of the control signal. However, in practice, the control signals are invariably subject to limitations in their amplitude and rate due to factors such as actuator saturation, considerations for ride comfort, and actuator wear. This neglect leads to a degradation in system performance and in severe cases results in the loss of closed-loop stability. This work primarily focuses on developing new control methods that can achieve consensus without violating the predefined input amplitude and rate limitations. The experiments on rendezvous control of quadrotor UAVs show the practical applicability of the presented algorithms, which yield satisfactory control performance as verified by theoretical analysis. This research contributes to the advancement of distributed control for multiagent systems, particularly in scenarios where input constraints are a critical consideration.}, archive = {J_TASE}, author = {Gang Wang and Zongyu Zuo and Peng Li and Yantao Shen}, doi = {10.1109/TASE.2024.3443402}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6354--6364}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Controllers for multiagent systems with input amplitude and rate constraints and their application to quadrotor rendezvous}, volume = {22}, year = {2025}, }
• (2025). Optimal gait planning and thruster force allocation for rough terrain climbing of an underwater hexapod robot. TASE, 22, 6340--6353. (www)

  @article{ , abstract = {The underwater hexapod robot, driven by eight thrusters and six C-shaped legs, can perform complex locomotion tasks such as climbing rough terrain. Unlike conventional point-contact legs, the C-shaped leg rolls on the terrain. The rolling fashion brings significantly complex loop-closure kinematic constraints and complicates the finding of feasible gaits. In addition, when C-shaped legs roll on rough terrain, their contact condition will change in real-time, leading to time-varying contact force, which may result in the leg slipping or even the robot falling. To address the two issues, we propose gait planning and thruster force allocating methods for rough terrain climbing. First, we propose a sampling-based gait planner that extends random trees in task space and finds feasible gaits to fulfill the loop-closure kinematic constraints, which avoids designing the complex sampling and steering procedures in an implicitly-defined manifold. Second, by designing a gait interval-based cost function, we propose an optimal sampling-based planner to find smooth climbing gaits. Third, by establishing a simplified single rigid body (SRB) model, we formulate an optimization problem to allocate thruster forces to guarantee that contact forces cannot lead to the leg slipping. Finally, the effectiveness and practicality of the proposed methods are validated via extensive Gazebo simulations as well as hardware experiments. Note to Practitioners—The motivation for this paper stems from the need to develop a practical rough terrain climbing algorithm for a thruster-assisted underwater hexapod robot that can walk the underwater structure with any dip angles to perform some meticulous small-range operations such as hull cleaning, fracture detection, and damage restoration. However, existing climbing algorithms mainly focus on finding legs’ torques or footsteps for point-contact legged robots. They may fail to be directly used in the underwater robot simultaneously driven by thrusters and C-shaped rolling-contact legs. Then, we propose an optimal gait planning method to find C-shaped legs’ smooth desired rotation angles and an optimal thruster force allocation method to regulate each support leg’s contact force to avoid slipping. Finally, the gait planning method can also be applied to other rolling-contact legged robots, and the thruster force allocation method can help other thruster-assisted legged robots perform complex locomotion tasks.}, archive = {J_TASE}, author = {Lepeng Chen and Rongxin Cui and Weisheng Yan and Yang Li and Kaiyang Xu}, doi = {10.1109/TASE.2024.3443399}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6340--6353}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal gait planning and thruster force allocation for rough terrain climbing of an underwater hexapod robot}, volume = {22}, year = {2025}, }
• (2025). Pose estimation of instruments for automatic chemical laboratories using multi-level template matching. TASE, 22, 6327--6339. (www)

  @article{ , abstract = {In chemical laboratories, robot operation of instruments often relies on structured auxiliary positioning and teaching methods, which are complicated and laborious. Moreover, the single color, sparse textures, and uneven scales of instruments make the accuracy and robustness of existing visual 6-Dof pose estimation methods difficult to meet the requirement of robot operation. Therefore, we propose a novel pose estimation approach for automatic chemical laboratories using multi-level template matching to assist robots in operating instruments. This approach matches the query image with templates step by step from three levels: template, image, and pixel. During the matching processes from global to local, it achieves prediction of 2D key-point coordinates to accurately estimate instrument pose. At the same time, a template searching method based on the distribution pattern of feature points is proposed to ensure the accuracy of template searching in the real world. The experimental results show that our approach has good robustness and accuracy and meets the requirement of robot operation in automatic chemical laboratories. Note to Practitioners—This paper was motivated by the positioning problem of robot operation for instruments in automatic chemical laboratories. Structured auxiliary positioning methods are complicated and most of the pose estimation works are not suitable for practical applications at present. This paper suggests a novel template searching method and a pose estimation approach using multi-level matching for instruments. We calculate the distances, angles and point number of feature points to construct a new feature searching the best matched template with the input image. Then the predefined key pixels of the template are transformed into the input image from coarse to fine matching to ensure the accuracy and robustness. Due to the instrument CAD model, we can obtain the 2D-3D correspondences and calculate pose of the instrument by the PnP method. The experiment results validate that our method is suitable for practical robot operation in automatic chemical laboratories.}, archive = {J_TASE}, author = {Xuchun Zhang and Fei Zhang and Xinsheng Tang and Qing Zhu and Luyuan Zhao and Hengyu Xiao and Shuang Cong and Jun Jiang and Weiwei Shang}, doi = {10.1109/TASE.2024.3443147}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6327--6339}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Pose estimation of instruments for automatic chemical laboratories using multi-level template matching}, volume = {22}, year = {2025}, }
• (2025). Asynchronous dynamic output feedback control for discrete nonlinear networked semi-markov jump models with cyber attacks and applications. TASE, 22, 6316--6326. (www)

  @article{ , abstract = {In this work, the asynchronous dynamic output feedback control is investigated for discrete nonlinear networked semi-Markov jump models with cyber attacks, in which asynchronous phenomenon refers to the mode mismatch between the controller and the system. For the potential uncertainty of system parameters, the interval type-2 fuzzy method is adopted to characterize nonlinear semi-Markov jump models. In light of actual state information unavailable in complex environment, the dynamic output feedback technique is proposed. The main novelty is to construct an appropriate dynamic output feedback control scheme, fully consider the asynchronous phenomenon of the controller mode, and introduce auxiliary variables to solve the matrix dimensional problem, so that the IT2 fuzzy network semi-Markov jump models under the influence of cyber attacks have better dynamic performance. According to stochastic theory, semi-Markov kernel, interval type-2 fuzzy method, and dwell-time-dependent Lyapunov function, sufficient criteria are established to ensure that the networked system is $\vartheta $ -error mean-square stable under random denial-of-service attacks. Furthermore, numerically checkable conditions are formulated to solve the asynchronous dynamic output feedback controller gain. Finally, a two-degree-freedom quarter-car suspension model is given to explain the superiorities of the proposed design approach. Note to Practitioners—As one of the research hotspots, networked control systems show outstanding advantages of low maintenance cost, easy installation, and high flexibility. However, cyber attacks often occur in networked control systems, posing a great threat to signal transmission. With the development of modern science and technology, semi-Markov jump models, owing to their excellent engineering background in modeling complex system, have a wide range of application prospects in solar receiver control, power electronics, chemical processes, and network communication. Note that some factors in the practical dynamical systems, such as parameter uncertainty and dwell information, cannot be completely obtained, and actual controller mode has switching delay with the system mode. In this paper, the asynchronous output feedback control is studied for discrete nonlinear networked semi-Markov jump models with cyber attacks. On the basis of interval type-2 fuzzy and dwell-time-dependent Lyapunov function, the $\vartheta $ -error mean-square stability is realized under random denial-of-service attacks. This research provides a new approach to study output feedback control strategies for discrete networked systems.}, archive = {J_TASE}, author = {Wenhai Qi and Runkun Li and Guangdeng Zong and Huaicheng Yan and Zheng-Guang Wu and Jun Cheng}, doi = {10.1109/TASE.2024.3441545}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6316--6326}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous dynamic output feedback control for discrete nonlinear networked semi-markov jump models with cyber attacks and applications}, volume = {22}, year = {2025}, }
• (2025). Noninvasive EEG-based intelligent mobile robots: A systematic review. TASE, 22, 6291--6315. (www)

  @article{ , abstract = {Brain-controlled mobile robotics can provide restoration of mobility for individuals with severe physical disabilities and empower healthy people with a broader reachable range in particular environments, which have been flourished over the past twenty years. This paper conducts a systematic state-of-the-art overview of noninvasive EEG-based intelligent mobile robots. We first present the general architecture and basic concepts, typical system types, and main research efforts on the whole-system design. Then, relevant key techniques associated with the brain-machine interfaces (BMIs), control strategies, and robot intelligence are reviewed to elucidate the research progress of the overall system. System performance evaluation is critical and complicated, here we summarize the conditions of the recruited participants, the experimental protocol, tasks and environments, with an emphasis on evaluation metrics regarding BMI performance, navigation performance, system robustness, and the user. We further highlight the remaining challenges and the potential research directions of future work. This study with informative outline is envisioned to enhance current understanding and suggest the future perspectives on EEG-based mobile robotic devices. Note to Practitioners—The brain-computer interfaces (BCIs) can provide a novel and feasible way to convey the users’ intentions to the terminal devices through the brain signals, which has exhibited emerging prospects in both civil and military applications thus having attracted increasing interests in researchers, industries, and government. The integration of noninvasive electroencephalogram (EEG)-based BCIs and expanded wheeled robotics have potentials to improve the human daily life with enhanced level of mobility, exhibiting significant socioeconomic impacts. This paper provides a systematic review of the EEG-based intelligent mobile robots from the aspects of key milestones, key techniques, and performance evaluation. To bring such the next generation assistive products from the laboratory to real applications, this paper further highlights the remaining challenges and shed light on the potential future work.}, archive = {J_TASE}, author = {Hongqi Li and Xiaoya Li and José R. del Millán}, doi = {10.1109/TASE.2024.3441055}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6291--6315}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Noninvasive EEG-based intelligent mobile robots: A systematic review}, volume = {22}, year = {2025}, }
• (2025). Secure bipartite consensus control for dynamic event-triggered multi-agent systems based on co-estimation of state and attacks. TASE, 22, 6280--6290. (www)

  @article{ , abstract = {The dynamic event-triggered (DET) bipartite consensus and attack estimation problem is investigated for multi-agent systems (MASs) under false data injection (FDI) attacks generated by exogenous systems. For the purpose of compensating for the system bias, a novel set of estimators are introduced to estimate the FDI attacks. To optimize the limited resources of the network bandwidth, a DET protocol dependant on the auxiliary variable and the local estimations is employed to regulate data transmission. By employing the designed state observer and attack observer, an novel attack tolerant event-triggered control strategy is established. Subsequently, sufficient criteria are obtained to ensure the bipartite consensus performance with $l_{2}$ - $l_{\infty }$ constraint. Then, by solving algebraic matrix equations and recursive linear matrix inequalities (LMI), the gains of state estimator, attack estimator and desired controller are determined, respectively. Finally, simulation examples are provided to demonstrate the validity of the derived results. Note to Practitioners—This paper discusses the secure bipartite consensus control problem for MASs subject to FDI attacks. Since the system state deviation caused by FDI attacks will bring about degradation of system performance, it is important to protect the MASs from being destroyed by designing observers to estimate the FDI attack and the system state. In view of the bandwidth constraints among the agents in MASs, optimization utilization of the network bandwidth should be taken into account seriously. Therefore, in this paper, the proposed observer-based DET controller is developed to guarantee bipartite consensus of the considered MASs while ensuring efficiency utilization of bandwidth resources. The current research provides a helpful reference for the secure bipartite consensus control of MASs with bandwidth constraints.}, archive = {J_TASE}, author = {Lijuan Zha and Danzhe Liu and Engang Tian and Xiangpeng Xie and Chen Peng and Jie Cao}, doi = {10.1109/TASE.2024.3440493}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6280--6290}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure bipartite consensus control for dynamic event-triggered multi-agent systems based on co-estimation of state and attacks}, volume = {22}, year = {2025}, }
• (2025). Adaptive tracking control for uncertain nonlinear multi-agent systems with partially sensor attack. TASE, 22, 6270--6279. (www)

  @article{ , abstract = {Although rich collection of research results on sensor attack (SA) exist, no attack detection mechanism has ever been designed based on output error information to detect whether an attack has occurred. The primary objective of this article is to build a backstepping adaptive tracking control protocol for heterogeneous nonlinear uncertain multi-agent systems (HNUMASs) with partially SA. A dynamic SA detection mechanism by using output error information of agent only is developed to identify the SA. After locating the attack, to circumvent the effects of unknown time-varying output gain caused by SAs, we introduce the Nussbaum function in backstepping design to compensate the unknown time-varying output gain. The result shows that the developed SA detection mechanism is able to detect the occurrence of attack in a timely manner, while the derived adaptive backstepping tracking controller can effectively handle the adverse effects of SA and ensure that all signals are bounded in the closed-loop system. At last, the effectiveness and benefits of the presented approach are verified by simulation example. Note to Practitioners—This paper aims to achieve the adaptive tracking control for HNUMASs under SA, which can be widely used in practice, such as power systems, vehicular platoon systems, etc. The control protocol consists of a attack detection mechanism and Nussbaum function that compensates for the unfavorable effects caused by SAs. Moreover, the system may also be influenced by uncertainties from its neighboring agents in practical applications. Therefore, an additional estimator is designed in each subsystem to handle the uncertainties involved in its neighbor dynamics. This design avoids the exchange of information related to local neighborhood consensus errors among connected subsystems. A feasible strategy is provided for industrial applications.}, archive = {J_TASE}, author = {Guangliang Liu and Qiuye Sun and Hanguang Su and Zhijian Hu}, doi = {10.1109/TASE.2024.3440394}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6270--6279}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tracking control for uncertain nonlinear multi-agent systems with partially sensor attack}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered boundary control for cyber-physical flexible riser systems subject to spoofing attacks. TASE, 22, 6259--6269. (www)

  @article{ , abstract = {This paper investigates the dynamic event-triggered boundary control of cyber-physical flexible riser systems under spoofing attacks. It aims to dampen vibrations resulting from riser pipe deformation, alleviate the communication load, and curtail the influence of such attacks. Firstly, a novel dynamic event-triggered boundary controller is constructed by utilizing external disturbances and event-triggering strategies. This controller overcomes the complex spatio-temporal coupling phenomena in riser systems and suppresses vibrations. Secondly, a controller is developed to minimize the impact of false data injection on the cyber-physical flexible riser system due to uncertain spoofing attacks. Finally, the stability of the system is mathematically analyzed to prevent the occurrence of Zeno phenomena. The simulation results further validate the effectiveness of the proposed method. Note to Practitioners—Cyber-physical flexible risers play an important role in various marine engineering and industrial fields, such as underwater oil and gas extraction, deep-sea resource exploitation, and seabed exploration. A challenge within these domains is the effective mitigation of cyber-physical riser vibrations amidst spoofing attacks, a problem that has yet to be resolved. This paper introduces a novel dynamic event-triggered boundary control strategy aimed at dampening vibrations induced by riser deformations, reducing communication overhead, and attenuating the effects of spoofing attacks. Based on PDEs, the proposed control strategy offers a solution for the regulation of flexible riser systems in forthcoming engineering endeavors.}, archive = {J_TASE}, author = {Xin-Yu Zhang and Xiang-Peng Xie and Ju H. Park}, doi = {10.1109/TASE.2024.3440310}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6259--6269}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered boundary control for cyber-physical flexible riser systems subject to spoofing attacks}, volume = {22}, year = {2025}, }
• (2025). A two-stage method for building evacuation with discrete time model. TASE, 22, 6246--6258. (www)

  @article{ , abstract = {This paper analyzes the evacuation process of people in a building, and constructs a discrete-time evacuation model that can accurately describe the evacuation problem according to evacuation scenarios. In the discrete time framework, the evacuation policy is dynamically adjusted based on the dynamic transfer of people. For the large-scale evacuation problem, this paper proposes a two-stage method to solve the evacuation policy of edge and node separately, which improves the solving efficiency. The results of case study prove the reliability and efficiency of our method. Note to Practitioners—In this paper, considering the influence of effective edge width and crowd density on the moving speed of people, a discrete time based evacuation model is constructed which accurately reflects the evacuation process. A two-stage method is proposed to solve the difficult problem of large-scale evacuation. The evacuation policy obtained by the two-stage method will be used as the evacuation plan, and the evacuation plan of different evacuation scenarios will be counted. When an evacuation event occurs, the distribution of people is matched with the scenario in the database, and the evacuation plan of the closest scenario is selected for evacuation.}, archive = {J_TASE}, author = {Yuandong Zhang and Qiaozhu Zhai and Zhanbo Xu and Jiang Wu and Xiaohong Guan}, doi = {10.1109/TASE.2024.3440265}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6246--6258}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A two-stage method for building evacuation with discrete time model}, volume = {22}, year = {2025}, }
• (2025). OMEPP: Online multi-population evolutionary path planning for mobile manipulators in dynamic environments. TASE, 22, 6234--6245. (www)

  @article{ , abstract = {This paper presents an online multi-population evolution path planning (OMEPP) algorithm to address the flexible path planning problem for mobile manipulators in complex dynamic environments. The OMEPP algorithm treats the mobile manipulator as a high-dimensional system to utilize its flexibility. The OMEPP algorithm is based on random sampling and evolutionary concepts: Optimization and passive obstacle avoidance operations are performed on the path at runtime, with superior paths replacing inferior ones within the same population. A novel path population partitioning approach is proposed to maintain diverse switchable paths, thereby improving robustness. This paper also proposes an efficient manipulator collision detection method and several other mechanisms to enhance the algorithm’s effectiveness. The experimental results demonstrate the algorithm’s ability to swiftly adapt and optimize paths in response to dynamic environmental changes. Note to Practitioners—This paper presents OMEPP, an online evolutionary algorithm for real-time path planning of mobile manipulators in dynamic environments. OMEPP employs novel techniques including path population partitioning, random sampling, and evolution to efficiently generate collision-free paths among moving obstacles. A novel path population partitioning approach is proposed to maintain diverse switchable paths, thereby improving robustness. Simulations have demonstrated that the OMEPP algorithm is effective for real-time path planning of mobile manipulators in complex dynamic environments. Future work will focus on trajectory generation respecting dynamics limits.}, archive = {J_TASE}, author = {Yangjun Pi and Xin Liu and Zuodong Yang and Yunlin Zhong and Tao Huang and Huayan Pu and Jun Luo}, doi = {10.1109/TASE.2024.3440252}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6234--6245}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {OMEPP: Online multi-population evolutionary path planning for mobile manipulators in dynamic environments}, volume = {22}, year = {2025}, }
• (2025). Lexicographic dual-objective path finding in multi-agent systems. TASE, 22, 6223--6233. (www)

  @article{ , abstract = {Path finding in multi-agent systems aims to identify collision-free and cost-optimized paths for all agents with distinct start and goal positions. It poses challenging optimization problems. Existing research typically treats all agents equally, overlooking their differences in practical scenarios where they undertake the tasks of varying importance. In many application scenarios, agents must be differentiated into critical (c-agents) and acritical ones (a-agents) due to premium/general service, no-loaded/full-loaded states, and urgent/non-urgent tasks. Facing this practical need, this work focuses on multi-agent systems in which the different importance of agents must be considered; and tackles a lexicographic dual-objective variant of path-finding problem. The consideration of c-agents makes the concerned problem more useful yet more challenging than basic multi-agent path-finding problems. Different from existing multi-agent path planning methods that minimize the sum-of-costs of all agents, we optimize two objectives with preferences to emphasize the influence of c-agents on the system. The primary one is to minimize the sum-of-costs of c-agents and the secondary one is to minimize that of a-agents. As existing methods are inadequate for this unique challenge, we adapt a conflict-based search framework and design new two-level lexicographic dual-objective optimization methods to deal with it. A high level is responsible for iteratively expanding a search tree and adding constraints to resolve conflicts among agents. A low level is responsible for finding the path of each agent for the node newly expanded in the high level. By conducting numerous computational experiments, we verify the great performance of the presented methods in solving the concerned problem. We further develop a prototype system incorporating our methods and make it public to promote their practical application. This research contributes valuable insights and solutions to pathfinding challenges in multi-agent systems with critical and acritical agents. Note to Practitioners—This work addresses a multi-agent path finding problem in multi-agent systems involving both critical and acritical agents. The former are more important than the latter since they are assigned to perform more important tasks. This is a common scenario in manufacturing and service environments. We propose a two-level lexicographic dual-objective optimization framework to deal with the problem and underscore the importance of c-agents in the path finding problem. Three solution approaches are designed for practitioners to select based on their practical application needs. The computational experimental results and statistical analyses highlight the exceptional performance of our proposed approaches. In order to facilitate the practical application, we further provide a prototype system with our proposed approaches embedded, which is openly accessible for practitioners to realize their specific applications.}, archive = {J_TASE}, author = {Ziyan Zhao and Siyi Li and Shixin Liu and MengChu Zhou and Xingyang Li and Xiaochun Yang}, doi = {10.1109/TASE.2024.3440169}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6223--6233}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Lexicographic dual-objective path finding in multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). T-TD3: A reinforcement learning framework for stable grasping of deformable objects using tactile prior. TASE, 22, 6208--6222. (www)

  @article{ , abstract = {Human tactile perception enables rapid assessment of deformable objects and the application of appropriate force to prevent slip or excessive deformation. However, this task remains challenging for robots. To address this issue, we propose the T-TD3 algorithm, which utilizes a multi-scale fusion neural network (MSF-Net) for the fused perception of multi-scale features, including the tactile prior information obtained through preprocessing. Our approach decomposes the robot task of grasping deformable objects into three subtasks: slip detection, stable grasping evaluation, and minimum grasping force tracking. We develop a simulation environment called CR5GraspStable-Env using PyBullet and TACTO for the network training. Our work reports a success rate of 94.81% in the robot task of grasping deformable objects in real, demonstrating an excellent sim-to-real capability. Moreover, the proposed approach has the potential to be extended to other stable grasping tasks that utilize tactile perception. Note to Practitioners—Traditional grasping strategies in stable grasping tasks typically apply significant grasping force to prevent slip. However, excessive grasping force for deformable objects may lead to excessive deformation and damage. While humans can flexibly control grasping force through tactile perception, it poses a significant challenge for robots. To overcome this challenge, we propose a novel method for the robot to learn an improved stable grasping strategy by incorporating tactile priors. Our method establishes a unified tactile prior representation for the visual-based tactile sensors mounted on the robot grippers, which enables them to sense the contact state. Additionally, it utilizes sensor distortion correction based on spatial symmetry to ensure the applicability of the tactile prior representation to any kind of visual-based tactile sensor. Furthermore, this method integrates multi-scale tactile priors and robot states, and utilizes reinforcement learning to autonomously make real-time decisions, aiming to minimize the grasping force while maintaining stable grasps on deformable objects. The primary research objective of this article is to address the challenge of achieving stable grasps on variable objects, while also being applicable to grasping rigid objects. We validated the practicality of this method by successfully achieving a 94.81% success rate in stably grasping deformable objects using the robotic arm CR5 in both simulated and real-world environments. Furthermore, this method can be readily applied to other robot systems. In the future, we plan to extend the application of our proposed method to more dexterous manipulators and perform more complex manipulation tasks. Additionally, we aim to introduce new fusion algorithms and decision-making strategies to further enhance the applicability of this method.}, archive = {J_TASE}, author = {Yanmin Zhou and Yiyang Jin and Ping Lu and Shuo Jiang and Zhipeng Wang and Bin He}, doi = {10.1109/TASE.2024.3440047}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6208--6222}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {T-TD3: A reinforcement learning framework for stable grasping of deformable objects using tactile prior}, volume = {22}, year = {2025}, }
• (2025). Privacy-preserving state estimation in the presence of eavesdroppers: A survey. TASE, 22, 6190--6207. (www)

  @article{ , abstract = {Networked systems are increasingly the target of cyberattacks that exploit vulnerabilities within digital communications, embedded hardware, and software. Arguably, the simplest class of attacks – and often the first type before launching destructive integrity attacks – are eavesdropping attacks, which aim to infer information by collecting system data and exploiting it for malicious purposes. A key technology of networked systems is state estimation, which leverages sensing and actuation data and first-principles models to enable trajectory planning, real-time monitoring, and control. However, state estimation can also be exploited by eavesdroppers to identify models and reconstruct states with the aim of, e.g., launching integrity (stealthy) attacks and inferring sensitive information. It is therefore crucial to protect disclosed system data to avoid an accurate state estimation by eavesdroppers. This survey presents a comprehensive review of the existing literature on privacy-preserving state estimation methods, while also identifying potential limitations and research gaps. Our primary focus revolves around three types of methods: cryptography, data perturbation, and transmission scheduling, with particular emphasis on Kalman-like filters. Within these categories, we delve into the concepts of homomorphic encryption and differential privacy, which have been extensively investigated in recent years in the context of privacy-preserving state estimation. Finally, we shed light on several technical and fundamental challenges surrounding current methods and propose potential directions for future research. Note to Practitioners—With the increasing openness and anonymization of the networked estimation systems, privacy concerns require to be paid more attention. The essence of the privacy-preserving approaches is to seek certain tradeoffs among privacy budget and various performance metrics, such as utility and energy. Cryptographic methods are suitable for high-performance processors because they need sufficient computation resources to generate and operate complicated secret keys. By contrast, perturbation methods can be realized faster, but the adverse impact on the legitimate systems should be limited not to violently disrupt the desired operations. In conclusion, the choice of these encryption approaches depends on practical demands. Moreover, general state-space models, which can represent most real-world dynamics, are the basis of the reviewed methods. Thus these approaches can be easily deployed to practical engineering systems to effectively guarantee their privacy, providing significant application values.}, archive = {J_TASE}, author = {Xinhao Yan and Guanzhong Zhou and Daniel E. Quevedo and Carlos Murguia and Bo Chen and Hailong Huang}, doi = {10.1109/TASE.2024.3440042}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6190--6207}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Privacy-preserving state estimation in the presence of eavesdroppers: A survey}, volume = {22}, year = {2025}, }
• (2025). PDE-based neuro adaptive control for multi-agent deployment with non-collocated observer. TASE, 22, 6179--6189. (www)

  @article{ , abstract = {A neuro adaptive control for deploying a partial differential equation-based multi-agent system in 3D space with a non-collocated observer is proposed in this study. Since the full states of the system are unavailable in practice, an observer-based control is developed to ensure stability of the underlying closed-loop system. In addition, the system uncertainty is addressed by introducing a neural network control. By choosing appropriate system parameters, the desired control objectives can be achieved. The proposed strategy is simple to implement and its implementation condition is easily satisfied. Finally, the effectiveness of the designed method is verified by the simulation results. Note to Practitioners— In this paper, we introduce a neuro-adaptive control strategy for a multi-agent system based on partial differential equations in 3D space, using a non-collocated observer. This approach is particularly relevant for practitioners dealing with dynamic and uncertain environments in control systems. Neural networks are employed to manage system uncertainties, adapting to changing conditions, which is crucial in environments with variable system parameters. The non-collocated observer allows for state estimation, beneficial in situations where direct measurement is impractical. Ensuring the observer’s accuracy is key for effective control. Our strategy focuses on simplicity and ease of implementation, making it accessible for integration into existing systems. The observer-based control ensures the stability of the closed-loop system, a critical factor for consistent performance.}, archive = {J_TASE}, author = {Zhijie Liu and Jianhui Zhang and Huiyang Song and Yan Yang and Zhijia Zhao and Keum-Shik Hong}, doi = {10.1109/TASE.2024.3439747}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6179--6189}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PDE-based neuro adaptive control for multi-agent deployment with non-collocated observer}, volume = {22}, year = {2025}, }
• (2025). A novel H-shaped soft magnetic microrobot for automatic manipulation in dynamic environments. TASE, 22, 6168--6178. (www)

  @article{ , abstract = {Untethered magnetically actuated soft microrobots are promising in micromanipulation applications. Autonomous navigation and micromanipulation in dynamic environments are challenging because uncertain dynamic obstacles lead to increased planning time and reduced real-time performance. Here, we developed a novel H-shape soft magnetic microrobot with finger behaviors, including gripping and releasing. Without additional stimuli (e.g., light, temperature, etc.), the finger behaviors and movements of the microrobot can be controlled simultaneously by adjusting the strength and direction of the magnetic fields. We proposed a sliding-window-based local dynamic path planning method in dynamic environments to address the drawbacks of inefficiency and the high computational cost of global search. An enhanced optimized rapidly-exploring random tree algorithm is developed for planning avoidance paths at a safe distance from the boundary of obstacles. In order to automatically manipulate multiple objects, the manipulation process is quantitatively divided into several subtasks, and a finite state machine (FSM) based task planner is proposed to control the switching between subtasks. Experiments demonstrate the autonomous obstacle avoidance and navigation of the microrobot in dynamic environments with obstacles moving 2.5 times faster than the robot. Eight spherical objects of 2 mm diameter were automatically transported to the corresponding target position with an average error of 0.29 mm and an average transport time of 17.25 seconds. In the future, this work promises to enable automated cell sorting by microrobots. Note to Practitioners—This article was motivated by the recent interest in utilizing the small-scale microrobots to perform micromanipulation tasks. For automated manipulation of objects, effective Untethered actuation, manipulation behavior control, and navigation are required. Besides, obstacle avoidance and dynamic obstacles in the environment need to be considered. The strategy proposed here is developed for automated multi-object manipulation via a novel soft magnetic microrobot. The developed task planner quantitatively divides the operation process into several subtasks and controls the switching between them. The dynamic path planning algorithm is used to efficiently search for collision avoidance paths in dynamic obstacle environments. The motion-behavior coordination controller is used to control the robot to perform manipulation subtasks. Experimental results have validated the proposed strategy in actuation, navigation, and multiple cargoes manipulation in dynamic obstacle environments.}, archive = {J_TASE}, author = {Chenyang Huang and Tiantian Xu and Hengyuan Yu and Xinyu Wu}, doi = {10.1109/TASE.2024.3439710}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6168--6178}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel H-shaped soft magnetic microrobot for automatic manipulation in dynamic environments}, volume = {22}, year = {2025}, }
• (2025). Event-based h∞ fuzzy control for implicit hybrid AVS systems embedding acceleration characteristic under random deception and DoS attacks. TASE, 22, 6154--6167. (www)

  @article{ , abstract = {This article considers the problem of event-based $H_{\infty }$ fuzzy control for implicit hybrid active vehicle suspension systems (IHAVSSs) embedding acceleration and Markovian jumping characteristics under random deception and denial-of-service (DoS) attacks. Aiming at conserving of communication resources, an event-triggered strategy is introduced in the communication link between sensor and controller. The accelerations of sprung, unsprung and tyre are taken into account, the active vehicle suspension systems are constructed as the implicit one. Admissibility conditions are derived under the framework of linear matrix inequalities by utilizing the piecewise Lyapunov-Krasovskii (L-K) functional. Via singular value decomposition (SVD) technique, the IHAVSSs are proved to be regular and impulse-free. By means of Takagi-Sugeno (T-S) fuzzy method and parallel-distributed compensation (PDC) technique, a mode-dependent (M-D) fuzzy feedback controller (FFC) is designed to ensure $H_{\infty }$ stochastic admissibilization of IHAVSSs under random deception and DoS attacks. The validity of the results is demonstrated by using the quarter active vehicle suspension model produced by Quanser company. Note to Practitioners—The active vehicle suspension systems hold paramount significance in practical applications, effectively dampening vibrations from road surfaces and enhancing ride comfort. Concurrently, as networking advances, interconnectedness among various components within the active vehicle suspension system grows. However, the communication links are vulnerable to malicious attacks that may decrease the performance of the AVSSs, which inspires our current research. This paper delves into the $H_{\infty }$ stochastic admissibilization of the AVSSs subjected to network malicious attacks, and makes full use of the dynamic variable of the system in the modeling process, that is, the accelerations of each components are selected as the system state variable. The method is applied to a platform simulation, and the stochastic admissibility of implicit hybrid active vehicle suspension systems under random deception and DoS attacks is realized.}, archive = {J_TASE}, author = {Guangming Zhuang and Qingyu Zhu and Xiang-Peng Xie and Jianwei Xia}, doi = {10.1109/TASE.2024.3439705}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6154--6167}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based h∞ fuzzy control for implicit hybrid AVS systems embedding acceleration characteristic under random deception and DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Stabilization of 2D markov jump systems with directional communication delays: Handling delayed modes and asynchronous modes. TASE, 22, 6143--6153. (www)

  @article{ , abstract = {This paper studies the stabilization problem of two-dimensional (2D) Markov jump systems (MJSs) with directional communication delays, where delays exist in both states and modes. Based on whether the delay mode can be directly observed, the mode-delayed and asynchronous controllers are designed, respectively. For the mode-delayed case, the closed-loop system with current modes and delayed modes is re-planned as a closed-loop 2D MJS. For the asynchronous case, an extended hidden Markov model is developed to describe the asynchronous modes in controllers. Based on the Lyapunov theory, sufficient conditions are derived to ensure the asymptotic mean square stability of the closed-loop 2D MJSs under these two cases. Finally, two different examples from a representative model of some thermal processes are verified in simulations to demonstrate the effectiveness of the designed approaches. Note to Practitioners—2D systems have found extensive applications in thermal processes, gas absorption, and water stream heating, etc. In these applications, sudden changes in parameters and structures are difficult to avoid, which will result in the system being unable to be described. Fortunately, this problem can be handled by the Markov model, which consists of modes and states. In the control problem of 2D MJSs, delayed states are usually considered in the plant. Consider a more practical case that delays exist in directional communication channels between the plant and the controller, resulting in directional delayed modes and states in the controller. In this case, how to handle these complex modes and state information and stabilize the system is of practical significance. Based on whether the delay mode can be directly observed, the mode-delayed and asynchronous controllers are designed, respectively. Finally, two different examples from a representative model of some thermal processes are verified in simulations to demonstrate the effectiveness of the designed approaches.}, archive = {J_TASE}, author = {Xiang Zhang and Shuping He and Zehua Jia and Dongsheng Guo and Weidong Zhang}, doi = {10.1109/TASE.2024.3439694}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6143--6153}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stabilization of 2D markov jump systems with directional communication delays: Handling delayed modes and asynchronous modes}, volume = {22}, year = {2025}, }
• (2025). A covert spoofing algorithm for SINS/GNSS tightly integrated navigation system. TASE, 22, 6134--6142. (www)

  @article{ , abstract = {In order to solve the spoofing problem of SINS/GNSS tightly integrated navigation system with inertial assist spoofing detection module, this paper proposes a covert spoofing algorithm for SINS/GNSS tightly integrated navigation system. Firstly, the influence mechanism of the GNSS spoofing signal on the innovation and state estimation parameters of the SINS/GNSS tightly integrated navigation system is analyzed, and then the GNSS spoofing signal model is established with the constraint condition of the influence of the GNSS spoofing signal on the innovation and state estimation parameters of the SINS/GNSS tightly integrated navigation system. Secondly, the mathematical model of covert spoofing is constructed by using the GNSS spoofing signal model. Finally, SINS/GNSS tightly integrated navigation system is spoofed by GNSS spoofing signal generated according to the mathematical model of covert spoofing. Note to Practitioners—The motivation of this paper is to solve the spoofing problem of SINS/GNSS tightly integrated navigation system, so as to realize the control of the illegal unmanned platform equipped with tightly integrated navigation system. The existing spoofing methods are easily detected by the spoofing detection module, and it is difficult to meet the spoofing requirements of SINS/GNSS tightly integrated navigation system equipped with inertial aided spoofing detection module. This paper proposes a covert spoofing algorithm. Firstly, the influence mechanism of GNSS spoofing signal on SINS/GNSS tightly integrated navigation system is analyzed. Then, the mathematical model of covert spoofing is established by taking the influence of GNSS spoofing signal on the innovation parameter and state estimation parameters of SINS/GNSS tightly integrated navigation system as the constraint condition. Finally, the GNSS spoofing signal generated by the covert spoofing mathematical model is used to spoof the SINS/GNSS tightly integrated navigation system. Simulation and experimental results verify the effectiveness and correctness of the proposed method. In the future research, we will further promote the experiment and application of spoofing technology in the control of various illegal unmanned platforms.}, archive = {J_TASE}, author = {Xingshou Geng and Yan Guo and Kanghua Tang and Wenqi Wu and Yanchao Ren and Guodong Duan}, doi = {10.1109/TASE.2024.3439544}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6134--6142}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A covert spoofing algorithm for SINS/GNSS tightly integrated navigation system}, volume = {22}, year = {2025}, }
• (2025). Data-based dynamic decoupling control for MIMO precision motion stages with position-dependent disturbances. TASE, 22, 6122--6133. (www)

  @article{ , abstract = {Decoupling control is a widely employed technique used to mitigate coupling effects and bridge the gap between multiple-input multiple-output (MIMO) and single-input single-output (SISO) control. In the field of precision motion control, the cross-talk resulting from coarse decoupling poses a significant challenge to achieving high performance. Static decoupling control fails to completely decouple the MIMO system due to dynamic disparities between drives, actuators, and the flexible modes of the plant. Consequently, dynamic decoupling control methods have gained attention for their potential to enhance performance. However, existing dynamic decoupling control methods suffer from limitations such as reliance on system models, and susceptibility to disturbances and noise. In this paper, these deficiencies are addressed by 1) a data-based optimization with no involvement of model knowledge, and 2) using the augmented vector and instrumental variable to eliminate the estimation bias caused by position-dependent disturbances and measurement noise. The effectiveness and superiority of the proposed method are substantiated through numerical simulations and experiments conducted on an ultra-precision wafer stage. Note to Practitioners—This paper presents a non-iterative dynamic decoupling method, which is suitable for LTI multivariable systems with stringent precision requirements, such as the wafer stage and the atomic force microscope. In the field of precision motion control, the feedback controller plays a crucial role in the stability and robustness of the system. However, to achieve the desired performance, additional methods may be necessary, such as feedforward control method, nonlinear control method, etc. Especially for multiple-input multiple-output systems, it is essential to maintain high decoupling accuracy to avoid the cross-talk, which emerges due to a coarse decoupling. Given the practical differences in dynamic characteristics between drives and actuators, as well as non-rigid modes in the system, dynamic decoupling is recommended over static decoupling. Furthermore, it is promising to develop a data-based method that enables inevitable uncertainties between the model and the actual plant. In the dynamic decoupling method proposed in this paper, the impact of position-dependent disturbances and the measurement noise are jointly taken into consideration, which has not been previously addressed in literature. The augmented vector and instrumental variable are introduced to eliminate the adverse effects of these irrelevant data. Consequently, an unbiased estimate of the dynamic decoupling controller can be obtained. The practicality and effectiveness of the proposed method have been demonstrated. In future work, we will focus on the optimal selection of the dynamic decoupling controller under bounded disturbances and its application to redundant systems.}, archive = {J_TASE}, author = {Kaixin Liu and Fazhi Song and Yue Dong and Yang Liu and Jiubin Tan}, doi = {10.1109/TASE.2024.3439102}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6122--6133}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-based dynamic decoupling control for MIMO precision motion stages with position-dependent disturbances}, volume = {22}, year = {2025}, }
• (2025). Periodic event-triggered and self-triggered control of spacecraft rendezvous system with input delay. TASE, 22, 6113--6121. (www)

  @article{ , abstract = {This paper solves the problem of spacecraft rendezvous with input delay by designing the periodic event-triggered control (PETC) and periodic self-triggered control (PSTC), respectively. Firstly, a PETC based on the discrete-time parametric Lyapunov equation (DPLE) is designed to stabilize the delayed spacecraft rendezvous systems. Moreover, in order to avoid monitoring the measurement errors, a PSTC algorithm that the updates of the next control law depend on the previous triggered states is also designed. Specially, by using the properties of the DPLE, this new approach is not only simple, but also provides an easy and explicit condition on the only parameter of DPLE to guarantee the non-triviality of the designed PETC and PSTC. Finally, the effectiveness of theoretical results is verified by simulations. Note to Practitioners—Although this paper was inspired by the problem of spacecraft rendezvous, the designed algorithms can also be applied to other time-delay systems. Existing general event-triggered control methods have been used to solve the problem of spacecraft rendezvous for reducing the communication load, although this often complicates the design of the controller due to preventing the occurrence of Zeno phenomenon. To overcome this drawback, This paper proposes a periodic event-triggered control to achieve the spacecraft rendezvous with input delay, which naturally avoids the occurrence of Zeno phenomenon. In addition, a periodic self-triggered control is also designed, which has never been designed to solve the problem of delayed spacecraft rendezvous.}, archive = {J_TASE}, author = {Kai Zhang and Meilin Li and Zhijian Hu and Xuefei Yang and Kang-Kang Zhang}, doi = {10.1109/TASE.2024.3439009}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6113--6121}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Periodic event-triggered and self-triggered control of spacecraft rendezvous system with input delay}, volume = {22}, year = {2025}, }
• (2025). Decoupled finite-time approximation auxiliary system-based three-dimensional integrated guidance and control. TASE, 22, 6100--6112. (www)

  @article{ , abstract = {In this paper, to develop a backstepping-based finite-time control (BFTC) method for the three-dimensional integrated guidance and control (IGC) system with target maneuvers and disturbances, a novel decoupled finite-time approximation auxiliary system (DFTAAS) is proposed together with a finite-time disturbance observer (FTDO). To accurately estimate the target maneuvers and the disturbances, the FTDO is designed. To approximate the virtual control law derivative and avoid the explosion of complexity problem, the DFTAAS is proposed. To solve the singularity issue and guarantee the boundedness of the virtual control law derivative, the piecewise functions are designed to derive the DFTAAS and the virtual control. Besides, to make the controller parameters independent of the norms associated to the control matrices, the control matrices are introduced into the DFTAAS so that the conservatism of the controller parameters is restrained. Furthermore, the finite-time stability properties of the system state stabilization and the disturbance estimation are integrated into the Lyapunov stability to guarantee the finite-time stability of the whole closed-loop IGC system. Finally, the hardware-in-the-loop experiment is implemented to verify the effectiveness of the proposed method. Note to Practitioners—The aim of this paper is to design a BFTC scheme for the three-dimensional IGC system with target maneuvers and disturbances. In the practical IGC application, to solve the explosion of complexity problem, the finite-time dynamic surface control is designed to implement the BFTC. Nevertheless, the boundedness of the virtual control law derivative, which is essential for the system stability, can not be guaranteed due to the singularity. Besides, the parameter conservatism is inevitable due to the control matrices. In addition, the finite-time stability properties of the system state stabilization and the disturbance estimation, which interact and influence each other in the control process, is rarely integrated in practical. In view of the above actual issues, the piecewise function is designed to develop the DFTAAS and the guidance law to solve the explosion of complexity problem and ensure the boundedness of the virtual control law derivative. The control matrices are introduced into the DFTAAS to solve the problem of parameter conservatism. The finite-time stability properties of the system state stabilization and the disturbance estimation are integrated into the Lyapunov stability, which is more in line with the practical needs to sufficiently consider the coupling of them.}, archive = {J_TASE}, author = {Yaohua Shen and Mou Chen and Mihai Lungu and Hongzhen Guo}, doi = {10.1109/TASE.2024.3438552}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6100--6112}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Decoupled finite-time approximation auxiliary system-based three-dimensional integrated guidance and control}, volume = {22}, year = {2025}, }
• (2025). Vision-based automatic control of a surgical robot for posterior segment ophthalmic surgery. TASE, 22, 6088--6099. (www)

  @article{ , abstract = {In ophthalmic surgery, especially in posterior segment procedures, clinicians face significant challenges, like the inherent tremor of the surgeon’s arm, restricted visibility, and heavy reliance on the surgeon’s skills for precise control of hand-held tools during micro-surgical movements. Automatic control of robotic-assisted ophthalmic surgical systems has the potential to overcome these challenges, simplifying complex surgical procedures. This paper proposes a novel image-guided automatic control method for an Ophthalmic micro-Surgical Robot (OmSR), specifically designed for posterior segment eye surgery. The method relies on forceps shadow tracking. The paper introduces a tip detection network (Net-SR), which accurately calculates the coordinates of the Tips of Surgical Forceps (ToSF) and Tips of Shadow (ToS) to enable automatic navigation. Additionally, through the Non-Uniform Rational B-Spline (NURBS) curve interpolation and speed look-ahead algorithm, dense and time-continuous data points are obtained to improve control accuracy and smoothness. The accuracy of the Net-SR network and motion of the ToSF, and the effectiveness of the proposed automatic controller are experimentally evaluated. Results demonstrate a significant 98.21% improvement in the Net-SR network accuracy over the normal keypoint detection network. The use of the speed look-ahead algorithm leads to a notable 41.7% improvement in optimal speed, and the ToSF successfully reaches the target lesion with vision-based navigation and no overscale motion. Note to Practitioners—The practical problem that motivated this research is the need for safer and more efficient surgical procedures, focusing on minimizing the risk of fundus tissue damage associated with intraoperative surgical instruments. To overcome challenges related to handheld and tele-operated control, we explore automatic control as a promising solution. In this paper, the tip of the instrument can consistently and accurately reach the target lesion with high precision and no overscale motion, allowing for deskilling of complex and repetitive tasks. This capability holds potential for the clinical needle insertion operation and membrane peeling operation. The proposed control methods can also be extended to other surgical procedures.}, archive = {J_TASE}, author = {Ning Wang and Xiaodong Zhang and Sophia Bano and Danail Stoyanov and Hongbing Zhang and Agostino Stilli}, doi = {10.1109/TASE.2024.3438452}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6088--6099}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Vision-based automatic control of a surgical robot for posterior segment ophthalmic surgery}, volume = {22}, year = {2025}, }
• (2025). Spatial attention-based capsule networks with guaranteed group equivariance. TASE, 22, 6076--6087. (www)

  @article{ , abstract = {Some capsule networks (CapsNets) reported lately aim to enforce capsule poses and descriptors to be equivariant and invariant respectively by adding extra loss functions as regularization but without providing rigorous proof. To address this problem, a group equivariant spatial attention mechanism (GSA) is proposed to rigidly guarantee the equivariance with mathematical proof while enhancing the spatial information in capsule poses. In addition, to alleviate the computation burden associated with the conventional routing algorithm, group poolings are developed to generate the descriptors and poses of capsules, which contribute greatly to preserving the invariance and equivariance of CapsNets. With the proposed components of GSA and group poolings, a new attentive CapsNet, namely spatial attentive group equivariant CapsNets (SAGE-CapsNets), is constructed in this paper. To validate the invariance and equivariance of SAGE-CapsNets, we conduct experiments involving classification, semantic segmentation, and visualization. The results obtained from these experiments provide empirical evidence of the effectiveness of our proposed approach. Note to Practitioners—This paper is motivated by the problem that existing affine transformations in the real world generally degrade the performance of neural networks in vision tasks like image classification, segmentation, and detection. While conventional capsule networks help to alleviate this problem by learning invariant spatial relationships between features, their robustness to affine transformations is shown through empirical results without rigorous proof. To tackle this issue, we propose a novel capsule network with equivariant components, including group spatial attention and group pooling layers. These components are rigorously proven to be equivariant and greatly contribute to the model’s robustness against affine transformations. Moreover, for practical applications, our proposed attention mechanism improves model performance without significantly increasing computation. Additionally, group pooling preserves model equivariance while reducing computation overhead. As a result, our computation-saving model can be applied to real-world vision applications that require robustness to affine transformations, such as bearing fault diagnosis and facial recognition.}, archive = {J_TASE}, author = {Ru Zeng and Yan Song and Yuzhang Qin}, doi = {10.1109/TASE.2024.3438190}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6076--6087}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Spatial attention-based capsule networks with guaranteed group equivariance}, volume = {22}, year = {2025}, }
• (2025). Modeling, robust control design, and experimental verification for quadrotor carrying cable-suspended payload. TASE, 22, 6061--6075. (www)

  @article{ , abstract = {This paper originates from two well-accepted challenges in the control of a quadrotor with a cable-suspended payload: 1) designing a refined controller based on high-precision payload swing modeling to achieve the quantized prescribed robustness; and 2) resolving the trajectory tracking performance degradation issue caused by input saturation. To combat these challenges, we start with establishing a precise payload swing model. The experimental investigation reveals the fact that neglect of realistic factors like cable-joint dry friction and cable elasticity has significant impacts on the precision of the existing models, especially under small swing angles. Therefore, we propose a new integrated drag model including a novel sign of the payload airspeed-dependent term, which lumps all payload swing damping factors together. This model is experimentally verified to be precise to provide the payload swing disturbance spectrum for quantitatively designing the bandwidth of the uncertainty and disturbance estimator (UDE). Furthermore, we resolve the input saturation issue by augmenting the classic UDE-based controller with a tracking differentiator (TD). Rigorous performance analysis derives a clear relationship between the control performance and the UDE parameter, which forms a simple yet effective parameter tuning guideline for practical applications to ensure the prescribed robustness and trajectory tracking accuracy. The effectiveness and advantage of the proposed controller are verified via comparative experiments in different flight scenarios. Note to Practitioners—The control input saturation issue and multi-parameter optimization are frequently encountered in engineering practices. When applying the TD to address the input saturation, the selection of parameter r in the TD depends on the reference continuity. Specifically, if the reference is continuous, then r can be large; otherwise r should be small to smooth the reference to avoid input saturation. For the multi-parameter optimization, the feedback gains $k_{p}$ and $k_{d}$ should be tuned first to guarantee the system stability, followed by decreasing the UDE parameter T to improve the system robustness. However, the feasible range of T may be restricted by the measurement noise and actuator bandwidth in practice. The proposed algorithm is applicable not only in aerial transportation systems but also in addressing other disturbance rejection problems.}, archive = {J_TASE}, author = {Yang Zhu and Zhiyuan Zheng and Jinliang Shao and Hailong Huang and Wei Xing Zheng}, doi = {10.1109/TASE.2024.3437747}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6061--6075}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling, robust control design, and experimental verification for quadrotor carrying cable-suspended payload}, volume = {22}, year = {2025}, }
• (2025). An artificial-delay-based looped functional for dynamic event-triggered fault-tolerant control of T-S fuzzy multi-agent systems. TASE, 22, 6050--6060. (www)

  @article{ , abstract = {This paper investigates the event-triggered (ET) fault-tolerant control problem of T-S fuzzy multi-agent systems (MASs) subject to actuator faults. A novel dynamic event-triggering mechanism is proposed to optimize communication transmission efficiency between agents. This mechanism includes some existing event-triggering mechanisms as exceptional cases and allows flexible adjustment of threshold parameters based on the system state variations. Under this triggering mechanism, the fault-tolerant control problem of MASs is transformed into a stability problem of time-delay systems with the help of artificial time-delay methods. Based on this, an artificial-delay-based looped functional is proposed to improve the existing Lyapunov functional. Then, a sufficient condition for fault-tolerance consensus of MASs, which can tolerate larger ET threshold parameters, is derived. Finally, the effectiveness of the proposed ET fault-tolerant control strategy is validated through application to the Duffing-Van der Pol oscillators. Note to Practitioners—The motivation of this paper is to optimize communication resources and deal with potential actuator failures in cooperative control of MASs, which are widely used in autonomous vehicles, industrial automation, and the Internet of Things. It is noted that frequent communication between MASs will cause communication congestion and reduce control efficiency, and actuator failure will inevitably occur due to physical damage or aging. Thus, the study of ET fault-tolerant control not only helps to optimize communication resources, but also ensures the reliable operation of MASs in the face of actuator abnormal situations, guaranteeing stability and reliability in practical applications. This paper proposes a novel dynamic ET fault-tolerant control protocol to achieve cooperative control objectives in the presence of actuator faults. Additionally, an artificial-delay-based looped functional is constructed to enhance the design flexibility of the ET mechanism and controller. The ET fault-tolerant control strategy is applied to the Duffing-Van der Pol oscillator model to verify its effectiveness and practicability.}, archive = {J_TASE}, author = {Huichao Lin and Jiuxiang Dong and Ju H. Park}, doi = {10.1109/TASE.2024.3436927}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6050--6060}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An artificial-delay-based looped functional for dynamic event-triggered fault-tolerant control of T-S fuzzy multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Self-triggered MPC for teleoperation of networked mobile robotic system via high-order estimation. TASE, 22, 6037--6049. (www)

  @article{ , abstract = {Since the teleoperation process of networked mobile robotic systems (NMRSs) is generally affected by the coupling of the human force, complex dynamical model, external environment and nonholonomic constraints, designing an effectiveness control to guarantee the desired performance for the teleoperation task is a challenging work. Besides, due to the limitation of the communication and control technology, the controller for the teleoperation system is usually required to have low computation load and low monitoring sampling frequency. To address these challenges, a novel self-triggered model predictive control (STMPC) framework, being consisted of the local STMPC, high-order estimation and M/R fixed-time controller, is constructed. Compared to traditional MPC methods, our proposed local STMPC offers lower computational load and requires fewer monitoring resources while retaining the ability to optimize control performance and handle multiple constraints. Using the presented STMPC framework in a hierarchical manner and newly designed high-order estimation, we successfully and simultaneously solve the teleoperation task while accounting for the self-triggering mechanism, decentralized control implementation, disturbance rejection, and complex nonholonomic model. Additionally, we derive sufficient conditions for ensuring the stability of the closed-loop system. Finally, the simulation and experiment results are provided to demonstrate the effectiveness of the proposed STMPC framework. Note to Practitioners—Our research introduces a groundbreaking framework, self-triggered model predictive control (STMPC), specifically designed to elevate control efficiency, quality, and reliability in the context of remote operation of networked mobile robotic systems (NMRSs). This innovation holds immense promise, particularly in domains requiring the utilization of autonomous vehicle fleets, such as large-scale exploration, search and rescue missions, and escort operations in hazardous or hard-to-reach areas (e.g., forest firefighting, warzone patrolling, saturation attacks, and space escort). These environments are often characterized by extreme external conditions and unpredictable external influences, such as meteorite impacts, flames, obstacles, and explosions, posing significant challenges for the control and task execution of unmanned vehicle fleets. Moreover, these missions demand high precision and control responsiveness from the robotic clusters to swiftly accomplish urgent tasks. Consequently, the need arises for autonomous vehicle fleets that not only possess the capacity to adapt to various constraints and external disturbances but also execute mission tasks swiftly and precisely. STMPC, as presented in this paper, demonstrates its capability to address these issues and requirements effectively, making it a versatile and powerful tool for enhancing control and system performance in various scenarios.}, archive = {J_TASE}, author = {Jing-Zhe Xu and Zhi-Wei Liu and Ming-Feng Ge and Yan-Wu Wang and Ding-Xin He}, doi = {10.1109/TASE.2024.3436922}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6037--6049}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-triggered MPC for teleoperation of networked mobile robotic system via high-order estimation}, volume = {22}, year = {2025}, }
• (2025). Optimization of embedded system with edge computing and sleep modes for balance between energy consumption and cache occupancy. TASE, 22, 6024--6036. (www)

  @article{ , abstract = {The paper considers the embedded system that can either compute tasks locally by itself or offload tasks to the edge server for remote computing during the running period (RP) and switch to the sleep mode to save energy once the RP ends, i.e., the idle period (IP) arrives. The tasks are stored in the cache, and the more tasks are computed during the RP, the less cache space will be occupied at the end of RP, while also leading to more energy consumption. Meanwhile, the sleep mode that the embedded system enters during the IP also influences the energy consumption. Therefore, how to make the optimal tradeoff between energy consumption and cache occupancy arises as an interesting issue. To address this issue, this paper first establishes an optimization-theoretical framework to formulate the energy consumption under the constraint of cache occupancy, then discovers the most energy-saving RP, computing mode (i.e., local or edge computing), and low-power mode. An algorithm based on our discovered theoretical results is proposed for the embedded system to minimize the energy consumption within the acceptable level of cache occupancy. Theoretical analysis and field experiments jointly verify its good performance. Note to Practitioners—This paper addresses the interesting tradeoff between energy consumption and cache occupancy in the embedded system that operates in the environments with limited available energy as well as cache space. It facilitates to improve the operation efficiency of the embedded systems in the area of Internet of Things (IoT) or Cyber-Physical Systems (CPS) that employs edge computing technology to empower embedded systems with more computing capability and sleep modes to guarantee embedded systems with more energy savings, in order to minimize the accumulative energy consumption, while maintaining the cache occupancy in terms of task data bits stored within an acceptable range. Experimental investigations show that the solution proposed here outperforms existing ones.}, archive = {J_TASE}, author = {Chen Hou}, doi = {10.1109/TASE.2024.3435880}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6024--6036}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimization of embedded system with edge computing and sleep modes for balance between energy consumption and cache occupancy}, volume = {22}, year = {2025}, }
• (2025). Social-learning coordination of collaborative multi-robot systems achieves resilient production in a smart factory. TASE, 22, 6009--6023. (www)

  @article{ , abstract = {This paper presents a novel resilient production problem in a multi-robot system (MRS) driven smart factory that suffers diverse degradation of accuracy among heterogeneous robots. The inevitable degradation in production robots’ accuracy undermines desirable productivity and efficiency. Although traditional maintenance and calibration strategies can be employed, they fail to support the continuity, flexibility and agility required in smart factories. Instead of pursuing accurate task execution relative to a consistent global reference, we formulate an innovative MRS coordination strategy, where a collaborative MRS autonomously pursues relative accuracy against accuracy degradation toward resilient operation. This new coordination strategy introduces challenges including adapting to time-dynamic production flows, difficulties in observing production robots’ accuracy, and the absence of a comprehensive architecture of resilient MRS in a smart factory. We propose a computational approach with social learning and AI to overcome these challenges. A cyber-physical MRS model is proposed, in which the physical domain represents the time-dynamic production flows, while the cyber domain forms a partially connected wireless network to automate the collection of data regarding production flows, accuracy degradation, and peer robot measurements. Social learning based on such data is employed to collaboratively estimate accuracy and AI decision-making, thereby enabling adaptive coordinated task execution against accuracy degradation. We further investigated a group decision-based predictive maintenance against point failures caused by accuracy degradation. Computational experiments demonstrate that the proposed approach improves the effective rate and mean-time-to-fail performances against accuracy degradation and point failures in a scalable manner. Note to Practitioners—The practical problem motivating our work revolves around resilient production, which is affected by diverse accuracy degradations among heterogeneous robots. This issue is particularly relevant for implementing smart factories with multi-robot systems (MRS), including production and transportation robots. We propose a cyber-physical multi-robot system (CPMRS) to comprehend the dynamics of MRS-driven smart factories, aiming to facilitate productivity, efficiency, and resilience through MRS coordination, as opposed to accurate production task execution achieved by frequent maintenance and calibration, which sacrifices the flexibility and agility of smart factories. CPMRS allows the autonomous collection and exchange of information about production flows, accuracy degradations, and peer robot measurements with limited wireless communications. CPMRS further enables computational solutions with social learning and reinforcement learning for robots as well as group-decision-based predictive maintenance. It facilitates autonomous coordinated task execution and resilient production. Computational experiments demonstrate a 124 percent improvement in the mean time to failure (MTTF) and a 317.27 percent improvement with predictive maintenance. As in our proposal to general manufacturing, further research may be adapted to specific manufacturing scenarios. This research can benefit broader applications, such as smart logistics, warehouses, and even smart cities, where similar collective actions among multiple robots or agents are required.}, archive = {J_TASE}, author = {Zixiang Nie and Kwang-Cheng Chen and Kyeong Jin Kim}, doi = {10.1109/TASE.2024.3435443}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {6009--6023}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Social-learning coordination of collaborative multi-robot systems achieves resilient production in a smart factory}, volume = {22}, year = {2025}, }
• (2025). A distributed ADMM-based deep learning approach for thermal control in multi-zone buildings under demand response events. TASE, 22, 5994--6008. (www)

  @article{ , abstract = {The increasing electricity use and reliance on intermittent renewable energy sources challenge power grid management during peak demand, making Demand Response programs and energy conservation measures essential. This research combines distributed optimization using ADMM with deep learning models to plan indoor temperature setpoints effectively. A two-layer hierarchical structure is used, with a central building coordinator at the upper layer and local controllers at the thermal zone layer. The coordinator must limit the building’s maximum power by translating the building’s total power to local power targets for each zone. Local controllers can modify the temperature setpoints to meet the local power targets. While most algorithms are either centralized or require prior knowledge about the building’s structure, our approach is distributed and fully data-driven. The proposed algorithm, called Distributed Planning Networks, is designed to be both adaptable and scalable to many types of buildings, tackling two of the main challenges in the development of such systems. The proposed approach is tested on an 18-zone building modeled in EnergyPlus. The algorithm successfully manages Demand Response peak events.Note to Practitioners—This article addresses the problems faced by multizone buildings in maintaining thermal comfort while participating in Demand Response events. A centralized control approach is nonscalable, raises privacy concerns, and requires huge communication bandwidth. This work introduces a distributed algorithm that adapts the temperature setpoints of multi-zone buildings to enhance Demand Response participation and energy conservation. The proposed approach is fully data-driven and only requires historical data on weather, indoor temperature, and HVAC power consumption. The blueprint of the building and details about the HVAC system’s architecture are not needed. Energy savings will be spread across all the zones depending on the user’s choice of comfort temperature interval for each zone. The distributed optimization makes the approach scalable to large buildings with multiple zones. The algorithm is designed to work for buildings with various insulated zones such as apartment buildings and hotels, and may be extended to other types of buildings.}, archive = {J_TASE}, author = {Vincent Taboga and Hanane Dagdougui}, doi = {10.1109/TASE.2024.3435073}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5994--6008}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A distributed ADMM-based deep learning approach for thermal control in multi-zone buildings under demand response events}, volume = {22}, year = {2025}, }
• (2025). Development of reconfigurable electromagnetic actuation system with large workspaces: Design, optimization, and validation. TASE, 22, 5982--5993. (www)

  @article{ , abstract = {Magnetically actuated robots have recently shown great capabilities for remote applications in medical procedures. However, the efficient actuation of magnetic robots with dexterous field and gradient generation in large workspaces remains challenging. To overcome the critical challenges, we report a reconfigurable electromagnetic actuation system (REMA) for regulating magnetic fields (maximum: 17 mT) and gradients (maximum: 120 mT/m) in large workspaces. Reconfigurable coil configurations are achieved by employing three mobile electromagnetic coils mounted on three independent 6-DOF robotic arms. Furthermore, the field characteristics generated by a single coil and three coils were modeled via Finite-element method (FEM) and measurements from experiments, respectively. Since there are non-linearities between desired field generation and coil configuration, we propose a multi-objective optimization (MOO) method for generating the Pareto-optimized coil configuration to achieve field and force control in large workspaces. Finally, extensive experiments were conducted to demonstrate the capability and dexterity of our system for autonomous magnetic manipulation in large workspaces, thus showing its potential for clinical applications. Note to Practitioners—This paper aims to address the dexterous generation of magnetic fields and gradients in large workspaces, aiming to realize accurate, efficient, and automated control of different magnetic robots. This paper introduces a reconfigurable electromagnetic actuation system based on three independent robotic arms with three electromagnetic coils. Subsequently, we propose a multi-objective optimization (MOO) method to regulate the coil configuration for generating different fields and gradients. This approach facilitates the application of magnetically driven helical robots, catheters, and capsule robots in various medical scenarios. The results demonstrate that our proposed platform and optimization strategy can effectively implement magnetic manipulations across diverse application scenarios. Looking ahead, we anticipate integrating our work with medical imaging devices to furnish doctors with enhanced tools for medical applications.}, archive = {J_TASE}, author = {Mingxue Cai and Zhaoyang Qi and Yanfei Cao and Xinyu Wu and Tiantian Xu and Li Zhang}, doi = {10.1109/TASE.2024.3435035}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5982--5993}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development of reconfigurable electromagnetic actuation system with large workspaces: Design, optimization, and validation}, volume = {22}, year = {2025}, }
• (2025). Output feedback with feedforward robust control for motion systems driven by nonlinear position-dependent actuators. TASE, 22, 5966--5981. (www)

  @article{ , abstract = {This paper introduces a control approach for a motion system driven by a class of actuators with multiple nonlinearities. The proposed approach presents a combination of a feedforward controller and an output feedback controller to achieve a tracking performance of the motion system. The feedforward controller is mainly proposed to address the actuator dynamics and provide a linearization without requiring measurements from the actuator. Subsequently, the output feedback controller is designed using the measured position to achieve a tracking objective for a desired reference signal, considering the unknown nonlinearities in the system and the error due to the open-loop compensation using feedforward control. The efficacy of the proposed control approach is validated through three applications: reluctance actuator, electrostatic microactuator, and magnetic levitation system. Both simulation and experimental results demonstrate the effectiveness of the proposed control approach in achieving the desired reference signal with minimal tracking error, considering that the actuator and system nonlinearities are unknown. Note to Practitioners—In precision-driven motion applications, the control of the motion system plays a pivotal role in attaining the desired motion profile with exceptional accuracy. Recently, modern actuators have garnered attention from industries and academia as they aim to develop the next generation of motion systems for various advanced applications. For instance, reluctance actuators are designed to drive the wafer scanner in lithography machines, and electrostatic actuators are used to drive the mirror optic systems in smartphones. However, the multiple nonlinearities and position dependency inherent in such actuators, where the mover of the actuator is part of the motion system, introduce unstable behavior, limit performance, and pose challenges for controllers. This paper presents a control approach combining feedforward and output feedback control based on the extended high-gain observer (EHGO). The proposed controller offers several advantages, including enhanced performance of motion systems driven by such actuators and increased robustness by estimating unknown nonlinearities or external disturbances. This results in more accurate and reliable motion profiles, particularly in precision applications. Moreover, the proposed control approach is easy to implement since it does not require adaptation, tuning, or training algorithms and involves fewer controller and observer parameters to design.}, archive = {J_TASE}, author = {Mohammad Al Saaideh and Almuatazbellah M. Boker and Mohammad Al Janaideh}, doi = {10.1109/TASE.2024.3434674}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5966--5981}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Output feedback with feedforward robust control for motion systems driven by nonlinear position-dependent actuators}, volume = {22}, year = {2025}, }
• (2025). Visual servoing-based anti-swing control of cable-suspended aerial transportation systems with variable-length cable. TASE, 22, 5955--5965. (www)

  @article{ , abstract = {By utilizing a suspension cable to connect the payload with the quadrotor, transport tasks can be accomplished while preserving the unmanned aerial vehicle’s agility and maneuverability, particularly in environments that are impassable for ground vehicles. Equipping onboard visual sensors and utilizing image-based visual servoing techniques, the application range of aerial transportation systems is poised to be significantly expanded in scenarios like autonomous landing and goods release. Unfortunately, within the system, there exist multiple layers of dynamic couplings between image features, quadrotor rotation, translation, and payload motion. These intricacies give rise to numerous difficulties in achieving smooth anti-swing transportation. To overcome the aforementioned difficulties, this paper presents the first image-based visual servoing control scheme for the aerial transportation system with variable-length cable. Specifically, the image moments defined on the rotated virtual image plane are taken as the image features, whose dynamics is independent of the quadrotor rotational motion. Subsequently, a generalized virtual image feature signal is introduced by organically combining the cable length and payload swing angles with the image feature, which is further exploited in the anti-swing control scheme design. The equilibrium point of the overall closed-loop system is proved to be asymptotically stable through Lyapunov techniques and LaSalle’s Invariance Theorem. Hardware experiments are conducted on a self-built aerial transportation platform to verify the proposed controller’s basic and functional performance in terms of rapid anti-swing and accurate target position and cable length tracking. Note to Practitioners—This paper is motivated by the requirement to improve the autonomy level and payload swing suppression ability of the aerial transportation system through visual servoing techniques. By installing onboard monocular camera and the cable length adjustment mechanism, the application scope of the aerial transportation system can be significantly expanded. However, due to the “double” underactuated characteristic, the visual features couple with both the quadrotor motion and the payload motion, hence, it is quite challenging to realize visual servoing control for cable-suspended aerial transportation systems with simultaneous payload swing suppression and quadrotor positioning. Accounting for the foregoing problems, this paper proposes an image-based visual servoing anti-swing control scheme. With the elaborately constructed generalized virtual image feature signal, the designed controller could improve the anti-swing ability with a completed theoretical analysis. Furthermore, two groups of hardware experiments are conducted to validate the effectiveness of the suggested control method. In future studies, we intend to design more effective control scheme for payload delivery issue with consideration of the visibility of the mobile platform.}, archive = {J_TASE}, author = {Hai Yu and Zhaopeng Zhang and Tengfei Pei and Jianda Han and Yongchun Fang and Xiao Liang}, doi = {10.1109/TASE.2024.3434637}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5955--5965}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Visual servoing-based anti-swing control of cable-suspended aerial transportation systems with variable-length cable}, volume = {22}, year = {2025}, }
• (2025). Grasping with occlusion-aware ally method in complex scenes. TASE, 22, 5944--5954. (www)

  @article{ , abstract = {Robotic arm target grasping by vision support is a commonly used method in grasping tasks and is usually used for multi-target complex scenes. Where vision support is generally used to identify the targets and to get their positions, categories and sizes. Most robotic arm grasping tasks using target recognition methods as visual inspection ignore the relationship between target objects such as the occlusion problem between objects. This limits the targets to be grasped and makes the crawling task inefficient. We propose Grasping with Occlusion-Aware aLly (GOAL) method based on binocular stereo-vision. Firstly, occlusion relationships in the view are directly inferred and targets are segmented as well as localized. Subsequently, multi-target grasping pose estimation is performed to obtain effective grasping positions. Ultimately, validation is conducted on a high-resolution dataset using the EPSON robotic arm. Note to Practitioners—This research significantly advances the field by addressing occlusion challenges in robotic grasping, offering effective methods, a valuable dataset, and practical insights. The proposed Grasping with Occlusion-Aware aLly (GOAL) method was validated on a high-resolution dataset using the EPSON robotic arm, showcasing its applicability and efficiency in real-world scenarios. This work provides valuable contributions to practitioners in the field of robotic manipulation and grasping tasks.}, archive = {J_TASE}, author = {Lulu Li and Abel Cherouat and Hichem Snoussi and Tian Wang}, doi = {10.1109/TASE.2024.3434610}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5944--5954}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Grasping with occlusion-aware ally method in complex scenes}, volume = {22}, year = {2025}, }
• (2025). Q-learning methods for LQR control of completely unknown discrete-time linear systems. TASE, 22, 5933--5943. (www)

  @article{ , abstract = {This paper focuses on solving the linear quadratic regulator problem for discrete-time linear systems without knowing system matrices. The classical Q-learning methods for linear systems can be divided into Q-learning value iteration and Q-learning policy iteration. Q-learning value iteration converges at a linear convergence rate. Q-learning policy iteration has a second-order convergence rate but requires an initial stabilizing control policy. This paper aims to propose efficient model-free algorithms for solving the optimal control problem without requiring an initial stabilizing control policy. In this paper, we first present an equivalent problem for an auxiliary system with the same optimal control policy as the LQR problem. A Q-learning algorithm is proposed to solve the equivalent problem, which is proven to converge monotonically to the optimal solution. The convergence rate of the Q-learning algorithm is heavily dependent on the auxiliary system, so we introduce a model-free homotopy method based on Q-learning to solve the LQR problem. This homotopy method can achieve the optimal solution in a finite number of iterations by solving an LQR problem in each iteration. Additionally, we propose a Q-learning Lyapunov iteration algorithm to solve the equivalent problem for an auxiliary system and analyze its properties. Finally, two examples are provided to demonstrate our results. Note to Practitioners—This paper proposes several Q-learning methods to solve the linear quadratic regulator problem for discrete-time linear systems. On the one hand, it is difficult to know the exact system dynamics knowledge in actual engineering, so this paper is devoted to developing model-free algorithms. On the other hand, this paper focuses on the LQR problem because it is widely spread in practical applications. We propose several model-free algorithms to solve the LQR problem, which provides the basis for optimal control of actual applications. Similar to policy iteration, our algorithms need to solve the Lyapunov equation. The advantage of our methods is that all of our algorithms do not have strict constraints on initial conditions compared with policy iteration. The properties of every algorithm proposed in this paper are provided. In addition, we focus on the efficiency of algorithms to obtain the optimal control policy faster. Two practical examples are used to verify the effectiveness of our methods. Finally, the applicable situations of each algorithm are summarized in the conclusion.}, archive = {J_TASE}, author = {Wenwu Fan and Junlin Xiong}, doi = {10.1109/TASE.2024.3434533}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5933--5943}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Q-learning methods for LQR control of completely unknown discrete-time linear systems}, volume = {22}, year = {2025}, }
• (2025). Affine formation maneuver control for NUSVs: An anti-competing interaction solution with random packet losses. TASE, 22, 5916--5932. (www)

  @article{ , abstract = {Network’s communication mechanism and reliability are main factors that affect the maneuverability and robustness of networked unmanned surface vehicles (NUSVs). This article investigates event-driven affine formation maneuver control (AFMC) of NUSVs. Two synchronously occurring configuration maneuvers, i.e., position formation and attitude consensus, can be achieved under the event-driven AFMC in manner of discrete neighboring communication. Nevertheless, the communication channel will be simultaneously occupied by multiple packets when different nodes trigger their events at the same time. Worse still, the broadcasted packets may lose when inter-vehicle Euclidean distances are maneuvered beyond the allowable communication range, failing to achieve the pre-specified maneuvering requirement. Regarding this, by introducing a reference system for each vehicle, a novel dynamic interleaved periodic event-triggered mechanism (DIPETM) is subsequently explored to prevent NUSVs from communication competition. Based on this framework, an inner-dynamic variable determined by the conditional-prescribed event detecting period is firstly constructed, which is not only designed to optimize the triggering frequency, but also responsible for evaluating the secondary damage caused by random packet losses. Numerical simulations are conducted to illustrate the efficacy of this work. Note to Practitioners—The swarm of NUSVs often needs to operate for extended periods in complex environments, which may challenge its maneuverability and robustness. One expectation is the ability to maneuver the entire formation in a distributed manner, performing arbitrary configuration transformation. However, limited communication resources problems are not ignorable, especially when confronting unreliable network connectivity during frequent formation maneuver processes. The reasons lie in three aspects. Firstly, physical hardware prefers the data to be detected and transmitted in manner of discrete form. Secondly, the available communication channel is always limited, leading to only a few nodes being permitted to broadcast packets at the same time. Thirdly, random packet losses phenomenon may cover the significant broadcasted packets, resulting in system instability. Within this context, we design a novel DIPETM based AFMC scheme to solve the aforementioned problems while satisfying the formation maneuver expectation.}, archive = {J_TASE}, author = {Xiaotao Zhou and Bing Huang and Bin Zhou and Cheng Zhu and Hongde Qin and Jianming Miao}, doi = {10.1109/TASE.2024.3433020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5916--5932}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Affine formation maneuver control for NUSVs: An anti-competing interaction solution with random packet losses}, volume = {22}, year = {2025}, }
• (2025). Self-evolving hermite fuzzy neural fractional-order sliding mode control of MEMS gyroscope. TASE, 22, 5906--5915. (www)

  @article{ , abstract = {In attempt to ensure that the proof mass maintains the desired vibration modes, a fractional order sliding mode control (FOSMC) for MEMS gyroscopes based on a self-evolving Hermite fuzzy neural network (SEHFNN) has been proposed, where the FOSMC is crucial in the controller design to guarantee the tracking performance and a Hermite fuzzy neural network with a structural self-evolutionary mechanism is engaged in the controller implementation. The SEHFNN combines the advantages of both self-evolving fuzzy neural network (SEFNN) and Hermite neural network (HNN) to compensate for the unknown model parameters. The SEFNN is adapted to the current application scenario by a real-time structural adjustment mechanism, performed by the lightweight computation. The Hermite polynomial function used in HNN is able to take a full range of inputs without restriction and its role as a basis function can improve the generalization neural network ability. The performance effect is measured by calculating the RMSE parameter of the tracking error. Simulation experiments verified the robust performance of the proposed controller, showing it has higher control accuracy and smoother control input, indicating the proposed self-evolutionary mechanism completes the optimal structure adjustment successfully. Note to Practitioners—This paper was motivated by the problem of advanced control of MEMS gyroscopes. a fractional order sliding mode control using a self-evolving Hermite fuzzy neural network is proposed in this paper to maintain the trajectory tracking of proof mass. A Hermite fuzzy neural network with a structural self-evolutionary mechanism is introduced to be engaged in the implementation of the controller. The introduction of Hermite polynomial increases the depth of the network while improving the generalization ability of SEHFNN by decomposing the signal. Simulation studies prove the proposed control scheme has superior performance.}, archive = {J_TASE}, author = {Juntao Fei and Jiapeng Xie}, doi = {10.1109/TASE.2024.3432937}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5906--5915}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-evolving hermite fuzzy neural fractional-order sliding mode control of MEMS gyroscope}, volume = {22}, year = {2025}, }
• (2025). Precise robotic picking up of polar body for biopsy application. TASE, 22, 5895--5905. (www)

  @article{ , abstract = {Polar body biopsy has been widely applied in preimplantation genetic diagnosis for assisted reproductive technology. The key step in the polar body biopsy is picking up the polar body from the oocyte/embryo using a micropipette. Unfortunately, the almost transparent appearance of the polar body as well as its dynamic drift when the micropipette approaches it inside the cell makes it a challenging task to pick it up with less cytoplasm loss for the cell. The unnecessary cytoplasm loss in the picking up process of the polar body easily causes damage to the development competence of the cell and may lead to disturbances to the biopsy results of the polar body. This paper proposes a precise robotic picking up method of polar bodies with less cytoplasm loss for biopsy purposes. First, a defocus imaging method is proposed to locate the polar body with an almost transparent appearance. Then, the dynamic drift of the polar body with the micropipette moving inside the cell is modeled online based on force analysis to determine an appropriate trajectory for the micropipette to approach the polar body. Further, an Active Disturbance Rejection Controller (ADRC) is designed to move the micropipette along the desired trajectory to approach the polar body and then aspirate it into the micropipette. The experimental results on porcine oocytes demonstrate that our system is capable of localizing the polar body with a success rate of 95% and an average error of $1.12\pm 0.14~\mu $ m. Moving along the determined trajectory, the micropipette is capable of approaching the edge of the polar body with an average error of $1.84\pm 0.31~\mu $ m (n =20), which is only 11% of the results obtained without dynamic drift estimation of the polar body. With this advantage, our system picks up the polar body with a close 60% improvement in success rate (95% vs 60%) and only half of the average cytoplasm loss (5% vs 10%) in comparison to operation results without dynamic drift estimation. Note to Practitioners—Picking up of polar body from the oocyte/embryo using a micropipette is a vital operation in the polar body biopsy. Precisely picking up the polar body with less cytoplasm is vital to maintaining the developmental competence of the embryo/oocyte and reducing disturbances to biopsy result. This article presented a precise robotic picking up process of polar body. This process introduced defocus imaging method for polar body localization, dynamic drift estimation of polar body, and micropipette trajectory design and motion control by Active Disturbance Rejection Controller (ADRC). Experimental results have demonstrated the efficiency of the proposed robotic picking up process. Application of this process may provide an economical and practical method to carry polar body biopsy for practitioners.}, archive = {J_TASE}, author = {Ruimin Li and Minghui Li and Jinyu Qiu and Ke Li and Yuzhu Liu and Chaoyu Cui and Shaojie Fu and Biting Ma and Qiongao Zhang and Maosheng Cui and Mingzhu Sun and Xin Zhao and Qili Zhao}, doi = {10.1109/TASE.2024.3432572}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5895--5905}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Precise robotic picking up of polar body for biopsy application}, volume = {22}, year = {2025}, }
• (2025). Constrained reinforcement learning using distributional representation for trustworthy quadrotor UAV tracking control. TASE, 22, 5877--5894. (www)

  @article{ , abstract = {Simultaneously accurate and reliable tracking control for quadrotors in complex dynamic environments is challenging. The chaotic nature of aerodynamics, derived from drag forces and moment variations, makes precise identification difficult. Consequently, many existing quadrotor tracking systems treat these aerodynamic effects as simple ‘disturbances’ in conventional control approaches. We propose a novel and interpretable trajectory tracker integrating a distributional Reinforcement Learning (RL) disturbance estimator for unknown aerodynamic effects with a Stochastic Model Predictive Controller (SMPC). Specifically, the proposed estimator ‘Constrained Distributional REinforced-Disturbance-estimator’ (ConsDRED) effectively identifies uncertainties between the true and estimated values of aerodynamic effects. Control parameterization employs simplified affine disturbance feedback to ensure convexity, which is seamlessly integrated with the SMPC. We theoretically guarantee that ConsDRED achieves an optimal global convergence rate, and sublinear rates if constraints are violated with certain error decreases as neural network dimensions increase. To demonstrate practicality, we show convergent training, in simulation and real-world experiments, and empirically verify that ConsDRED is less sensitive to hyperparameter settings compared with canonical constrained RL. Our system substantially improves accumulative tracking errors by at least 70%, compared with the recent art. Importantly, the proposed ConsDRED-SMPC framework balances the trade-off between pursuing high performance and obeying conservative constraints for practical implementations. Note to Practitioners—This work is motivated by challenges in training Reinforcement Learning (RL) for autonomous navigation in unmanned aerial vehicles, but its implications extend to other high-criticality applications in, for example, healthcare and financial services. The implementation of RL algorithm policies may exhibit various deficiencies, including (i) opaque or unstable training due to the blackbox nature of deep neural networks, (ii) difficulty in reproducing RL outcomes, and (iii) heightened costs associated with robustness investigation, e.g., hyperparameter tuning and generalization. We present a novel ‘Constrained Distributional REinforced-Disturbance-estimator’ (ConsDRED) to identify uncertainties arising from aerodynamic disturbances in agile flight. The proposed algorithm demonstrates at least an optimal convergence rate for both global optimization and constraint violations. Theoretical guarantees offered by this approach can make training convergence processes more transparent and provide a degree of confidence in the expected convergence outcomes in real-world applications, thereby addressing, to a considerable extent, the deficiencies related to (i) and (ii). Our robustness investigation, related to (iii), empirically demonstrates that ConsDRED is significantly less sensitive to hyperparameter settings compared to traditional constrained RL approaches. The experiments further showcase the generalization capability of ConsDRED by introducing new and previously unseen external forces in real-world scenarios. While ConsDRED strikes a reasonable balance between performance and interpretability, there remains a trade-off with conservative decision-making. Hence, we must strike a balance between the desired degree of interpretability and the acceptable level of scalability in RL algorithms. Our proposed flight control framework evidences such a balance between high performance and safety constraints. A fundamental limitation is the computational complexity, where we sample at 16 Hz for disturbance estimation owing to limited onboard computational resources. This may be improved in the future by implementing dedicated hardware.}, archive = {J_TASE}, author = {Yanran Wang and David Boyle}, doi = {10.1109/TASE.2024.3432405}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5877--5894}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Constrained reinforcement learning using distributional representation for trustworthy quadrotor UAV tracking control}, volume = {22}, year = {2025}, }
• (2025). An integrated uncertainty quantification model for longitudinal and time-to-event data. TASE, 22, 5863--5876. (www)

  @article{ , abstract = {We present a novel joint prognostic framework for the integrated analysis and uncertainty quantification of longitudinal (i.e., multi-sensor degradation signals) data and time-to-event data. Specifically, the proposed method models longitudinal data using a functional principal component analysis (FPCA), while the time-to-event data is characterized by a Bayesian neural network-based Cox (BNN-Cox) model. The proposed method delivers several unique advantages: 1) Providing accurate remaining useful life (RUL) predictions while seamlessly integrating the uncertainties of both the longitudinal and time-to-event sub-models; 2) Demonstrating great flexibility in modeling both data types; 3) Allowing online, real-time updates of the RUL distribution as new measurements are collected; and 4) Making reliable predictions under limited data availability. Compared to existing methods that provide limited uncertainty information restricted to a single sub-model, the proposed approach offers more accurate and comprehensive uncertainty information via uncertainty propagation. The numerical evaluations on simulated and real-world data suggest that the proposed method achieves outstanding performance compared to existing benchmarks. Note to Practitioners—This paper is motivated by the practical issue of extracting prognostic insights from longitudinal and time-to-event data. There are two fundamental research questions involved: 1) How to accurately model both types of data without resorting to restrictive parametric assumptions; and 2) how to seamlessly integrate the uncertainties from both sub-models into the final RUL predictions. The proposed method is particularly useful in cases when there are modeling uncertainties in both longitudinal and time-to-event data, such as complex manufacturing or energy systems with multiple sensors, such as aircraft engines. There are four main steps involved when implementing the proposed method: 1) fit the historical longitudinal data using an FPCA-based degradation sub-model; 2) construct a BNN-Cox sub-model using the fitted longitudinal data and time-to-event data; 3) predict the degradation status and remaining useful life of the in-service units based on their longitudinal data and time-to-event data; and 4) provide uncertainty quantifications of the RUL estimates by integrating the uncertainties across the two sub-models. A key advantage of the proposed method is that practitioners can assess the reliability of RUL predictions by leveraging the well-quantified uncertainty estimates, allowing them to make well-informed maintenance decisions and avoid unnecessary operational expenses.}, archive = {J_TASE}, author = {Ye Kwon Huh and Minhee Kim and Kaibo Liu and Shiyu Zhou}, doi = {10.1109/TASE.2024.3432400}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5863--5876}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An integrated uncertainty quantification model for longitudinal and time-to-event data}, volume = {22}, year = {2025}, }
• (2025). Low-complexity decentralized output-feedback fault-tolerant control of general unknown interconnected nonlinear systems. TASE, 22, 5852--5862. (www)

  @article{ , abstract = {This paper is concentrated on the problem of decentralized output-feedback control of interconnected strict-feedback systems with actuator failures. It is focused on the cases where the virtual control coefficients of the plant are unknown; the global boundedness, matching conditions or global Lipschitz conditions of the interconnections are not assumed; the control algorithm is as simple as possible. They render the existing decentralized output-feedback fault-tolerant control designs infeasible. To address the problem, a low-complexity decentralized robust prescribed performance control approach based on a linear state transformation and an input-driven filter is put forward in this paper. It achieves the system outputs to track the corresponding references with the preassigned speed and accuracy. It is also inherently robust against the unknown system dynamics, the actuator failures, and the disturbances, thus without parameter estimation, function approximation, derivative computation, command filtering, fault detection, fault isolation or fault estimation. Finally, a comparative simulation on two inverted pendulums linked by a spring is conducted to demonstrate the developed control design. Note to Practitioners—Many complex systems, such as power systems, aerospace systems, and chemical systems, can be modeled as interconnected systems. Moreover, due to the increasing scale and complexity of engineering systems, actuator failures are becoming more likely to occur during system operation. On the other hand, both the transient and steady-state tracking performance of the systems are required to be preassigned in practical scenarios, e.g., missile interception. Existing approaches to compensate for the actuator failures guarantee only the boundedness of the tracking error under nonparametric uncertainties in the system model. This paper presents a decentralized robust prescribed performance control approach. It is inherently robust to the system nonlinearities, the actuator failures, and the disturbances. It exhibits lower costs in computation, higher efficiency in design, and is more user-friendly in implementation. It achieves trajectory tracking with preassigned rate and accuracy, despite the actuator failures. Extension of the approach to multi-agent systems with actuator failures is an interesting topic for future investigations.}, archive = {J_TASE}, author = {Jin-Zi Yang and Jin-Xi Zhang and Tianyou Chai}, doi = {10.1109/TASE.2024.3432131}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5852--5862}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Low-complexity decentralized output-feedback fault-tolerant control of general unknown interconnected nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). ROSCOM: Robust safe reinforcement learning on stochastic constraint manifolds. TASE, 22, 5841--5851. (www)

  @article{ , abstract = {Reinforcement Learning (RL) has demonstrated remarkable success across various domains. Nonetheless, a significant challenge in RL is to ensure safety, particularly when deploying it in safety-critical applications such as robotics and autonomous driving. In this work, we develop a robust and safe RL methodology grounded in manifold space. Initially, we construct a constrained manifold space, taking safety constraints into consideration. We then propose a robust safe RL approach, supported by theoretical analysis, based on the value at risk and conditional value at risk, in order to enhance the robustness of safety. Our methodology is designed to ensure safety within stochastic constraint environments. Following the theoretical analysis, we develop a practical, safe algorithm to search for a robust safe policy on stochastic constraint manifolds (ROSCOM). We evaluate the effectiveness of our approach through circular motion and air-hockey tasks. Our experiments demonstrate that ROSCOM outperforms existing baselines in terms of both reward and safety. Note to Practitioners—Real-world applications often involve inherent uncertainties, noise, and high-dimensional spaces. This complexity accentuates the urgency and challenge of ensuring safety in robot learning, especially when implementing RL in practical environments. To address this critical issue, we build a stochastic constraint manifold to delineate the safety space, thus establishing a rigorous framework for robot learning at each iteration. Compared with state-of-the-art baselines, our method can provide remarkable performance regarding safety and reward performance. For example, in an air hockey robot learning task, our method has demonstrated a remarkable 50% enhancement in safety performance compared to the ATACOM framework, while concurrently exhibiting superior reward performance. Moreover, in contrast to traditional algorithms, including CPO, PCPO, our method has achieved a 99% improvement in safety performance, coupled with significantly superior reward performance. These empirical insights render our approach not only theoretically sound but also practically efficacious, indicating its potential as a useful tool in real robot learning and beyond.}, archive = {J_TASE}, author = {Shangding Gu and Puze Liu and Alap Kshirsagar and Guang Chen and Jan Peters and Alois Knoll}, doi = {10.1109/TASE.2024.3431530}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5841--5851}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ROSCOM: Robust safe reinforcement learning on stochastic constraint manifolds}, volume = {22}, year = {2025}, }
• (2025). Co-design of a switching-type control scheme for nonlinear networked systems with protocol-based communication and its application to circuits. TASE, 22, 5828--5840. (www)

  @article{ , abstract = {This article investigates the resilient security control problem of nonlinear network systems under randomly activated deception attacks and a two-phase data transfer mechanism (TP-DTM). Initially, in order to solve the problem of high conservatism caused by the one-order free-weighting matrix (OFM) method in previous research results, this study introduces a high-order multi-mode switching free-weighting matrix (HMSFM) mechanism to establish free-weighting matrix groups under various switching modes, At the same time, sometime-varying balance matrices (TVBMs) are proposed to cooperate with HMSFM to collect proprietary features about each switching mode. Furthermore, in the context of the system subjected to deception attacks, a TP-DTM employing adaptive event-triggered mechanism (AETM) and round robin (RR) protocol is proposed to alleviate the impact of limited communication resources. Then, considering the above factors, the asymptotic stability of the augmented fuzzy system can be guaranteed under the given sufficient criteria. Consequently, a numerical example and a tunnel diode circuit example are executed to verify the advancement of the developed results. Note to Practitioners—The motivation of this paper comes from the problem that nonlinear networked control systems are vulnerable to malicious network attacks in the actual environment. In particular, malicious attacks will bring security risks such as data tampering and functional damage to the actual communication circuit system, and many state variables cannot be directly monitored due to objective conditions or financial constraints. Based on this, this paper proposes a HMSFM-based co-design mechanism for nonlinear NCSs under malicious attacks. The improved co-design scheme can deal with some difficult-to-measure state variables through observer-assisted tracking control. The main difficulty of this paper is how to reduce the conservatism of resilient control design. The improved co-design scheme can be adjusted according to the dynamic characteristics of the system by combining with TVBMs, making the constraints more relaxed and the state estimation more accurate.}, archive = {J_TASE}, author = {Yu Shan and Xiang-Peng Xie and Zehui Mao}, doi = {10.1109/TASE.2024.3431235}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5828--5840}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Co-design of a switching-type control scheme for nonlinear networked systems with protocol-based communication and its application to circuits}, volume = {22}, year = {2025}, }
• (2025). Recognizing video activities in the wild via view-to-scene joint learning. TASE, 22, 5816--5827. (www)

  @article{ , abstract = {Recognizing video actions in the wild is challenging for visual control systems. In-the-wild videos show actions not seen in training data, recorded from various angles and scenes with the same labels. Most existing methods address this challenge by developing complex frameworks to extract spatiotemporal features. To achieve view robustness and scene generalization cost-effectively, we explore view consistency and scene joint understanding. Based on this, we propose a neural network (called Wild-VAR) to learn view and scene information jointly without any 3D pose ground truth labels, a new approach to recognizing video actions in the wild. Unlike most existing methods, first, we propose a Cubing module to self-learn body consistency between views instead of comprehensive image features, boosting the generalization performance of across-view settings. Specifically, we map 3D representations to multiple 2D features and then adopt a self-adaptive scheme to constrain 2D features from different perspectives. Moreover, we propose temporal neural networks (called T-Scene) to develop a recognizing framework, enabling Wild-VAR to flexibly learn scenes across time, including key interactors and context, in video sequences. Extensive experiments show that Wild-VAR consistently outperforms state-of-the-art methods on four benchmarks. Notably, with only half the computation costs, Wild-VAR improves accuracy by 2.2% and 1.3% on the Kinetics-400 and the Something-Somthing V2 datasets, respectively. Note to Practitioners—In human-robot interaction tasks, video action recognition technology is a prerequisite for visual control. In real applications, humans move freely in 3D space, which results in significant changes in the view of video capture and constantly changing scenes. Deep Neural Networks are limited by the perspectives and scenarios contained in the training data, resulting in most existing methods are only effective for identifying actions from 2–4 fixed views, and the background is single. Therefore, existing models are often difficult to generalize to unconstrained application environments. Human view and video scene understanding are often treated separately. Inspired by the human visual system, this paper proposes a view-to-scene video processing method in a cost-efficient way. In real-world applications, this lightweight method can be integrated into robots to help identify human behavior in complex environments. Fewer parameters indicate that the method can be easily migrated to different types of behaviors, and the reduced computational costs represent the ability to achieve real-time performance under limited hardware conditions.}, archive = {J_TASE}, author = {Jiahui Yu and Yifan Chen and Xuna Wang and Xu Cheng and Zhaojie Ju and Yingke Xu}, doi = {10.1109/TASE.2024.3431128}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5816--5827}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Recognizing video activities in the wild via view-to-scene joint learning}, volume = {22}, year = {2025}, }
• (2025). L₁ adaptive control-based formation tracking of multiple quadrotors without linear velocity feedback under unknown disturbances. TASE, 22, 5804--5815. (www)

  @article{ , abstract = {This paper addresses the problem of formation control for a quadrotor swarm (QS) system with directed graph topology under external environmental disturbances and unreliable internal state acquisition. The proposed distributed robust control framework, based on a gemetric controller, incorporates ${\mathcal {L}}_{1}$ adaptive controllers and differentiator systems. First, the geometric formation controller is designed to implement the formation control of the nominal system. Then, ${\mathcal {L}}_{1}$ adaptive controllers are designed separately for each quadrotor’s position loop and attitude loop subsystems to address the effects of uncertainties such as external time-varying disturbances (matched and unmatched disturbances) and different mass variations of quadrotors. Furthermore, the differentiator system is devised to accurately estimate the higher-order derivatives of the non-directly-measurable velocity information and the virtual translation control signal, which enhances system accuracy while reducing computational complexity. The Lyapunov stability theory is employed to analyze the stability of the closed-loop system. Finally, the effectiveness and exceptional performance of this approach in QS formation control were validated through numerical simulation and experimental results. Note to Practitioners—The inspiration for this article comes from the issue of formation control in a cluster of quadrotor drones, which is also applicable to formation control in other types of drones. In this paper, a formation control algorithm based on ${\mathcal {L}}_{1}$ adaptive control strategy and arbitrary-order differentiation is designed. This algorithm can address not only the issue of time-varying wind disturbances frequently encountered during quadrotor drone flights but also the effects of unpredictable velocities and inconsistent masses of quadrotor drones. The disturbance rejection capability of this scheme enables quadrotor drones to be applied more safely and reliably in complex environments for search and rescue missions and surveillance tasks. Eliminating the need for linear velocity measurements reduces sensor costs and enhances system reliability and stability. The proposed formation control scheme allows the QS system to have different masses for each UAV, which can be applied to tasks such as collaboration logistics transportation, material delivery and crop spraying. Preliminary physical experiments have validated the feasibility of the proposed scheme, although it has not been applied in practical scenarios yet. In future research, we intend to equip each drone in the QS system with objects of different masses to achieve collaboration material transportation and delivery in complex environments.}, archive = {J_TASE}, author = {Yang Hu and Zhiqiang Miao and Yaonan Wang and Haoming Tang and Xiangke Wang and Wei He}, doi = {10.1109/TASE.2024.3431019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5804--5815}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {L₁ adaptive control-based formation tracking of multiple quadrotors without linear velocity feedback under unknown disturbances}, volume = {22}, year = {2025}, }
• (2025). Progressive feature enhancement network for automated colorectal polyp segmentation. TASE, 22, 5792--5803. (www)

  @article{ , abstract = {In recent years, colorectal polyp segmentation has attracted increasing attention in academia and industry. Although most existing methods can achieve commendable outcomes, they often confront difficulty when localizing challenging polyps with complex background, variable shape/size, and ambiguous boundary, because of the limitations in modeling global context and in cross-layer feature interaction. To cope with these challenges, this paper proposes a novel Progressive Feature Enhancement Network (PFENet) for polyp segmentation. Specifically, PFENet follows an encoder-decoder structure and utilizes the pyramid vision transformer as the encoder to capture multi-scale long-term dependencies at different stages. A cross-stage feature enhancement (CFE) module is embedded in each stage. The CFE module enhances the feature representation ability from interaction among adjacent stages, which helps integrate scale information for recognizing polyps with complex background and variable shape/size. In addition, a foreground boundary co-enhancement (FBC) module is used at each decoder to simultaneously enhance the foreground and boundary information by incorporating the output of the adjacent high stage and the coarse segmentation map, which is generated by fusing features of all four stages via a coarse map generation module. Through top-down connections of FBC modules, PFENet can progressively refine the prediction in a coarse-to-fine manner. Extensive experiments show the effectiveness of our PFENet in the polyp segmentation task, with the mIoU and mDic values over 0.886 and 0.931 tested on two in-domain datasets and over 0.735 and 0.809 tested on three out-of-domain datasets.Note to Practitioners—Automated and accurate polyp segmentation in colonoscopy images is a critical prerequisite for subsequent detection, removal, and diagnosis of polyps in clinical practice. This paper proposes a novel deep neural network for polyp segmentation, termed PFENet, with a CFE module to enhance the feature representation ability for better capturing polyps with complex background and variable shape/size, and a FBC module to simultaneously enhance the foreground and boundary information on the feature representation provided by the CFE module. Qualitative and quantitative results on five public datasets show that our PFENet yields accurate predictions and is superior to 9 state-of-the-art polyp segmentation methods. The proposed PFENet will facilitate potential computer-aided diagnosis systems in clinical practice, in which it can better promote medical decision-making than competing methods in polyp detection and removal.}, archive = {J_TASE}, author = {Guanghui Yue and Houlu Xiao and Tianwei Zhou and Songbai Tan and Yun Liu and Weiqing Yan}, doi = {10.1109/TASE.2024.3430896}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5792--5803}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Progressive feature enhancement network for automated colorectal polyp segmentation}, volume = {22}, year = {2025}, }
• (2025). Adaptive event-triggered heterogeneous consensus of leader-following multi-agent systems with nonlinear dynamic topology. TASE, 22, 5781--5791. (www)

  @article{ , abstract = {This paper is interested in the event-triggered leader-following consensus problem for discrete-time multi-agent systems subject to nonlinear dynamic topology. Considering that the on-board resources of each agent are generally limited, a novel adaptive event-triggered strategy is presented in this paper through establishing a nonlinear transformation law of consensus errors, which provides an effective way to weaken the impact of small fluctuations on triggering behaviors after the error systems converge. Then, an interval type-2 fuzzy model is introduced to describe the nonlinear dynamic topology where the dynamic characteristics contain both nonlinear time-varying law and uncertain parameter. In view of the discrepancies in adjacency relationships between different agents, the heterogeneous fuzzy-dependent controllers are employed to further decrease the consensus error. Meanwhile, some sufficient conditions are deduced to solve the designed controllers while ensuring that the consensus of multi-agent systems can be achieved with desired $H_{\infty }$ performance. Ultimately, advantages of the presented leader-following control strategy are illustrated by two examples.Note to Practitioners—Many practical control scenarios, such as formation of unmanned air vehicles, can be described by a leader-following consensus problem of multi-agent systems with limited onboard resources. To save the resource consumption of agent, a novel event-triggered scheme is provided, which has more potential to reduce the computational tasks to a greater extent. In addition, aiming to promote the convergence of tracking error, the agent will tend to strengthen connections with the one that have not yet reached consensus, and weaken connections with the others. Therefore, a nonlinear dynamic topology with time-varying weight is introduced during the design of heterogeneous controller, such that the adjacency relationship between agents can be adjusted online. The effectiveness of the obtained results is demonstrated by an actual experiment. It is expected that the presented strategy can be further extended to multi-agent systems where network structures are more complex, such as the one with multiple groups of agents.}, archive = {J_TASE}, author = {Zehui Xiao and Zijing Xiao and Jie Tao and Chang Liu and Peng Shi and Imre J. Rudas}, doi = {10.1109/TASE.2024.3430895}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5781--5791}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive event-triggered heterogeneous consensus of leader-following multi-agent systems with nonlinear dynamic topology}, volume = {22}, year = {2025}, }
• (2025). Data-driven energy modeling of machining centers through automata learning. TASE, 22, 5769--5780. (www)

  @article{ , abstract = {The paper addresses the problem of estimating the energy consumed by production resources in manufacturing so that alternative process designs can be compared in terms of energy expenditure. In particular, the proposed methodology focuses on Computer Numerical Controlled (CNC) machining centers. Classical approaches to energy modeling require high expertise and large development effort since, for example, data acquisition is resource-specific and must be repeated frequently to avoid obsolescence. An automated and flexible data-driven methodology is designed in this work. A data-driven method is employed to learn a hybrid and stochastic model of a CNC machining center’s energetic behavior. The learned model is used to provide offline energy consumption estimates of simulated part-programs before the actual execution of the cutting. Numerical results show the performance of the proposed method on a set of case studies. The methodology is also applied to a real industrial application, including data collected during machine production. Note to Practitioners—This article provides a flexible and autonomous data-driven approach to building models representing the energetic behavior of production resources, particularly CNC machining centers. The learned models can predict machine energy consumption while executing complex part-programs. The algorithm uses data that are commonly acquired by contemporary machine monitoring systems and does not require ad-hoc experimental tests for training. Specifically, it requires the spindle rotary speed signal, part load/unload signal, and spindle (or machine) power signal during the learning phase, whilst the estimation phase uses only the load/unload and spindle speed simulated signals.}, archive = {J_TASE}, author = {Livia Lestingi and Nicla Frigerio and Marcello M. Bersani and Andrea Matta and Matteo Rossi}, doi = {10.1109/TASE.2024.3430394}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5769--5780}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven energy modeling of machining centers through automata learning}, volume = {22}, year = {2025}, }
• (2025). An adaptive finite-time sliding mode control for retinal vein micro-puncture with silicon phantom. TASE, 22, 5757--5768. (www)

  @article{ , abstract = {Retinal vein occlusion (RVO) is a prevalent ocular pathology that may result in hemorrhage and even blindness. Currently, a procedure termed retinal vein cannulation (RVC), involving puncturing the retinal vein and injecting medication, has been developed. However, RVC requires extremely high precision at the micron scale. To address the challenges in the micro-puncturing process of the RVC, an adaptive finite-time sliding mode (AFSM) control scheme with a smooth motion generator has been proposed to assist surgeons in achieving precise micro-punctures using a piezo-actuated stage. Firstly, an S-curve-based smooth motion planning approach incorporating force feedback is designed to detect the successful micro-puncture state, thus addressing the challenge of limited force perception during the procedure. Subsequently, an AFSM control scheme has been developed to track the desired motion. Finally, a micro-puncture system, equipped with a silicon phantom, is established for experimental purposes. The experimental results demonstrate that the proposed control scheme significantly enhances the tracking performance during the micro-puncture process. The smooth motion planning and AFSM control scheme prove to be effective for the automatic control of the piezo-actuated end-effector, thereby providing improved assistance to surgeons in the RVC process. Note to Practitioners—During the procedure of retinal vein micro-puncture, it is crucial to ensure a smooth motion planning and accurate tracking to guide the needle tip into the retinal vein lumen. In light of the difficulties in lack of depth perception, a motion generator has been proposed with an adaptive micro-puncture state detection mechanism based on force feedback. To reliably track the desired motion, the AFSM controller is designed to ensure tracking accuracy and robustness, and finite-time stability. In particular, the adaptive gain of the AFSM controller does not require uncertain prior information, making it friendly to clinical applications. The experimental results based on silicone phantom, demonstrate the effectiveness of the controller in achieving successful micro-puncture with precise tracking performance. The implementation of the AFSM controller enables the automated micro-puncture task, reducing the risk of damage during operation.}, archive = {J_TASE}, author = {Bo Hu and Shiyu Xu and Rongxin Liu and Xin Zhao and Mingzhu Sun}, doi = {10.1109/TASE.2024.3430386}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5757--5768}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptive finite-time sliding mode control for retinal vein micro-puncture with silicon phantom}, volume = {22}, year = {2025}, }
• (2025). Design of an assistive controller for physical Human–Robot interaction based on cooperative game theory and human intention estimation. TASE, 22, 5741--5756. (www)

  @article{ , abstract = {This article aims to design an assistive controller for physical Human-Robot Interaction (pHRI) based on Dynamic Cooperative Game Theory (DCGT). In particular, a distributed Model Predictive Control (dMPC) is formulated based on the DCGT principles (GT-dMPC). For proper implementation, one crucial piece of information regards human intention, which is defined as the desired trajectory that a human wants to follow over a finite rolling prediction horizon. To predict the desired human trajectory, a learning model is composed of cascaded Long-Short Term Memory (LSTM) and Fully Connected (FC) layers (RNN+FC). Iterative training and Transfer Learning (TL) techniques are proposed to adapt the model to different users. The behavior of the proposed GT-dMPC framework is thoroughly analyzed with simulations to understand its applicability and the tuning of its parameters for a pHRI assistive controller. Moreover, real-world experiments were carried out on a UR5 robotic arm equipped with a force sensor was installed. First, a brief validation of the RNN+FC model integrated with the GT-dMPC is proposed for the iterative procedure and the TL. Finally, an application scenario is proposed for co-manipulating two objects and comparing the obtained results with other controllers typically used in the pHRI. Results show that the proposed controller reduces the required force of the human in completing tasks, even in the presence of unknown and different loads and inertia. Moreover, the proposed controller allows for precise reaching of the target point and does not introduce any undesirable oscillations. Finally, a subjective questionnaire shows that the proposed controller is, in general, preferred by different users. Note to Practitioners—This work presents a method to design an assistive controller to help a human perform physically coupled shared tasks with a robot. The target applications of this work are co-handling tasks of large or heavy objects. Such tasks require two agents to be performed easily, and the proposed work aims to make the robot a companion for the human partner. The proposed approach also quickly adapts to new users or tasks, making it feasible for real production systems or daily scenarios. Another possible target application is the co-manipulating large flexible components such as carbon fiber plies. This application would require small modifications, particularly in how the force is exchanged. Some additional/different sensors should be used, such as vision to map object deformations with virtual forces. The present work does not directly consider these kinds of applications. Indeed, this work strictly relies on force measurements that are not reliable when dealing with flexible materials, at least in a compression state. Such an issue will be investigated in future works by using vision systems to measure a virtual force that allows this method to be applicable even in the case of flexible components.}, archive = {J_TASE}, author = {Paolo Franceschi and Davide Cassinelli and Nicola Pedrocchi and Manuel Beschi and Paolo Rocco}, doi = {10.1109/TASE.2024.3429643}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5741--5756}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design of an assistive controller for physical Human–Robot interaction based on cooperative game theory and human intention estimation}, volume = {22}, year = {2025}, }
• (2025). Real-time metric-semantic mapping for autonomous navigation in outdoor environments. TASE, 22, 5729--5740. (www)

  @article{ , abstract = {The creation of a metric-semantic map, which encodes human-prior knowledge, represents a high-level abstraction of environments. However, constructing such a map poses challenges related to the fusion of multi-modal sensor data, the attainment of real-time mapping performance, and the preservation of structural and semantic information consistency. In this paper, we introduce an online metric-semantic mapping system that utilizes LiDAR-Visual-Inertial sensing to generate a global metric-semantic mesh map of large-scale outdoor environments. Leveraging GPU acceleration, our mapping process achieves exceptional speed, with frame processing taking less than $7ms$ , regardless of scenario scale. Furthermore, we seamlessly integrate the resultant map into a real-world navigation system, enabling metric-semantic-based terrain assessment and autonomous point-to-point navigation within a campus environment. Through extensive experiments conducted on both publicly available and self-collected datasets comprising 24 sequences, we demonstrate the effectiveness of our mapping and navigation methodologies. Note to Practitioners—This paper tackles the challenge of autonomous navigation for mobile robots in complex, unstructured environments with rich semantic elements. Traditional navigation relies on geometric analysis and manual annotations, struggling to differentiate similar structures like roads and sidewalks. We propose an online mapping system that creates a global metric-semantic mesh map for large-scale outdoor environments, utilizing GPU acceleration for speed and overcoming the limitations of existing real-time semantic mapping methods, which are generally confined to indoor settings. Our map integrates into a real-world navigation system, proven effective in localization and terrain assessment through experiments with both public and proprietary datasets. Future work will focus on integrating kernel-based methods to improve the map’s semantic accuracy.}, archive = {J_TASE}, author = {Jianhao Jiao and Ruoyu Geng and Yuanhang Li and Ren Xin and Bowen Yang and Jin Wu and Lujia Wang and Ming Liu and Rui Fan and Dimitrios Kanoulas}, doi = {10.1109/TASE.2024.3429280}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5729--5740}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time metric-semantic mapping for autonomous navigation in outdoor environments}, volume = {22}, year = {2025}, }
• (2025). A strong UAV vision tracker based on deep broad learning system and correlation filter. TASE, 22, 5714--5728. (www)

  @article{ , abstract = {Object detection and tracking is always a challenging issue in UAV (unmanned aerial vehicle) application. Especially, in the scene of UAV-ASV (autonomous surface vehicle) cooperative system, UAV vision based target tracking performance has been suffering from the target rotation and fast motion. Aiming at optimizing the related UAV vision tracking performance, a SDSST tracker(strong discriminative scale space tracking) with automatic initialization and self-adjusting is developed in this paper. Firstly, in the step of initialization, combining the advantages of both fast optimization of BLS (Broad Learning System) and efficient image processing of CNN (convolutional neural networks), a novel DBLS (Deep Broad Learning System) is posed for the target detection. Meanwhile, a Q-learning based DBLS architecture searching is further introduced. Then, in terms of width/height ratio self-adjusting, this article proposed a novel filter state supervisor that helps to find the abnormal estimated state caused by rotation in target scale estimation. Basically, this so called filter state supervisor could take RSV (Rolling Standard Value) as input feature and give out the filter state. Finally, the abnormal filter state would be adjusted by an appropriate alternative by searching in the proposed rotation angles memory, so that an optimized self-adjusting could be realized. Meanwhile, extensive experiments are performed on data set of USV center in Qiandao Lake, yielding a competitive result compared with five other prevalent trackers. Note to Practitioners—This paper was motivated by the problem of target lost caused by sudden change of target motion in the UAV-ASV vision tracking. Usually, the rotation motion of ASV is a major operation when ASV carries out a maritime assignment. However, the poor tracking state supervision and inflexible scale updating method as well as manual initialization in the existing approaches lead to inappropriate scale and target lost when the target undergoes rotational motion. Therefor, this paper proposes a strong vision tracker (SDSST) to enhance the original DSST in the three aspects: automatic initialization, filter state supervisor, and self-adjusting. This can allow the tracker to initialize without human interference, also bad tracker state can be informed by filter state supervisor. When bad state happens that the target scale in the tracker will be updated flexibly based on proposed rotation angles memory. Finally, The proposed method is implemented on the real filed UAV vision data collected by UAV-ASV system in the Qiandao Lake. The results show that SDSST achieves competitive result compared to 7 other prevalent trackers.}, archive = {J_TASE}, author = {Mengmeng Wang and Quanbo Ge and Bingtao Zhu and Changyin Sun}, doi = {10.1109/TASE.2024.3429161}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5714--5728}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A strong UAV vision tracker based on deep broad learning system and correlation filter}, volume = {22}, year = {2025}, }
• (2025). Enhancing E-commerce warehouse order fulfillment through predictive order reservation using machine learning. TASE, 22, 5700--5713. (www)

  @article{ , abstract = {Order batching plays a pivotal role in enhancing order fulfillment efficiency within both manual and robotic warehousing systems. Rare attention has been devoted to the impact of future incoming orders on online order batching. This study addresses this gap by exploring the potential benefits of reserving suitable orders when upcoming orders share similarities with existing orders in the order pool. Specifically, we investigate the online order batching problem with predictive order reservation, employing the Ensemble Learning method, to predict similarities between current and future orders. Our proposed approach involves deliberate reservation of certain orders upon arrival, deferring their batching to a subsequent period for additional efficiency gains. To operationalize this predictive order reservation, we develop an algorithmic framework that comprehensively addresses online order batching, encompassing batching, sequencing, and assignment. Experimental results, conducted on real data from an e-commerce warehouse, demonstrate the superiority of our proposed approach over fixed and variable time-window online batching algorithms in terms of order turnover time, with improvements of up to 6.1%. Notably, the benefits are more pronounced when the order arrival rate aligns with the available picking resources. Note to Practitioners—This paper was motivated by the practical problem of predictively reserving orders to improve the holistic efficiency of online batch picking in e-commerce warehouses. Existing approaches generally have assumed that warehouse management system (WMS) releases all orders in the order pool to generate the order batch and not considered the potential benefits of order reservation. This paper proposes a new online order batching approach with order reservation using machine learning. The proposed approach could bring many benefits to practitioners. Firstly, the proposed algorithm could improve the quality of order batch and could easily be embedded in the existing WMS without interference to current operation process. Secondly, the proposed algorithm framework could be easily re-configured to cover diverse scenarios of operation strategy combination and used to evaluate the performance of different methods. Thirdly, the proposed online order batching approach with order reservation using machine learning can be used to various order picking systems including manual order picking system, human-robot collaborative order picking system and robotic order picking system. In future research, we will address the order reservation mechanism considering different objective functions, such as minimizing tardy orders or variance of order turnover time and explore its applicable scenarios.}, archive = {J_TASE}, author = {Yuexin Kang and Zhiqiang Qu and Peng Yang}, doi = {10.1109/TASE.2024.3428541}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5700--5713}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing E-commerce warehouse order fulfillment through predictive order reservation using machine learning}, volume = {22}, year = {2025}, }
• (2025). Fault-tolerant consensus control with privacy-preserving virtual layer for heterogeneous multi-agent system. TASE, 22, 5687--5699. (www)

  @article{ , abstract = {The paper investigates the privacy-preserving cooperative tracking problem under actuator faults, system uncertainties and unmatched disturbances for the heterogeneous multi-agent system (MAS). The virtual system is constructed to preserve the initial states and transient processes with the dynamic perturbation method against internal and external eavesdroppers. The real system achieves tracking to the virtual system and heterogeneous consensus by the adaptive fault-tolerant (FT) controller. It is demonstrated that for groups of agents with different initial states, the designed privacy-preserving algorithm can achieve consistent output to the eavesdropper while ensuring ultimate consensus. Compared with the existing results in the privacy-preserving problem, the requirements for an accurate fault-free model are avoided by the hierarchical design approach. Simulation examples validate the effectiveness of the suggested control scheme. Note to Practitioners—The study is motivated by the privacy-preserving problem of communication between MASs, and it applies to groups like UAVs, vehicles and robots that need to exchange information to achieve common goals. There are hostile parties who eavesdrop on messages sent between agents via network interception in order to get sensitive private information. To prevent eavesdroppers from obtaining privacy of the sent information and achieve the agent group’s common objective, communication information with perturbation and corresponding control mechanisms are established. The information protected by the paper is the agent’s initial states and transient process, and the genuine states will be wrapped with dynamic perturbation signals. However, due to the interference of the real measuring mechanism and the actuator faults, precisely measuring and controlling the genuine state is challenging. As a result, this research proposes a virtual system with privacy-preserving performance. The disguised signals of virtual system are communicated between agents, while the real system is designed to track the virtual system while suppressing disturbances and faults. Future studies will concentrate on privacy preservation with security constraints.}, archive = {J_TASE}, author = {Jialu Liu and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3428408}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5687--5699}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fault-tolerant consensus control with privacy-preserving virtual layer for heterogeneous multi-agent system}, volume = {22}, year = {2025}, }
• (2025). Parallel control with event-based adaptive critic implementation for robust optimal tracking of uncertain nonlinear systems. TASE, 22, 5673--5686. (www)

  @article{ , abstract = {This paper investigates event-based robust optimal parallel tracking control for a class of uncertain nonlinear systems via adaptive dynamic programming (ADP). Analysis reveals that optimal control of the nominal system with sufficient feedback gain leads to robust optimal control. Then, optimal control is implemented online employing a critic neural network (NN), and event-triggered mechanisms (ETMs) are explored to update the weights intermittently. Among them, a unique dynamic event-triggered mechanism (DETM), which releases data in the light of an auxiliary term designed for stability verification, merits significant focus, and the comparison emphasizes its potential for better handling practical control challenges. Finally, experimental findings highlight the feasibility of the proposed robust control method while validating the characteristics and superiority of the ETMs. Note to Practitioners—As a significant issue in practical applications, event-based robust optimal control motivates this study. Distinguished from the previous efforts, this study develops the practical control difficulty into virtual space via parallel control and proportionally increases the nominal system’s optimal control rule for robust optimal control. The relaxation of both the system’s prior knowledge and the assumption related to the input dynamics allows the proposed method to be more compatible with practical systems. Moreover, ETMs are adopted to better balance control performance and resource occupation, with segmented DETM providing a fresh insight to effectively address practical control issues, i.e., releasing more data to prevent the system from being damaged by the disturbance while invoking less data to improve resource efficiency when the system is stable. Finally, the experimental comparisons validate the theoretical results.}, archive = {J_TASE}, author = {Shanshan Jiao and Qinglai Wei and Jun Xiao}, doi = {10.1109/TASE.2024.3427771}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5673--5686}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Parallel control with event-based adaptive critic implementation for robust optimal tracking of uncertain nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Model-free inverse H-infinity control for imitation learning. TASE, 22, 5661--5672. (www)

  @article{ , abstract = {This paper proposes a data-driven model-free inverse reinforcement learning (IRL) algorithm tailored for solving an inverse $H_{\infty } $ control problem. In the problem, both an expert and a learner engage in $H_{\infty } $ control to reject disturbances and the learner’s objective is to imitate the expert’s behavior by reconstructing the expert’s performance function through IRL techniques. Introducing zero-sum game principles, we first formulate a model-based single-loop IRL policy iteration algorithm that includes three key steps: updating the policy, action, and performance function using a new correction formula and the standard inverse optimal control principles. Building upon the model-based approach, we propose a model-free single-loop off-policy IRL algorithm that eliminates the need for initial stabilizing policies and prior knowledge of the dynamics of expert and learner. Also, we provide rigorous proof of convergence, stability, and Nash optimality to guarantee the effectiveness and reliability of the proposed algorithms. Furthermore, we showcase the efficiency of our algorithm through simulations and experiments, highlighting its advantages compared to the existing methods.Note to Practitioners—Generally, the cost function for optimal tracking or imitation control is manually defined, which is a challenging task and may result in large tracking errors and slow tracking. In such cases, IRL is a powerful tool for reconstructing proper cost functions. Real-world systems, as demonstrated in practical cases, are frequently exposed to external disturbances and come with unknown models. Employing $H_{\infty } $ control is an effective strategy to handle disturbances. However, applying model-free IRL to solve the inverse problem of $H_{\infty } $ control for imitation remains an underexplored domain. This paper explores model-free inverse $H_{\infty } $ control for imitating expert behaviors, specifically addressing the time-consuming nature of the existing IRL studies that employ a two-loop iteration structure. We propose an efficient single-loop IRL algorithm with a new framework to do this. It is data-driven and model-free, eliminating the need to find an initial stabilizing control policy, which is typically challenging. Additionally, it ensures convergence, stability, and optimality with provable guarantees.}, archive = {J_TASE}, author = {Wenqian Xue and Bosen Lian and Yusuf Kartal and Jialu Fan and Tianyou Chai and Frank L. Lewis}, doi = {10.1109/TASE.2024.3427657}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5661--5672}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Model-free inverse H-infinity control for imitation learning}, volume = {22}, year = {2025}, }
• (2025). H∞ high-order repetitive control for functional electrical stimulation in intention tremor suppression. TASE, 22, 5648--5660. (www)

  @article{ , abstract = {Intention tremor is a rhythmic and involuntary limb swing movement that causes significant inconvenience to the daily life of patients. Repetitive control is suitable for functional-electrical-stimulation-based intention tremor suppression because it can significantly attenuate the periodic signals. However, the performance of the repetitive controller may be weakened in tremor suppression due to uncertainties in the dynamics of the musculoskeletal model and the tremor frequency. In this paper, we propose an advanced methodology for tremor suppression by combining $H_{\infty }$ control with high-order repetitive control. The proposed controller can not only guarantee the robust stability of the system subjected to model uncertainty, but also effectively suppress tremors with varying frequency. Comparative experiments on the unimpaired subjects and intention tremor patients were carried out to verify the effectiveness of the proposed method. The experimental statistical analysis results show that the proposed $H_{\infty }$ high-order repetitive controller can suppress tremors by up to 84.97%, which is about 11% and 31% higher than the single memory loop repetitive controller and the traditional filter-based controller, respectively. Note to Practitioners—Patients afflicted with wrist intention tremors encounter significant challenges while executing routine activities such as eating, writing, and dressing. Functional electrical stimulation for tremor suppression operates by generating electrical pulses that oppose the tremor motion, thereby inducing muscle contractions and diminishing tremor magnitude. Given the periodic nature of tremor signals acting as system disturbances, repetitive control emerges as an effective method for tremor suppression. However, the conventional repetitive controller cannot significantly improve the tremor suppression performance in practice due to the uncertain property of the musculoskeletal dynamics and the variation of tremor frequency. In this paper, a robust controller combining $H_{\infty }$ control with high-order repetitive control is proposed to address above issues. The high-order repetitive controller can effectively suppress the periodic tremor signals with varying frequency, and the $H_{\infty }$ controller can provide robust stability and the desired tracking performance by properly choosing of the weighting functions. The comparative experimental results, conducted on our self-built wrist tremor suppression experimental platform, involving unimpaired subjects and patients with intention tremors, validate that the proposed control approach enhances tremor suppression efficacy by 11% and 31%, compared to the traditional repetitive controller and the filter-based controller, respectively. The feasibility and effectiveness of the proposed approach is only initially verified by small scale test, and more clinical verifications will be carried out in tremor patients in the future.}, archive = {J_TASE}, author = {Zan Zhang and Benyan Huo and Yanhong Liu and Anqin Dong and Hongnian Yu}, doi = {10.1109/TASE.2024.3427638}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5648--5660}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {H∞ high-order repetitive control for functional electrical stimulation in intention tremor suppression}, volume = {22}, year = {2025}, }
• (2025). Time-varying formation h∞ tracking control and optimization for delayed multi-agent systems with exogenous disturbances. TASE, 22, 5637--5647. (www)

  @article{ , abstract = {This paper presents a novel time-varying formation (TVF) $H_{\infty }$ tracking controller for multi-agent systems (MASs) under exogenous disturbances and time-varying delays. In order to address the issue that most of the existing formation controllers are designed under the predetermined small time delay, a formation controller which can tolerate a large delay upper bound is designed. This is achieved by employing the Lyapunov functional method. At the same time, in order to suppress the influence of exogenous disturbances on the formation process, the robustness of the formation controller is improved through optimizing the $H_{\infty }$ performance index. To derive the gain matrix of the formation protocol, a mere two linear matrix inequalities (LMIs) need to be resolved. On this basis, differential evolution (DE) algorithm is employed to further adjust and optimize the parameters of the formation controller. Finally, it is proved that the obtained results can be directly applied to resolving the leader-follower consensus problem. Two simulation experiments are carried out to verify the superiority and effectiveness of the proposed scheme, where four followers achieve the $H_{\infty }$ formation tracking following the leader. Note to Practitioners—The motivation of this article is to address the TVF $H_{\infty }$ tracking control problem for MAS under limited network communication resources. The TVF model is extensively applied in the flocking control domain, such as robot formation control, spacecrafts formation flight and so on. In the engineering application scenarios, the communication between agents usually has the following problems: 1) The information interaction among the agents is often influenced by exogenous disturbances and time delays; 2) The existing formation tracking controllers are only suitable for handling predetermined small time delay. Thus, this paper presents a TVF $H_{\infty }$ tracking controller with excellent robustness to deal with these problems, and this controller is optimized based on a DE algorithm.}, archive = {J_TASE}, author = {Xiao-Jie Peng and Yong He and Zhouzhou Liu and Le You and Hongyi Li}, doi = {10.1109/TASE.2024.3427422}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5637--5647}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time-varying formation h∞ tracking control and optimization for delayed multi-agent systems with exogenous disturbances}, volume = {22}, year = {2025}, }
• (2025). A novel chattering-free discrete sliding mode controller with disturbance compensation for zinc roasting temperature distribution control. TASE, 22, 5627--5636. (www)

  @article{ , abstract = {Precise control of roasting temperature is paramount for optimizing production efficiency in the zinc smelting process. However, existing research mainly focuses on average temperature control, and there is little research on temperature distribution control. To achieve this, a roasting temperature distribution model is first established based on the principles of heat transfer. Second, accounting for modeling errors and environmental disturbances, a discrete sliding mode control with disturbance compensation is proposed. Besides, continuous reaching law is implemented to address issues related to chattering, so as to ensure stable roasting temperature. Finally, the quasi-sliding-mode domain of the proposed method is obtained by boundary analysis. The simulation results of roasting temperature distribution control substantiate the efficacy of the proposed approach.Note to Practitioners—Roasting temperature is the most critical temperature that directly determines product quality and stable production during the roasting process. Currently popular schemes all use average temperature as the control target. However, the average temperature does not represent the actual temperature inside the roaster. This paper aims to achieve the temperature distribution of the roaster, thereby ultimately improving product quality and ensuring safe production. This paper proposes a roasting temperature control scheme based on discrete sliding mode control. During the implementation of this method, the current temperature error distribution is used as input to adjust the zinc concentrate feeding rate in real time. Experimental simulations verified the feasibility of this method, but it has not yet been applied in actual production.}, archive = {J_TASE}, author = {Huiping Liang and Bei Sun and Biao Huang and Yonggang Li and Chunhua Yang}, doi = {10.1109/TASE.2024.3426328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5627--5636}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel chattering-free discrete sliding mode controller with disturbance compensation for zinc roasting temperature distribution control}, volume = {22}, year = {2025}, }
• (2025). Dynamic-inner white component analysis: A new latent variable model for dynamic data analytics and process monitoring. TASE, 22, 5614--5626. (www)

  @article{ , abstract = {Recently, dynamic latent variable (DLV) models have been prevalent in dynamic data modeling and process monitoring. They maximize the covariance or canonical correlation between latent components and one-step ahead prediction thereof; however, auto-correlations may still be present in residuals, resulting in unmodeled dynamics and hence compromising the efficacy of generic variability-based monitoring indices. In this work, a novel dynamic-inner white component analysis (DiWCA) model is put forth. Its salient feature lies in that residuals of DLVs are enforced to be as “white” as possible to suppress unmodeled dynamics such that process dynamics can be adequately described. The whiteness index is used to quantify the closeness of finite-length innovation time series to white noise, based on which a bi-level program is formulated to extract DLVs. Thanks to the enforced whiteness, two new monitoring statistics can be constructed in DiWCA to effectively detect whether static variations become “colore” and thus yield interpretable information that complements generic variability-based indices. Finally, case studies on the Tennessee Eastman process and a real-world vacuum furnace system show that DiWCA is capable of adequately capturing process dynamics from data and providing deep insight into process operating status via two whiteness-based monitoring statistics. Note to Practitioners—Interpretability is a crucial aspect of statistic design in process monitoring schemes. Generic test statistics in DLV models are mostly constructed based upon variability of residuals, which are only a partial indicator of the existence of temporal dynamics. In the proposed DiWCA model, residuals are endowed with desirable noise-like properties so that new whiteness-based monitoring statistics can be constructed to testify whether the residuals are “colored”. This offers meaningful information about the existence of temporal dynamics in an interpretable way that well complements generic variability-based statistics. Furthermore, the integration of whiteness-based and variability-based monitoring statistics in DiWCA can provide a panoramic view of process operating status. Moreover, on realistic industrial data DiWCA shows better generalization capability than existing DLV models, indicating that process dynamics can be faithfully captured through the maximization of whiteness.}, archive = {J_TASE}, author = {Kaida Huo and Dexian Huang and Fan Yang and Chao Shang}, doi = {10.1109/TASE.2024.3426092}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5614--5626}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic-inner white component analysis: A new latent variable model for dynamic data analytics and process monitoring}, volume = {22}, year = {2025}, }
• (2025). BDHT: Generative AI enables causality analysis for mild cognitive impairment. TASE, 22, 5601--5613. (www)

  @article{ , abstract = {Effective connectivity estimation plays a crucial role in understanding the interactions and information flow between different brain regions. However, the functional time series used for estimating effective connectivity is derived from certain software, which may lead to large computing errors because of different parameter settings and degrade the ability to model complex causal relationships between brain regions. In this paper, a brain diffuser with hierarchical transformer (BDHT) is proposed to estimate effective connectivity for mild cognitive impairment (MCI) analysis. To our best knowledge, the proposed brain diffuser is the first generative model to apply diffusion models to the application of generating and analyzing multimodal brain networks. Specifically, the BDHT leverages structural connectivity to guide the reverse processes in an efficient way. It makes the denoising process more reliable and guarantees effective connectivity estimation accuracy. To improve denoising quality, the hierarchical denoising transformer is designed to learn multi-scale features in topological space. By stacking the multi-head attention and graph convolutional network, the graph convolutional transformer (GraphConformer) module is devised to enhance structure-function complementarity and improve the ability in noise estimation. Experimental evaluations of the denoising diffusion model demonstrate its effectiveness in estimating effective connectivity. The proposed model achieves superior performance in terms of accuracy and robustness compared to existing approaches. Moreover, the proposed model can identify altered directional connections and provide a comprehensive understanding of parthenogenesis for MCI treatment.Note to Practitioners—Diagnosing MCI allows for timely intervention and treatment measures to potentially slow down or even halt further cognitive decline. Exploring causal relations between brain regions enables a better understanding of pathogenic mechanisms and the development of effective biomarkers for MCI diagnosis. The current practice heavily relies on the software to analyze MCI causality, leading to large computing errors and degrading MCI analysis performance because of different parameter settings. This work aims to provide a unified framework for the estimation of brain effective connectivity using generative artificial intelligence. Due to their ability to generate high-quality samples, diffusion models have demonstrated remarkable performance in cross-modal medical image synthesis through iterative denoising processes. Our model provides a new insight into how to transform four-dimensional functional magnetic resonance imaging into effective connectivity without relying on software toolkits. The proposed model achieves good disease prediction performance and identifies altered directional connections that may be potential biomarkers for MCI treatment. Our work enables practitioners to develop deep learning model-based medical tools to assist clinicians with disease diagnosis and pathological analysis in an efficient way. Our work can also extend to the intelligently assisted diagnosis of other neurological diseases.}, archive = {J_TASE}, author = {Qiankun Zuo and Ling Chen and Yanyan Shen and Michael Kwok-Po Ng and Baiying Lei and Shuqiang Wang}, doi = {10.1109/TASE.2024.3425949}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5601--5613}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {BDHT: Generative AI enables causality analysis for mild cognitive impairment}, volume = {22}, year = {2025}, }
• (2025). An adaptively direct decoupling register control structure for the accelerating phase of roll-to-roll manufacturing. TASE, 22, 5590--5600. (www)

  @article{ , abstract = {The accelerating phase of R2R printing systems is an uncertain, complex and coupling process, which makes the register control challenging. In this paper, an online learning algorithm is presented for a register error model to automatically approximate the accelerating dynamics in the closed-loop control process. Then, an adaptively direct decoupling control (ADDC) method is proposed to eliminate the impacts of the complex coupling and acceleration on the register errors by an adjustable decoupling compensation. The effectiveness of the ADDC method is demonstrated via experiments, while the experimental results indicate that ADDC can rapidly reduce the register errors of the accelerating process and always keep them within $\pm 5 \times 10^{-5}$ m. Moreover, comparisons of ADDC, well-tuned PD and model-based feed-forward PD control methods are also carried out, and the results show that ADDC owns the highest register precision and least waste material in the three control methods. Note to Practitioners—In this paper, an adaptively direct decoupling control (ADDC) method is proposed for the register control of the accelerating phase of Roll-to-Roll (R2R) printing systems. In the ADDC method, an online learning algorithm is utilized in a general register error model to automatically approximate the accelerating dynamics in the closed-loop control process, thereby the decoupling compensation in the ADDC method can automatically adjust to directly eliminate the impacts of the complex coupling and acceleration on the register errors. Experiments are carried out on an industrial R2R printing machine to show the effectiveness of the ADDC method, and the results show that ADDC can rapidly reduce the register errors of the accelerating process and always keep them within $\pm 5 \times 10^{-5}$ m. Moreover, comparisons of ADDC, well-tuned PD and model-based feed-forward PD control methods are also carried out, and the results show that ADDC owns the highest register precision and least waste material in the three control methods.}, archive = {J_TASE}, author = {Tao Zhang and Haopeng Tan and Zhihua Chen and Youmin Rong}, doi = {10.1109/TASE.2024.3425432}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5590--5600}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptively direct decoupling register control structure for the accelerating phase of roll-to-roll manufacturing}, volume = {22}, year = {2025}, }
• (2025). Decision making for multi-robot fixture planning using multi-agent reinforcement learning. TASE, 22, 5578--5589. (www)

  @article{ , abstract = {Within the realm of flexible manufacturing, fixture layout planning allows manufacturers to rapidly deploy optimal fixturing plans that can reduce surface deformation that leads to crack propagation in components during manufacturing tasks. The role of fixture layout planning has evolved from being performed by experienced engineers to computational methods due to the number of possible configurations for components. Current optimisation methods commonly fall into sub-optimal positions due to the existence of local optima, with data-driven machine learning techniques relying on costly to collect labelled training data. In this paper, we present a framework for multi-agent reinforcement learning with team decision theory to find optimal fixturing plans for manufacturing tasks. We demonstrate our approach on two representative aerospace components with complex geometries across a set of drilling tasks, illustrating the capabilities of our method; we will compare this against state of the art methods to showcase our method’s improvement at finding optimal fixturing plans with 3 times the improvement in deformation control within tolerance bounds. Note to Practitioners—Fixture layout planning is one of the most fundamental manufacturing tasks that must be carried out before production cycles can begin, to ensure that deformation is within tolerances to avoid crack propagation and component damage. However, reliance on humans to generate these plans has led to sub-optimal solutions, leading to manufacturers incurring losses and wanting to explore analytical methods for fixture planning. In this vein, there may be the temptation to find a single solution that can be applied to all problems regardless of complexity. However, in the age of flexible manufacturing, the benefits of building tailored solutions within a framework - referred to as “freedom within a framework” - become more apparent. This paper outlines the framework for multi-agent reinforcement learning for fixture layout planning, and demonstrates the capabilities of this framework to outperform current state of the art methods with a simple algorithm through extensive experiments on representative aerospace components. We provide code implementations of our work on GitHub (Multi-Agent Fixture Planner on GitHub).}, archive = {J_TASE}, author = {Ethan Canzini and Marc Auledas-Noguera and Simon Pope and Ashutosh Tiwari}, doi = {10.1109/TASE.2024.3424677}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5578--5589}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Decision making for multi-robot fixture planning using multi-agent reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). All agents connectivity-preserving and error-based cooperative learning control with data-filter memory-based event-triggered strategy. TASE, 22, 5566--5577. (www)

  @article{ , abstract = {This paper proposes an all agents connectivity-preserving method and a data-filter memory-based event-triggered (ET) strategy to design cooperative learning control algorithm. Firstly, a type of error functions are proposed to achieve that all agents are within the communication boundary, which do not limit the initial values of agents. The designed method can dynamically adjust the boundary function based on the initial position of agents and gradually converge to the preset communication boundary, without abandoning any agents. Secondly, a data-filter memory-based ET strategy is proposed, which includes the designed error-based data filtering rules. The filtering rules avoid the problem that the ET mechanism stores abnormal historical data when the system has faults. Moreover, the presented error-based cooperative learning adaptive protocol does not need to presuppose that neighbor weights are bounded, reducing the conservatism. Finally, based on the above works, the constructed ET control method can achieve control objectives, and the effectiveness is demonstrated through theoretical analysis and simulation results. Note to Practitioners—In practice, multiagent systems (MASs) communicate through wireless communication mostly, which inevitably leads to an upper limit on the communication distance between agents. Exceeding the limitation of communication module will cause the problem that MASs cannot achieve signal transmission. Therefore, it is crucial to design appropriate constraint methods for different initial positions of agents to ensure that all agents can enter the predetermined communication boundary. In addition, the memory ET strategy can calculate the threshold of conditions based on the historical data. But in practice, it cannot guarantee the continuous normal operation of the system. If there are abnormal values in the stored signal data, this will lead to unreasonable calculation of the threshold. In response to this issue, this paper considers additional data filtering rules to avoid storing data when systems exist faults. Meanwhile, the cooperative learning algorithm proposed in this paper can adjust the learning information weights based on the control performance of neighbor agents, and remove the assumption of bounded neighbor learning laws.}, archive = {J_TASE}, author = {Zeyi Liu and Huaguang Zhang and Jiayue Sun and Lei Wan}, doi = {10.1109/TASE.2024.3424458}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5566--5577}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {All agents connectivity-preserving and error-based cooperative learning control with data-filter memory-based event-triggered strategy}, volume = {22}, year = {2025}, }
• (2025). Meta reinforcement learning of locomotion policy for quadruped robots with motor stuck. TASE, 22, 5551--5565. (www)

  @article{ , abstract = {Significant progress has been made in enhancing the motion capabilities of quadruped robots in unstructured environments due to advancements in hardware and control algorithms. However, limited research has been conducted on the fault-tolerant control of quadruped robots, which is crucial for their operation in remote or extreme environments like disaster sites. In this paper, we primarily focus on fault-tolerant strategies for common joint-stuck situations. By leveraging the static stability of quadruped robots, it becomes possible to adjust their control policies and enable them to continue following predetermined trajectories. We introduce a contextual meta-reinforcement learning (Meta-RL) method to design fault-tolerant policies. This method infers task-related latent vectors from the context to assist in training the policy network, ensuring both conciseness and optimality in various situations. Additionally, to expedite algorithm training, we propose a reference action generator (RAG). To validate the proposed algorithm, extensive simulations and physical experiments are conducted. The results demonstrate that our method allows the robot to maintain its trajectory even when faced with motor locking. Furthermore, our method outperforms all baseline algorithms, highlighting its superiority in terms of fault tolerance. Note to Practitioners—The motivation of this article is to provide fault-tolerant policies for quadruped robots, specifically referring to the policies for joint-stuck situations. Previous fault-tolerant strategies either require individually designing control strategies for each joint stuck task, which brings a significant workload to designers, or adopting a unified strategy that cannot provide the optimal strategy for each task. In this article, we utilize the Meta-RL method to handle the joint stuck issue in robots for the first time. By combining the context encoder and RAG, we can provide more suitable policies for various motor-stuck tasks. Both the simulation and physical experiments validate the effectiveness and applicability of this method.}, archive = {J_TASE}, author = {Ci Chen and Chao Li and Haojian Lu and Yue Wang and Rong Xiong}, doi = {10.1109/TASE.2024.3424328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5551--5565}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Meta reinforcement learning of locomotion policy for quadruped robots with motor stuck}, volume = {22}, year = {2025}, }
• (2025). Probability-guaranteed distributed set-membership secure fusion estimation against nonlinear hybrid attacks. TASE, 22, 5539--5550. (www)

  @article{ , abstract = {This paper investigates the distributed secure fusion estimation problem under stochastic nonlinear hybrid attacks. Specifically, this work analyzes a hybrid attack scenario where the attacker employs a randomized approach to launch false data injection (FDI) attacks and Denial-of-Service (DoS) attacks on the measurement information communication channel. Then, an innovative distributed secure fusion estimation model is proposed, addressing three situations: DoS attacks, FDI attacks, and the absence of attacks. Following this, an existence condition is derived for the secure fusion estimator, utilizing probability-guaranteed set-membership filtering technology, to ensure that the fusion estimation error will consistently be bounded within an expected ellipsoid with the specified probability. Subsequently, a convex optimization problem involving constrained recursive matrix inequalities is formulated to compute the secure fusion estimation weight matrices. Finally, the effectiveness of the proposed probability-guaranteed set-membership secure fusion estimation (SSFE) algorithm is demonstrated through a simulation example. Note to Practitioners—The research in this paper is dedicated to addressing the problem of fusion state estimation in practical engineering tasks such as intelligent transportation, industrial manufacturing and military defense. With the increase in application requirements and process accuracy, the majority of projects demand that the true state must be bounded within a certain range, e.g., unmanned vehicle obstacle avoidance and missile precision strikes. To overcome this challenge, probability-guaranteed set-membership filtering is introduced to ensure that the fusion estimation error is bounded with a certain probability. However, due to the expanding scope of engineering applications, the system may be deployed to perform tasks in a non-secure environment, which increases the risk of malicious attacks that may lead to functional failures as well as performance degradation. Therefore, this paper simultaneously considers the scenario where the communication channels are subjected to the stochastic hybrid attacks, which can be effectively handled by utilizing the proposed probability-guaranteed SSFE algorithm. Preliminary simulations demonstrate the feasibility of the algorithm. Since the actual system may also encounter problems such as sensor energy constraints, bandwidth resource limitations, and data processing asynchrony. Therefore, in our future work, we will focus on addressing the various limitations present in the fusion estimation system and improving their resolution.}, archive = {J_TASE}, author = {Kaizhou Chen and Haiyu Song and Peng Shi and Wen-An Zhang and Li Yu}, doi = {10.1109/TASE.2024.3424190}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5539--5550}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Probability-guaranteed distributed set-membership secure fusion estimation against nonlinear hybrid attacks}, volume = {22}, year = {2025}, }
• (2025). Active data-driven model and robust control scheme for twisted tendon-sheath hysteresis system using koopman operator. TASE, 22, 5526--5538. (www)

  @article{ , abstract = {Hysteresis is a typical nonlinear characteristic that exists in mechanical systems, which brings significant challenges to the robust tracking control of twisted tendon-sheath systems. In this paper, an active data-driven model is proposed to describe the hysteresis phenomenon of a twisted tendon-sheath system based on the Koopman operator, and a robust controller is designed to cancel the effect of the model error and deal with the physical constraints in practical applications. First, by utilizing the Koopman theory, an active data-driven model is built to describe the twisted tendon-sheath hysteresis system in a straightforward linear form. Then, an active model is proposed based on a modified set-membership filter to estimate the finite-dimensional approximation error. Furthermore, a robust controller is developed by taking advantage of both the magnitude and bound of the model error (obtained by the active model) to enhance the control performance while considering security constraints. To the best of our knowledge, the rule-based constraint term is first considered in the data-driven model-based control scheme to prevent potential instabilities for the twisted tendon-sheath system. The theoretical stability of the closed-loop system is proven by using the barrier Lyapunov theory to ensure the security boundary. Extensive experiments are also carried out on a self-built robotic ureteroscopy prototype to demonstrate the superior tracking performance and robustness of the proposed method. Note to Practitioners—This paper is motivated by the accurate transmission problems of twisted tendon-sheath hysteresis systems, which aims to provide a precise active modeling method and a robust controller for the robotic-assisted instrument (e.g., endoscope, catheter, etc.) twisting in the sheath/orifice. Most existing studies on tendon-sheath hysteresis systems realize trajectory tracking controllers by using parametric-model-based compensation, which still lacks a practical data-driven modeling approach to characterize the hysteresis phenomenon in the linear form, and ignore the security constraints of tendon outputs. Based on the set-membership filter, this paper builds an active Koopman-based model, which is a practical method to follow for systems characterized by complex dynamics. Subsequently, by employing the constructed active model and a rule-based term to handle output constraints, a robust controller is elaborately designed to realize accurate tracking control for twisted tendon-sheath hysteresis systems. In particular, no priori knowledge of the complex dynamics is required in the implementation and gains selection of the proposed controller, which holds theoretically and practically significance for various tendon-sheath hysteresis systems. A series of comparative hardware experiments further validate the effectiveness and robustness of the suggested control scheme. In future work, we will aim to extend the applicability of the proposed active modeling and control scheme to interventional procedures of endoscopic operation robots for complex steerings with varying sheath configurations.}, archive = {J_TASE}, author = {Xiangyu Wang and Yongchun Fang and Jianda Han and Ningbo Yu}, doi = {10.1109/TASE.2024.3423789}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5526--5538}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Active data-driven model and robust control scheme for twisted tendon-sheath hysteresis system using koopman operator}, volume = {22}, year = {2025}, }
• (2025). Fixed-time optimal bipartite containment control for stochastic nonlinear multiagent systems with unknown hysteresis. TASE, 22, 5516--5525. (www)

  @article{ , abstract = {This paper is concerned with the fixed-time optimal bipartite containment control issue for stochastic nonlinear multiagent systems in the presence of unknown Bouc-Wen hysteresis. Initially, under the optimized backstepping framework, the approximation solution of the Hamilton-Jacobi-Bellman equation is obtained via the reinforcement learning algorithm. Then, the neural networks are utilized to shape an identifier-critic-actor structure to approximate unknown dynamics, evaluate system performance, and implement control behavior, respectively. Moreover, an intermediate variable is devised to cope with the unknown control coefficient resulting from Bouc-Wen hysteresis. Theoretical analysis shows that all states of the closed-loop system are semi-global practical fixed-time stable, and its convergence time is capable of adjusting by choosing appropriate design parameters. Finally, a numerical example is performed to solidify the effectiveness of the proposed strategy. Note to Practitioners—It is well known that fixed-time convergence is one of the significance performance metrics in many industrial systems including manipulator systems, robotic systems as well as unmanned aerial systems. Motivated by this fact, this article devotes the first attempt to investigate the fixed-time optimal bipartite containment control issue for stochastic nonlinear multiagent systems in the presence of unknown Bouc-Wen hysteresis according to the reinforcement learning control strategy. Theoretical analysis and simulation examples both show that all state signals of the closed-loop system are semi-global practical fixed-time stable, and its convergence time is capable of adjusting by choosing appropriate design parameters.}, archive = {J_TASE}, author = {Weiwei Guang and Xin Wang and Wei Zhang and Huaqing Li and Hongyi Li and Tingwen Huang}, doi = {10.1109/TASE.2024.3423466}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5516--5525}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time optimal bipartite containment control for stochastic nonlinear multiagent systems with unknown hysteresis}, volume = {22}, year = {2025}, }
• (2025). RASEC: Rescaling acquisition strategy with energy constraints under fusion kernel for active incision recommendation in tracheotomy. TASE, 22, 5501--5515. (www)

  @article{ , abstract = {Tracheotomy is commonly performed for patients needing prolonged intubation, airway obstruction, and neck injuries. Accurate placement of the incision and the tracheal window is paramount in order to avoid complications. Current surgical technique heavily relies on palpating cartilage landmarks on the neck to place the incision. In order to achieve the accelerated goals of the robot-assisted subtask in a tracheotomy, this paper proposes a novel autonomous palpation-based acquisition strategy - RASEC in the tracheal region, which can interactively predict the next acquisition point to maximize the expected information and minimize the costs of palpation procedure. We employ a Gaussian Process (GP) to model the distribution of hardness and utilize anatomical information as a priori input to guide the point of palpation for medical robots. The dynamic tactile sensor based on the resonant frequency is introduced to measure tissue hardness in the tracheal region by millimeter-scale contact to secure the interaction. We investigate the kernel fusion method to blend the Squared Exponential (SE) kernel with the Ornstein-Uhlenbeck (OU) kernel and optimize the Bayesian optimization search by leveraging the anatomical information of the larynx as a priori knowledge. Moreover, we further regularize the exploration and greed factors. The tactile sensor’s moving distance and the robotic base link’s rotation angle during the incision localization process are considered new factors in the acquisition strategy. Simulation and physical phantom experiments are conducted for comparison with state-of-the-art GP-based exploration approaches. The results show that the sensor’s moving distance was reduced by 53.1% and the rotation angle of the base was reduced by 75.2% of the previous values without sacrificing overall performance capabilities. The satisfying algorithmic index (average precision 0.932, average recall 0.973, average F1 score 0.952) with fewer central estimation distance errors (0.423 mm) and high resolution (1 mm) indicates the performance of the proposed RASEC in terms of exploration efficiency, cost awareness, and localization accuracy for incision localization and recommendation in real robot-assisted subtask in the tracheotomy procedure.Note to Practitioners—This work is well motivated to introduce the Level of Autonomy (LoA) 2 - task-level autonomy, specifically in the context of tracheotomy procedures. The incorporation of robotic palpation techniques aims to provide surgeons with enhanced capabilities for incision recommendations, which directly benefit surgeons to visualize hands-on information and localize the trachea regions more efficiently and further reduce cognitive load. To detect the trachea region for intubation incision without costly ergodic acquisition, this article suggests a highly efficient acquisition strategy utilizing the fusion kernel function and regularized impact factors, eliminating the time consumption for such localization task. The actual clinical value is that our proposed strategy can earn more time for further increasing the probability of patient resuscitation, to facilitate supervised autonomy in the real clinic scene.}, archive = {J_TASE}, author = {Wenchao Yue and Fan Bai and Jianbang Liu and Feng Ju and Max Q.-H. Meng and Chwee Ming Lim and Hongliang Ren}, doi = {10.1109/TASE.2024.3423391}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5501--5515}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RASEC: Rescaling acquisition strategy with energy constraints under fusion kernel for active incision recommendation in tracheotomy}, volume = {22}, year = {2025}, }
• (2025). Modeling-based decoding of the point pattern for active stereo vision. TASE, 22, 5491--5500. (www)

  @article{ , abstract = {Active stereo vision (ASV) computes the parallax and depth information from the coded structured light patterns. Thus, it could overcome the difficulties of measuring objects without textures and colors. Pattern decoding is used to obtain the indexed points with pixel coordinates from the acquired pattern image for the subsequent stereo matching. However, decoding of the structured light patterns at locations of color crosstalk, specular reflection and occlusion remains challenging. In this paper, we propose a pattern modeling method to decode the structured light point pattern robustly. The mapping model between the points in the designed pattern and the points in the segmented pattern is computed based on the natural neighbor interpolation. Based on the modeled points, the missing points in the segmented pattern are restored and the noise points in the segmented pattern are removed. The robustness is achieved in the sense of hundred percent point segmentation completeness. Due to the hundred percent completeness, the points in the corresponding blocks are matched directly according to their indexes. Experimental results verified the effectiveness of the proposed active stereo vision approach and also proved that the proposed pattern modeling method could solve the segmentation problems caused by specular reflection, color crosstalk and occlusion effectively. Note to Practitioners—The proposed approach tries to solve the segmentation problems caused by color crosstalk, specular reflection and occlusions for exiting point-pattern based 3D surface imaging techniques. The proposed approach utilizes the natural neighbor interpolation technique to compute the mapping model between the designed points and the segmented points robustly. Based on the mapping model, the missing points could be restored and the noise points could be removed. As a result, the robust segmentation results with all the projected points could be obtained. The segmentation problems caused by specular reflection, color crosstalk and occlusions are thus solved successfully. Although the proposed approach is developed based on our designed point pattern, it could also be easily expanded to other structured light point patterns. Currently, the proposed approach only works for the point patterns and much more work is still required before it could be applied to other types of structured light patterns, such as the line patterns or the phase patterns.}, archive = {J_TASE}, author = {ZhenZhou Wang and YongCan Shuang}, doi = {10.1109/TASE.2024.3422424}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5491--5500}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling-based decoding of the point pattern for active stereo vision}, volume = {22}, year = {2025}, }
• (2025). Intermittent-type quantized control based fixed-time stability of switched filippov systems on networks. TASE, 22, 5479--5490. (www)

  @article{ , abstract = {In this paper, fixed-time and practical fixed-time control of switched Filippov systems on networks with nondifferential delays are studied. Firstly, by using the comparison principle and intermittent control interval method, intermittent fixed-time stability lemmas with unified exponential conditions are proposed. The estimation of the settling-time shows that the elasticity number in the control interval plays a vital role. In order to solve the unavailability of global information, without using the auxiliary timer function, intermittent practical fixed-time lemmas with unified exponential conditions are established, which can improve the related ones greatly. It is necessary to point out that the residual set is relevant to the constant parameters of the system. The constant parameters of the system can be used to get a needed attraction region. Based on newly established stability lemmas, the fixed-time control and practical fixed-time control of the addressed switched Filippov systems are achieved based on the designed intermittent-type quantized control strategies, which enables the quantized control to be performed in intermittent intervals. Some previous control strategies are improved. Finally, numerical examples and simulations are provided to verify the validity of the obtained results. Note to Practitioners—Intermittent control has become popular recently since it saves more costs and is more practical in engineering. As for the complex networks, it is not enough to use intermittent control only, but also to consider the bandwidth of signal transmission. So, how to enable the quantized control to be performed in intermittent intervals is an interesting question. In addition, the global information of many models is unidentifiable, practical fixed-time control will be more realistic. Is the residual set in practical fixed-time control relevant to the intermittent control interval? These questions puzzled the authors and prompted this study.}, archive = {J_TASE}, author = {Rongting Tao and Fanchao Kong and Yong Ren and Quanxin Zhu and Tingwen Huang}, doi = {10.1109/TASE.2024.3422419}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5479--5490}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intermittent-type quantized control based fixed-time stability of switched filippov systems on networks}, volume = {22}, year = {2025}, }
• (2025). Accurate and capable GNSS-inertial-visual vehicle navigation via tightly coupled multiple homogeneous sensors. TASE, 22, 5464--5478. (www)

  @article{ , abstract = {Continuous and reliable estimation of navigation states is of paramount importance in ensuring the safe operation of intelligent vehicles. The conventional global navigation satellite system (GNSS)-Inertial-Visual navigation systems have demonstrated the capability to achieve locally accurate and globally drift-free pose estimation. However, challenges such as frequent satellite signal interference, limited visual features, and sudden sensor failures can severely degrade performance or even lead to complete system collapse when utilizing minimal sensor configurations. For this reason, we propose a novel and globally drift-free tightly coupled (TC) system that can integrate any number of GNSS, inertial measurement units (IMUs), and cameras to enable accurate and robust vehicle navigation. Specifically, a stacked state estimator centered on the IMU is designed to fuse information from all sensors at the raw measurement level. The pseudorange and carrier phase measurements from all GNSS terminals are directly correlated with the core IMU, ensuring accurate and fast positioning of the system in a global frame. The feature measurements from multiple independent cameras are also used to update the states by exploiting inter-epoch geometric constraints. In addition, the multiple homogeneous IMUs can not only further improve the state estimation of the system by imposing rigid constraints on the core IMU, but also switch over in time for smooth state estimation when the core IMU are faulty. We comprehensively evaluate the state estimation accuracy and robustness of the proposed approach through a series of in-vehicle experiments and simulation experiments in real urban scenarios. The results indicate that the proposed system can achieve 93.2% availability with a horizontal position error less than 0.5 m and 97.8% availability with a heading error less than 0.2 deg in typical urban environments, significantly outperforming both conventional and state-of-the-art approaches. Note to Practitioners—This study focuses on the tight integration of multiple homogeneous and heterogeneous sensors with the goal of addressing frequent interference and degradation challenges in wide-area vehicle navigation applications. We propose a general-purpose GNSS-Inertial-Visual tight coupled framework capable of integrating any number of GNSS, IMUs, and cameras at the raw measurement level. It maximizes the use of as much sensor information as possible to achieve accurate and robust state estimation and is resilient to anomalous measurements and sensor unavailability. This solution holds practical and effective for autonomous vehicles that are now commonly equipped with multiple sensors.}, archive = {J_TASE}, author = {Zhiheng Shen and Xingxing Li and Yuxuan Zhou and Shengyu Li and Zongzhou Wu and Xuanbin Wang}, doi = {10.1109/TASE.2024.3422081}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5464--5478}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Accurate and capable GNSS-inertial-visual vehicle navigation via tightly coupled multiple homogeneous sensors}, volume = {22}, year = {2025}, }
• (2025). Enabling framework for constant complexity model in autonomous inter-reconfigurable robots. TASE, 22, 5448--5463. (www)

  @article{ , abstract = {In reconfigurable robotics, intra-reconfiguration enables a robot to change its functional abilities, while inter-reconfiguration manipulates the specification limits of the robot hardware. Although the versatility of inter-reconfigurable robots is desired in advanced autonomous systems, the O(n3) algorithm computational time complexity challenge comes when multiple modular robots combine and reconfigure into a bigger form structure for autonomous navigation tasks. This phenomenon has limited the inter-reconfiguration potential of expansion, versatility, and robustness. In this paper, a navigation framework with non-complex transformation states is proposed for inter-reconfigurable robots to perform combining and splitting control dimensions. Simulations have shown the complexity from O(n) to constant time O(1) in the reconfiguration states of the framework on a considerable number of robot agents. Additionally, a set of inter-reconfigurable robots, Wasp Biggie, was used to demonstrate the proof-of-concept in experiments as a fully functional centralized planner system. These experiments showed outperforming results on the consistent utility of CPU consumption while performing navigation and reconfiguration. Note to Practitioners—This study aims to provide controls for combining multiple robots into a single system. The research study enables the robots in the system to vary and manipulate their physical structure and mechanism limitations. Onboard computation resources are often finite and incapable of computing for multiple functions that are actively in demand. Hence, this paper is motivated by the severe increase in computational demands common in a multi-robot centralized system. The paper has provided the technical details of the system architecture of the state machine and how it integrates with a typical navigation stack. The state-machine of the framework can be easily constructed using the SMACH package in the Robot Operating System and integrated by reconstruction of the navigation stack by providing the command velocity as an input and producing the transformed command velocity for the controls of the respective robots. The outcome of the paper provides a method of control in inter-reconfigurability and takes in a consistent amount of computational resources as the number of robots varies in utilization.}, archive = {J_TASE}, author = {Ash Yaw Sang Wan and Anh Vu Le and Chee Gen Moo and Vinu Sivanantham and Mohan Rajesh Elara}, doi = {10.1109/TASE.2024.3421533}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5448--5463}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enabling framework for constant complexity model in autonomous inter-reconfigurable robots}, volume = {22}, year = {2025}, }
• (2025). Asynchronous secure control of fuzzy MJSs with time delays: A multi-level probabilistic event-triggered mechanism. TASE, 22, 5436--5447. (www)

  @article{ , abstract = {This paper investigates the event-triggered secure control problem for fuzzy Markov jump systems (MJSs) with time-varying delays. In order for higher resource efficiency, a multi-level probabilistic event-triggered mechanism (MLPETM) is newly established, which reduces the data transmission frequency by investigating possible variations of triggering thresholds. Given the network-induced phenomena, the MLPETM’s output may undergo cyber-attacks, and the controller may not switch synchronously along with the system. Asynchronous secure controllers are thus developed to perform secure control of fuzzy MJSs. Sufficient conditions dependent on fuzzy basis functions, state delays and jumping modes are established to guarantee the dissipative finite-time boundedness of the resultant closed-loop system. Afterward, an easy-to-check fuzzy secure control algorithm is developed to compute the controller gains. Finally, the one-link robotic arm system and the truck-trailer system are borrowed to validate the proposed results. Note to Practitioners—Networked control systems (NCSs) play a key role in real applications like automobile industry, power systems, and autonomous mobile robots. As a key part, the communication network such as Controller Area Network is embedded into the NCSs to perform the communication tasks between different control units. In spite of great benefits from the communication network, both the security and the efficiency issues of the data transmission are challenging for the engineering designer. Besides, the target physical system is usually nonlinear, and is very likely to undergo abrupt changes in the networked environment. In terms of these challenges, this paper studies the asynchronous security control of networked Markov jump systems with time-varying delays and randomly occurring cyber-attacks. A multi-level probabilistic event-triggered mechanism (MLPETM) is newly established, which studies the effect of the stochastic variation regarding the triggering threshold on the triggering performance. Potential cyber-attacks are modelled in a randomly occurring fashion and then absorbed into to the control law which allows operation modes asynchronous to the system modes. Finally, the one-link robotic arm system and the truck-trailer system are borrowed to check the proposed findings. Note that both the transmission security and communication traffic regulation are well concerned, which renders the developed method applicable to practical systems.}, archive = {J_TASE}, author = {Haiyang Chen and Guangdeng Zong and Xudong Zhao}, doi = {10.1109/TASE.2024.3421530}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5436--5447}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous secure control of fuzzy MJSs with time delays: A multi-level probabilistic event-triggered mechanism}, volume = {22}, year = {2025}, }
• (2025). Robust sensor fusion and biomimetic control of a lower-limb exoskeleton with multimodal sensors. TASE, 22, 5425--5435. (www)

  @article{ , abstract = {This article presents a systematic approach to robustly control a lower-limb exoskeleton in real-time using multimodal sensors. The control adopts two sensor bands that combine an array of force-sensitive resistors (FSR) and an inertial measurement unit (IMU) to measure both force myography (FMG) signals and limb motion. A robust sensor fusion algorithm that combines FMG and IMU signals with artificial neural networks is developed to accurately estimate the wearer’s hip and knee rotation angles. Moreover, a mathematical model of the lower-limb exoskeleton is calibrated and validated with real-time experimental data. Finally, a model-based controller is designed to track position references generated from the network through linear matrix inequalities. The biomimetic control algorithm is tested in simulation and a physical setup to show the effectiveness of the novel control method.Note to Practitioners—This work addresses the challenge of real-time control of lower-limb exoskeletons. Our approach to address the trajectory generation problem involves emulating the movements of a healthy limb. This is achieved through a robust sensor fusion algorithm to integrate data from two multimodal sensors, namely, FMG and IMU sensors, enabling a reliable estimation of the hip and knee’s angular positions. Signals are processed and neural networks are trained with exoskeleton encoder measurements as targets. Physical tests show both the accuracy and robustness of the control method. Potential applications of the new method include real-time gait analysis, control of upper limb exoskeletons, and study of the induction of neural plasticity.}, archive = {J_TASE}, author = {Juan Carlos Arceo and Lingzhou Yu and Shaoping Bai}, doi = {10.1109/TASE.2024.3421318}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5425--5435}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust sensor fusion and biomimetic control of a lower-limb exoskeleton with multimodal sensors}, volume = {22}, year = {2025}, }
• (2025). Cross-modality self-attention and fusion-based neural network for lower limb locomotion mode recognition. TASE, 22, 5411--5424. (www)

  @article{ , abstract = {Although there are many wearable sensors that make the acquisition of multi-modality data easier, effective feature extraction and fusion of the data is still challenging for lower limb locomotion mode recognition. In this article, a novel neural network is proposed for accurate prediction of five common lower limb locomotion modes including level walking, ramp ascent, ramp descent, stair ascent, and stair descent. First, the encoder-decoder structure is employed to enrich the channel diversity for the separation of the useful patterns from combined patterns. Second, a self-attention based cross-modality interaction module is proposed, which enables bilateral information flow between two encoding paths to fully exploit the interdependencies and to find complementary information between modalities. Third, a multi-modality fusion module is designed where the complementary features are fused by a channel-wise weighted summation whose coefficients are learned end-to-end. A benchmark dataset is collected from 10 health subjects containing EMG and IMU signals and five locomotion modes. Extensive experiments are conducted on one publicly available dataset ENABL3S and one self-collected dataset. The results show that the proposed method outperforms the compared methods with higher classification accuracy. The proposed method achieves a classification accuracy of 98.25% on ENABL3S dataset and 95.51% on the self-collected dataset. Note to Practitioners—This article aims to solve the real challenges encountered when intelligent recognition algorithms are applied in wearable robots: how to effectively and efficiently fuse the multi-modality data for better decision-making. First, most existing methods directly concatenate the multi-modality data, which increases the data dimensionality and brings computational burden. Second, existing recognition neural networks continuously compress the feature size such that the discriminative patterns are submerged in the noise and thus difficult to be identified. This research decomposes the mixed input signals on the channel dimension such that the useful patterns can be separated. Moreover, this research employs self-attention mechanism to associate correlations between two modalities and use this correlation as a new feature for subsequent representation learning, generating new, compact, and complementary features for classification. We demonstrate that the proposed network achieves 98.25% accuracy and 3.5 ms prediction time. We anticipate that the proposed network could be a general scientific and practical methodology of multi-modality signal fusion and feature learning for intelligent systems.}, archive = {J_TASE}, author = {Changchen Zhao and Kai Liu and Hao Zheng and Wenbo Song and Zhongcai Pei and Weihai Chen}, doi = {10.1109/TASE.2024.3421276}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5411--5424}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cross-modality self-attention and fusion-based neural network for lower limb locomotion mode recognition}, volume = {22}, year = {2025}, }
• (2025). Latency-aware control for wireless cloud-fog automation: Framework and case study. TASE, 22, 5400--5410. (www)

  @article{ , abstract = {The development of wireless communication has indeed promoted cloud-fog automation in the industry. It also introduces new issues of reliability and latency for control systems. This study sought to investigate the impacts of the commonly used industrial wireless network on the control performance parameters using a ball-and-beam (BB) time-critical balancing control system. An internal model control-inspired latency-aware wireless control framework (IMC-LA) is presented and employed in the BB system to handle the time delays and instability-creating elements introduced by wireless communication. A preliminary control assessment of the BB system under two new-generation wireless technologies, Wi-Fi 6 and 5G, was delivered. The correlation between network performance and control performance was analyzed statistically compared to the wired Ethernet condition. Test results show a dramatic decrease in position error after utilizing the proposed latency-aware wireless control framework. This study provides practical insights into the potential impacts of industrial wireless networks on control systems with a workable latency-aware wireless control approach. The methodology presented in this work has the potential to expedite the adoption of wireless communication in time-critical control. Note to Practitioners—Many factory automation processes experience undesirable latencies when transitioning from classic architecture to cloud automation architecture, particularly with the implementation of wireless communication. This paper aims to address the potential instability problem introduced by practical wireless networks and proposes a latency-aware control framework using the concept of internal model control. Meanwhile, this work also provides a way for practical operators to explore the relationship between critical parameters of communication networks and control performance parameters. The purpose of this paper is not to judge which wireless technology is better, instead we only want to demonstrate the effectiveness of the proposed latency-aware control in improving control performance under various wireless scenarios. The proposed framework is verified using the BB system for the commonly used cascaded PID control under two advanced wireless networks, 5G and Wi-Fi 6. It is also applicable to other industrial wireless networks with millisecond-class latency in the other regulatory control cases. The proposed IMC-LA wireless control framework reduces the significant effort and cost associated with constructing and tuning the controller in a real control system.}, archive = {J_TASE}, author = {Honghao Lyu and Zhibo Pang and Anna Bengtsson and Sofie Nilsson and Alf J. Isaksson and Geng Yang}, doi = {10.1109/TASE.2024.3420770}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5400--5410}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Latency-aware control for wireless cloud-fog automation: Framework and case study}, volume = {22}, year = {2025}, }
• (2025). Resilient consensus control of heterogeneous multi-UAV systems with leader of unknown input against byzantine attacks. TASE, 22, 5388--5399. (www)

  @article{ , abstract = {This paper addresses the consensus control problem of heterogeneous multi-UAV systems against Byzantine attacks. A drone compromised by Byzantine attacks transmits erroneous values to its neighbors while applying wrong input signals for itself, which is aggressive and challenging to defend. Inspired by the concept of digital twin technology, we introduce a new hierarchical protocol equipped with a virtual twin layer (TL), which decouples the challenges into two defense schemes: one against Byzantine edge attacks on the TL and the other against Byzantine node attacks on the cyber-physical layer (CPL). In the TL, we provide a topology reconfiguration strategy that enhances the resilience of the communication network by judiciously adding a minimal number of key edges. We rigorously demonstrate that the control strategy attains asymptotic consensus within a finite timeframe, given that the topology on the TL adheres to a strongly $(2f+1)$ -robustness criterion. Within the CPL, decentralized chattering-free controllers are proposed to ensure the resilient output consensus for the heterogeneous multi-UAV systems against Byzantine node attacks. Furthermore, the derived consensus controller exhibits an exponential convergence characteristic. The effectiveness and practicality of the obtained theoretical results are verified by a UAV swarm flight experiment. Note to Practitioners—Cooperative control of UAVs presents significant prospects for application and development, becoming a focal point of automatic control. By modeling UAV swarms as multi-agent systems, various complex distributed control methods have been conveniently proposed and implemented economically in practical systems. However, when certain agents are compromised and interfere with their neighbors, the whole network may become highly susceptible to failure. This paper specifically studies the resilient consensus control against the significant active internal threats, Byzantine attacks. The published results have primarily focused on cases where the leader UAV has no input signals. In practical applications, however, the leader often has a pre-established trajectory sent by the host and the followers are unaware of this input information. This significantly complicates the task of identifying Byzantine attackers. In this work, we introduce a new hierarchical protocol inspired by the concept of digital twin technology, which decouples the challenges into defense against Byzantine edge attacks on the TL and the defense against Byzantine node attacks on the cyber-physical layer. The experiment shows the feasibility and security of our control scheme, which provides valuable guidance for the practical applications of drones.}, archive = {J_TASE}, author = {Yukang Cui and Yiwen Liang and Qianyue Luo and Zhan Shu and Tingwen Huang}, doi = {10.1109/TASE.2024.3420697}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5388--5399}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient consensus control of heterogeneous multi-UAV systems with leader of unknown input against byzantine attacks}, volume = {22}, year = {2025}, }
• (2025). An error learning scenario-based scheme to quantized identification in wiener-hammerstein systems subject to deadzone nonlinearity. TASE, 22, 5375--5387. (www)

  @article{ , abstract = {Most of the existing estimation methods for nonlinear systems have been developed by using non-self-error data (e.g., prediction error and observation error, etc.), potentially resulting in a tricky problem. In this paper, we propose a new estimator design for nonlinear Wiener-Hammerstein systems subject to quantised measurements, where the self-error data (i.e., initial error and estimation error) are used. For this purpose, the estimation error information is derived by introducing several auxiliary variables with an error feedback filter. Then, a compensated estimation error variable is designed to remove the hostile effect of the regressor matrix on the estimator. A novel adaptive parameter estimation learning law is proposed based on a performance evaluation function, where the compensated estimation error term, initial error term and several restraint conditions are used to construct the aforementioned evaluation function. In addition, the online verification of persistent excitation (PE) condition is also provided. Finally, the efficiency and availability of the proposed scheme are validated through numerical examples and experiment in comparison with the available estimation algorithms. Note to Practitioners—This study was motivated by the system modelling and identification problem of the servomechanism but is also uses to other nonlinear systems that have deadzone, saturation, backlash and hysteresis nonlinearity characteristics. Available methods to use common error data to establish an estimator that produces biased estimate and initial-value problems. This study introduces a new scheme using the self-error data to establish an estimator, to provide a new framework of identification method design, and to improve above-mentioned problems. The self-error data are directly related to parameter adaptive updates, thus giving positive estimation performance. In this study, we mathematically characterize the dynamical equations for the servomechanism. We introduce how to extract self-error data from the input and output data of system. This is the key step for us to construct an estimator in the future. Then, based on self-error data and several constraint conditions, a novel estimator is provided using recursive pattern. Preliminary practical experiments indicate that the proposed method is feasible but it has not yet been tested in the complex industrial production.}, archive = {J_TASE}, author = {Linwei Li and Yanfeng Wang and Xin Wang and Huanlong Zhang and Xuemei Ren}, doi = {10.1109/TASE.2024.3420447}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5375--5387}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An error learning scenario-based scheme to quantized identification in wiener-hammerstein systems subject to deadzone nonlinearity}, volume = {22}, year = {2025}, }
• (2025). Optimization of steel plate yard outbound operation toward multiple cutting lines. TASE, 22, 5360--5374. (www)

  @article{ , abstract = {As the most important session of the intelligent yard, this paper considers the optimization of steel plate yard outbound operation that orients to multiple cutting lines. According to the cutting requirements, the yard should retrieve cutting plates distributed in different places, during which outbound orders of cutting plates and relocation strategy of blocking plates should be considered. Since outbound and cutting operations are performed simultaneously, cutting influence should be considered when formulating outbound operations. Therefore, instead of improving the outbound efficiency only, this problem aims to reduce the makespan of the entire process. We introduce a hybrid BVNS-RSP algorithm to solve this problem. The algorithm uses basic variable neighborhood search to decide outbound order and implements a relocation sub-problem model to determine the relocation strategy. Two initial heuristics are developed to fasten the search efficiency. Experimenting with actual data demonstrates the performance of the algorithm. Besides, results under different cut time variations illustrate their effect on makespan, thus providing suggestions for cutting plan formulation. Note to Practitioners—This paper considers a practical problem in shipbuilding’s steel plate yard. Currently, the outbound operation of steel plate yards mainly relies on manual experience, which is inefficient. With upgrading technologies such as wireless sensors and the Internet of Things, it is possible to establish an intelligent yard. So that, given the cutting plan, the required plates can be retrieved and sent to cutting lines automatically without human intervention. The efficiency of outbound operation is vital to the whole process, and the operational planning session that translates complex cutting plans into executable operation instructions is the most crucial part of this system. This work fills in this section and improves the outbound efficiency by optimizing the crane’s trajectory and eliminating redundant moves. Besides, this paper also provides some constructive suggestions for cutting plan development, helping the shipbuilding technical center to formulate cutting plans more scientifically.}, archive = {J_TASE}, author = {Lebao Wu and Zuhua Jiang}, doi = {10.1109/TASE.2024.3420389}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5360--5374}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimization of steel plate yard outbound operation toward multiple cutting lines}, volume = {22}, year = {2025}, }
• (2025). Low overhead minimum variance time synchronization for time-sensitive wireless sensor networks. TASE, 22, 5349--5359. (www)

  @article{ , abstract = {Targeting to improve the time synchronization accuracy of multi-hop Time-Sensitive Wireless Sensor Networks (TS-WSNs) for mission-critical industrial automation applications, a Minimum Variance Time Synchronization (MVTS) algorithm utilizing the concept of Packet-Coupled Oscillators (PkCOs) is proposed. This MVTS algorithm utlizes an output feedback approach to mitigate the impact of communication noise on the accumulation of synchronization errors. In addition, a Time-Division Multiple Access (TDMA) packet-exchange superframe is introduced to achieve efficient and low-overhead time synchronization. The optimal gain matrix of the MVTS algorithm is obtained by the Linear Matrix Inequality (LMI) optimization with theoretic analysis. The proposed MVTS algorithm is evaluated by both simulation and experiments on an IEEE 802.15.4 hardware testbed. The experimental results show that the proposed algorithm can effectively reduce the growth rate of clock offset along multi-hop nodes and improve the time synchronization accuracy of the TS-WSNs. Note to Practitioners—This paper explores a method to achieve precise time synchronization in TS-WSNs. The primary challenge being addressed is the accumulation of synchronization errors that occur in multi-hop TS-WSNs, which can compromise the accuracy of time synchronization. To mitigate this challenge while taking communication overhead into account, we propose a solution that combines a TDMA-based packet-exchange superframe structure with the MVTS algorithm. This approach introduces an output feedback consensus control scheme to minimize synchronization error variance. The optimal gain matrix for this consensus control scheme is derived through LMI optimization. The algorithm is implemented on an IEEE 802.15.4-compatable wireless node SAM R21 by Microchip and the experimental results of a 10-hop netowrk shows that the maximum synchronization error is $8.32\mu s$ , reduced by 56% and 32%, repsectivly, compared to the baseline method FTSP and the recent PISync.}, archive = {J_TASE}, author = {Zhian Jia and Dongliang Cui and Xuewu Dai and Zhi-Wei Liu and Shuxin Liu and Tianyou Chai}, doi = {10.1109/TASE.2024.3419142}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5349--5359}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Low overhead minimum variance time synchronization for time-sensitive wireless sensor networks}, volume = {22}, year = {2025}, }
• (2025). Error sensitivity flexibility compensation of joints for improving the positioning accuracy of industrial robots. TASE, 22, 5335--5348. (www)

  @article{ , abstract = {Flexibility models based on the virtual joint approach (VJA) are essential for error compensation to improve the positioning accuracy of industrial robots across a range of payloads. However, current flexibility models are not accurate enough due to less consideration of deformation, or incorporate too many factors leading to difficulties in practical applications. This paper proposes a flexibility model based on the error sensitivity analysis to improve the positioning accuracy and stability of industrial robots. First, the effects of the six directions flexible deformation of the joint on the positioning error are analyzed by introducing the Sobol’s method. It indicates that the rotational deformations around X, Y and Z-axes cause the majority of positioning errors, and only a tiny minority is originated from translational deformations along X, Y, and Z-axes. Then, a mapping equation between the flexible deformations around X, Y and Z-axes and the positioning error is derived based on this observation. Finally, a flexibility model for N degrees of freedom (DoF) industrial robots is established and an identification method is presented for flexibility coefficients. The verification experiments are performed on a 6-DoF robot, and an application example is provided for error compensation in robotic assembly tasks. The experimental results show that the proposed model has higher accuracy and stability but lower calculation cost than conventional models. Moreover, after compensation, the pose error is reduced to 0.1mm and 0.03° meeting the assembly requirements in the application example. Note to Practitioners—The joint deformation under payload and link gravity is mainly responsible for the degraded positioning accuracy of industrial robots. Error compensation by flexibility models is an effective way to improve positioning accuracy. Traditional 6-DoF flexibility models are too complex to be used in industrial scenarios, while 1-DoF flexibility models are not accurate enough. This paper proposes an error sensitivity flexibility compensation that considers the main factors affecting the positioning error, while removing factors with less impact. Compared with traditional models, the proposed model combines both high accuracy and applicability. Through establishing the mathematical equation between the joint deformation and the positioning error, a new flexibility model is derived, and an easy-to-implement method is provided to identify flexibility coefficients. The proposed model can be used conveniently for high-precision compensation of errors, including applications with constant payloads, such as assembly, cutting and welding, as well as tasks with variable payloads, such as milling, drilling and de-burring. In addition, the model can also be used to optimize poses to reduce robot flexibility and enhance resistance to deformation in one workspace. Experiments indicate that high compensation accuracy can be obtained by applying the flexibility model to robot assembly tasks.}, archive = {J_TASE}, author = {Yingjie Li and Guanbin Gao and Jing Na and Yashan Xing}, doi = {10.1109/TASE.2024.3419105}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5335--5348}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Error sensitivity flexibility compensation of joints for improving the positioning accuracy of industrial robots}, volume = {22}, year = {2025}, }
• (2025). Fixturing scheme design in the shipbuilding industry: A fixture amount and layout optimization for curved compliant parts. TASE, 22, 5322--5334. (www)

  @article{ , abstract = {With the demand for flexible manufacturing of curved ship blocks, a bracket system with fixtures has emerged for positioning curved compliant parts. However, current experience-based fixturing schemes struggle to determine the appropriate amount and layout of flexible fixtures for compliant parts from various curved ship blocks. Thus, it is urgent to design fixturing schemes for corresponding parts to mitigate the impact of redundant fixtures and suboptimal layouts, ultimately reducing the total cost. In this paper, a novel fixturing scheme design policy (FSDP) is proposed for curved compliant parts. Firstly, a finite element equation-based method is built to calculate the dimensional deformation of parts under different fixturing schemes. Secondly, a fixturing scheme optimization model is formulated considering fixture amount and layout as the decision variables. The objective is to minimize the overall cost, which includes the fixture cost, quality loss, and penalty cost for violating the accuracy requirement. Finally, three curved shell plates with different aspect ratios from the shipyard are used to demonstrate the effectiveness and expandability of the FSDP. The results indicate that compared with the current fixed fixture layouts, the proposed FSDP can significantly reduce the fixture cost and quality loss by 70-85% while controlling the part deformation within the accuracy requirement. It can help shipyards adaptively optimize the flexible fixturing schemes under different compliant parts during ship construction. Note to Practitioners—This work is motivated by designing a fixturing scheme for various curved compliant parts in the shipbuilding industry. Traditional fixturing schemes result in redundant costs and poor geometrical quality of curved shell panels. The large and uneven deformation of parts will hinder the welding of subsequent stiffening members with manual trimming, which is time-consuming and labor-intensive. Thus, it is not suitable for a wide range of diverse compliant parts in the flexible construction transformation of curved blocks. In this paper, we introduce the FSDP to find optimal fixturing schemes for different curved compliant parts. Compared with the current practice, the FSDP significantly reduces the total cost. Besides, the geometrical quality of the parts is improved. The proposed policy enables practitioners to implement the optimal fixturing schemes instead of relying on experience-based current practice in the flexible construction of curved blocks.}, archive = {J_TASE}, author = {Ge Hong and Tangbin Xia and Shuo Gao and Xiaofeng Hu and Ershun Pan and Lifeng Xi}, doi = {10.1109/TASE.2024.3418813}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5322--5334}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixturing scheme design in the shipbuilding industry: A fixture amount and layout optimization for curved compliant parts}, volume = {22}, year = {2025}, }
• (2025). Event-triggered feedback control for nonlinear parabolic distributed parameter systems with time-varying delays. TASE, 22, 5308--5321. (www)

  @article{ , abstract = {This paper presents an innovative event-triggered control approach for a class of nonlinear parabolic distributed parameter systems with time-varying delays. The novel event triggering mechanism enables control or measurement signals to be updated only when a predefined trigger condition exceeds a specified threshold. Multiple actuators and sensors, strategically distributed at specific points or partial regions of the spatial domain, are employed to perform pointwise/piecewise control and measurement. Two variations of event-triggered feedback (ETF) controllers are designed to address the collocated and non-collocated observation cases based on the distributions of actuators and sensors in space, respectively. The well-posedness of the open-loop and closed-loop systems is analyzed via the $C_{0}$ -semigroup theory, respectively. Furthermore, the non-existence of zeno behavior is guaranteed by demonstrating that the inter-event time intervals are nontrivial. Finally, the proposed method is applied to address the temperature control problem in the catalytic reaction process. Numerical simulation results validate the effectiveness of the proposed ETF control method in practical applications. Note to Practitioners—This work is motivated by the temperature control challenges in catalytic reaction process, with an extended application to the production of hot-rolled steel strips. This paper proposes an innovative event-triggered control method to reduce the demands on communication and computational resources. Extensive comparative experimental results have thoroughly validated the effectiveness of the proposed method in practical applications.}, archive = {J_TASE}, author = {Yaqiang Liu and Weili Zhang and Jun-Wei Wang and Yanhong Liu and Jinzhu Peng}, doi = {10.1109/TASE.2024.3418489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5308--5321}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered feedback control for nonlinear parabolic distributed parameter systems with time-varying delays}, volume = {22}, year = {2025}, }
• (2025). Fast motion planning in dynamic environments with extended predicate-based temporal logic. TASE, 22, 5293--5307. (www)

  @article{ , abstract = {Formal languages effectively outline robots’ task specifications, yet current temporal logic struggles to balance semantic expression with solution speed. To address this challenge, we propose extended predicate-based temporal logic (E-pTL), augmenting conventional linear temporal logic with more expressive atomic predicates to reflect the time, space, and order attributes of task and represent complex tasks with dynamic propositions, time windows, and even relative time window. This approach, blending automata-based task abstraction and extended predicates, offers enhanced expressiveness and conciseness for intricate specifications. To cope with E-pTL, we introduce a novel planning framework, the Planning Decision Tree (PDT). PDT incrementally builds a tree through automata and system state searches, recording potential task plans. The proposed pruning method can reduce the exploration space. This method swiftly handles complex temporal tasks defined by E-pTL. Rigorous analysis confirms PDT-based planning’s feasibility (ensuring satisfactory planning aligned with task specifications) and completeness (guaranteeing a feasible solution if available). Moreover, PDT-based planning proves efficient, with solution times approximately linearly proportional to automaton states squared. Extensive simulations and experiments validate its effectiveness and efficiency. Note to Practitioners—Temporal logic, as a formal language, has been widely applied to describe the task of robotic systems. However, linear temporal logic (LTL) struggles with the explicit time constraints, while other temporal logics, e.g., signal temporal logic (STL), cannot compactly represent task progress via an automaton. Consequently, current solutions for complex task with temporal logic specification heavily lean on optimization-based approaches, which are time-consuming and impractical for real-time systems. This work presents a novel approach for fast motion planning in dynamic environments with extended predicate-based temporal logic (E-pTL). The proposed E-pTL not only enhances semantic expression but also can be checked by an automaton, simplifying progress checks. However, existing automata-based methods encounter limitations in dynamic tasks. Graph search methods require replanning at each time step and lack completeness guarantees, while sampling methods can’t efficiently handle the feasibility of each atomic proposition. In contrast, the developed planning decision tree (PDT)-based framework incrementally searches for automata and validates propositions without a transition system, significantly reducing the search space. The suggested pruning method further trims the search space without compromising feasibility. Overall, PDT enables rapid, reliable planning, outperforming optimization-based methods like ILP/MILP and proving well-suited for real-time applications.}, archive = {J_TASE}, author = {Ziyang Chen and Mingyu Cai and Zhangli Zhou and Lin Li and Zhen Kan}, doi = {10.1109/TASE.2024.3418409}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5293--5307}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast motion planning in dynamic environments with extended predicate-based temporal logic}, volume = {22}, year = {2025}, }
• (2025). Distributed filter under homologous sensor attack and its application in GPS meaconing attack. TASE, 22, 5284--5292. (www)

  @article{ , abstract = {This study investigates the state estimation problem of multi-agent systems under a homologous sensor attack. A distributed filter is proposed to achieve a minimum variance unbiased (MVU) estimation of system states and attacks in the presence of measurement noise. A gain matrix selection method for implementing the MVU estimation is also provided. The proposed filter can be used for positioning corrections affected by global positioning system (GPS) meaconing attacks. This study treats the positioning offset caused by GPS meaconing attacks as a zero-mean white random variable and verifies the validity of this hypothesis through experiments with real GPS signals. Moreover, this study comprehensively analyses the integration of the filters into practical systems. Finally, the effectiveness of the proposed results is verified using simulation examples. Note to Practitioners–This study introduces a filter that can achieve GPS positioning calibration under meaconing attacks. This filter treats the true positioning of the system as a state and the deviation caused by meaconing attacks as a homologous attack. The filter employs a collaborative filter to reconstruct the state, thereby enabling positioning calibration. Additionally, the study explores the relationship between the homology of meaconing attacks and the estimation error of filters. This result reveals that as the homology of attacks increases, the performance of filters also improves.}, archive = {J_TASE}, author = {Yukun Shi and Wenjing He and Li Liang and Youqing Wang}, doi = {10.1109/TASE.2024.3418386}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5284--5292}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed filter under homologous sensor attack and its application in GPS meaconing attack}, volume = {22}, year = {2025}, }
• (2025). Multi-product multi-stage multi-period resource allocation for minimizing batch-processing steel production cost. TASE, 22, 5272--5283. (www)

  @article{ , abstract = {Rational allocation of resources can improve the profit margin of a steel enterprise. This paper deals with a multi-product multi-stage multi-period resource allocation problem. In it, product manufacturing involves multiple continuous production stages, each of which has parallel machines. According to process requirements, the tasks assigned to a machine need to be produced in batches. The process route of a product is a sequential combination of machines each of which is to be selected from a stage. The process route for each product and the batching rules of each machine are known in advance. Multi-period production means that the tasks released before a planning period can be processed in any of its periods. The demand for each product type in each period and the capacity of each machine are predetermined. Considering a customer’s demand, we optimally allocate machines for products in each planning period to achieve their efficient utilization. The objective is to minimize the sum of various costs related to transportation, resources, unmet demand, and product inventory. A mixed integer linear program is developed for the concerned problem. A fix-and-optimize heuristic with variable neighborhood size is newly designed to obtain high-quality solutions. Its solutions are compared with those of CPLEX (a commercial software) given a fixed solution time. Experimental results show that it can accurately solve small-scale instances and find better solutions than CPLEX for most large-scale instances. Comparison experiments are conducted and the results show that the proposed algorithm has excellent accuracy, speed, and stability in addressing the concerned problem. Note to Practitioners—As demand for steel products gradually shows a trend towards multiple varieties, small batches, and personalized customization, it increases the difficulty for practitioners to rationally allocate resources for their production in a steel enterprise. It is hard to achieve rational material and machine resource allocation subject to complex constraints for processing multiple products in multiple production stages and periods. To deal with a multi-product multi-stage multi-period resource allocation problem, it is essential to design efficient and stable algorithms. A fix-and-optimize heuristic with variable neighborhood size is thus proposed for addressing it. The method can decompose the problem into a series of subproblems according to a decomposition scheme. They are iteratively solved. In this work, our goal is to help practitioners to deal with the challenging resource allocation problem in a short time. The effectiveness of the proposed algorithm is validated and tested by comparing its results with those of a commercially available exact solver called CPLEX on various problem instances. Extensive experimental results demonstrate its effectiveness. It can quickly solve small-scale instances with no statistically significant difference from the optimal solutions obtained by CPLEX. When addressing large-scale instances, the proposed algorithm shows better solution performance than CPLEX in a given running time. The algorithm is flexible, accurate, and fast, which implies its great application potential for resource allocation in steel enterprises.}, archive = {J_TASE}, author = {Zhuohan Zhang and Ziyan Zhao and Shujin Qin and Shixin Liu and Mengchu Zhou}, doi = {10.1109/TASE.2024.3418370}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5272--5283}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-product multi-stage multi-period resource allocation for minimizing batch-processing steel production cost}, volume = {22}, year = {2025}, }
• (2025). Adaptive control for uncertain nonlinear systems against DoS attacks using quantized output only and application to a single-link robot. TASE, 22, 5262--5271. (www)

  @article{ , abstract = {This paper investigates the problem of adaptive output feedback control for uncertain nonlinear systems with input/output quantization subject to intermittent denial-of-service (DoS) attacks. When a DoS attack becomes active, the system is unable to obtain output signal. In order to address this challenge, we compensate the output signal by introducing an attack compensator. At the same time, a novel quantization compensator-based (QC) state observer is designed, which uses only the quantized compensated signal to reconstruct the system states. In addition, the problem of over-parameterisation is avoided by using parameter projection technique to design the parameter estimator. By combining Lyapunov stability theory and a modified average dwell time (ADT) approach, the stability of the closed-loop system is ensured. Furthermore, the proposed controller ensures that all closed-loop signals are bounded, and the control performance of mean-square error can be adjusted by appropriately selecting certain design parameters. Finally, the effectiveness of the proposed control strategy is validated through a single-link robot simulation results.Note to Practitioners—Considering that resources are limited, the DoS attacks mentioned in this paper are aperiodic, which means that attackers engage in frequent attacks within a relatively short time frame. This aligns more closely with real-world scenarios of network attacks. When the attack is active, the output signal is not available for the system. In this paper, a new QC state observer is designed, which in turn enables the estimation of the system state signals. Furthermore, this paper considers both input and output quantization, which reduces duplicate transmission of samples and improves the use of communication resources. This makes the work in this paper more relevant to the practical engineering context.}, archive = {J_TASE}, author = {Shenghang Liu and Xinjun Wang and Wendi Chen and Ben Niu and Xinmin Song and Huanqing Wang and Xudong Zhao}, doi = {10.1109/TASE.2024.3418194}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5262--5271}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive control for uncertain nonlinear systems against DoS attacks using quantized output only and application to a single-link robot}, volume = {22}, year = {2025}, }
• (2025). Design and stiffness control of a variable-length continuum robot for endoscopic surgery. TASE, 22, 5251--5261. (www)

  @article{ , abstract = {Continuum robots, owing to their inherent compliance, have become essential in endoscopic surgical procedures, such as mucosal ablation. However, the prevalent design of endoscopic manipulators, which typically features only a single active bending segment, often results in limited dexterity and accessibility. Additionally, the incorporation of variable stiffness in these robots has attracted significant interest, with the aim to improve manipulation capabilities in confined spaces. In the paper, we propose a novel variable-length continuum robot with variable stiffness for endoscopic surgery. The robot’s stiffness can be altered either by modifying the catheter’s length or solid-liquid transition of low-melting-point alloy (LMPA). The design and fabrication methods of the robot are meticulously detailed. Additionally, a quasi-static stiffness model along with a learning-based stiffness compensation approach for accurate stiffness estimation are proposed. Leveraging this model, a contact force controller is designed for ablation procedure. The experimental results show that our robot possesses good flexibility and accessibility, making it highly adept at manipulating in confined spaces. Its variable stiffness feature significantly enhances its ability to counteract external disturbance and prevent tip deformation (with a average position change of 1.1mm). Finally, through force control experiments and a surgical demonstration in a gastrointestinal model, we have further validated the robot’s applicability in surgical contexts. Note to Practitioners—This paper proposed a variable-length continuum robot with variable stiffness for endoscopic surgery. The robot can achieve axial elongation and omnidirectional bending motion, having better dexterity and accessibility than traditional medical continuum robots with one active bending segment. The robot’s stiffness can be adjusted by the length changes or solid-liquid transition of low-melting-point alloy (LMPA). Besides, an accurate stiffness model and a contact force controller are proposed for endoscopic ablation surgery. By experimental results, the robot shows high flexibility and accessibility, allowing access to confined spaces for manipulation, and good control accuracy and variable stiffness capability for endoscopic surgery.}, archive = {J_TASE}, author = {Jingyu Zhang and Qin Fang and Lilu Liu and Rui Jin and Pingyu Xiang and Rong Xiong and Yue Wang and Haojian Lu}, doi = {10.1109/TASE.2024.3418092}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5251--5261}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and stiffness control of a variable-length continuum robot for endoscopic surgery}, volume = {22}, year = {2025}, }
• (2025). Learning target-directed skill and variable impedance control from interactive demonstrations for robot-assisted soft tissue puncture tasks. TASE, 22, 5238--5250. (www)

  @article{ , abstract = {A framework is proposed in this paper for learning variable impedance in percutaneous puncture surgery, with the aim of simplifying the robotic puncture of soft tissues. The framework involves simulating the dynamic changes that occur when the human arm interacts with human tissues and transferring the resulting adaptive capabilities to the robot through learning movement trends and stiffness changes. To enhance performance during task execution, we integrate the variable impedance control framework with the interactive operation and feedback controllers. To provide flexibility for trajectory modification during operation, derivative Gaussian processes are introduced to identify the target position and obtain a model of motion trends. This control law is combined with virtual dynamics that describe puncture dynamics, enabling the robot to regulate interactions and plan its trajectory. We present experiments involving tissue puncturing tasks performed by the Franka-Emika Panda robot with varying degrees of hardness. The results demonstrate that our framework is capable of learning manipulation skills for physical interaction with humans, thereby reducing application complexity in tasks involving complex force interactions for robots. Compared to using fixed or variable impedance gain controllers, our approach effectively improves the success rate, stability, and efficiency of percutaneous puncture. Note to Practitioners—This paper is motivated by the limitations encountered by robots when handling deformed objects. In traditional robot control processes, the assumption of a fixed and unchanging contact object poses a significant challenge in applying robot control to the medical industry. Consequently, it becomes imperative for robot control systems to develop stable intelligent approaches capable of interacting with deformed objects. In this paper, we propose a framework for robot-assisted puncture that combines robotic impedance control techniques with sensing mechanisms. By integrating these approaches, our framework demonstrates effectiveness in performing tasks involving soft tissues with varying levels of hardness. Our proposed method encompasses three main ideas: 1) Sensing muscle activity during task execution enables the acquisition of task parameters from the human arm. 2) The utilization of a robot control method enhances the stability of the robot’s execution process. 3) The proposed method shows potential for application in processing and treating objects with low stiffness, deformed objects, and thin-walled parts. Experimental results validate the effectiveness of the developed method. In future work, it is important for the robot-assisted puncture system to consider recognizing and localizing more diverse targets to enhance its generalization capabilities.}, archive = {J_TASE}, author = {Xueqian Zhai and Li Jiang and Hongmin Wu and Haochen Zheng and Dong Liu and Xinyu Wu and Zhihao Xu and Xuefeng Zhou}, doi = {10.1109/TASE.2024.3418018}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5238--5250}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning target-directed skill and variable impedance control from interactive demonstrations for robot-assisted soft tissue puncture tasks}, volume = {22}, year = {2025}, }
• (2025). A new stick-slip piezoelectric actuator using dual-arch bridge-type compliant mechanisms. TASE, 22, 5226--5237. (www)

  @article{ , abstract = {This paper reports a new stick-slip piezoelectric linear actuator with large load density, compact size, and high resolution. It is devised using a dual-arch bridge-type driving mechanism (DBDM) and an integrated elliptical preload mechanism(IEPM). The DBDM can generate a coupling motion on the driving point, which provides clamping and releasing actions during stick-slip motion. Also, the DBDM can magnify the clamping force with large amplification ratios, thus improving driving force and load-carrying capacity. The IEPM uses compact flexure hinges to adjust preload force, reducing structural dimension. Then, theoretical analysis and simulations are conducted, and a prototype is fabricated. A modified driving voltage with smooth turning points is designed to improve motion properties further. Experiments show that load density is 0.037 g/mm $^{\mathbf {3}}$ under the driving voltage of 100 V. The actuator has an overall size of 33.6 mm $\times 20$ mm $\times 8$ mm and a displacement resolution of 22 nm.Note to Practitioners—Most research on linear actuators focused on motion displacement improvement, and more research needs to be done on improving load-carrying capacity. Therefore, this paper designs a force amplification linear actuator with a compact structure using the DBDM. In the experimental stage, the sudden change of the driving signal affects the output performance, and a modified sawtooth voltage is used to improve the output speed and load-carrying capacity. The experimental results show that under the modified sawtooth wave voltage, the maximum speed of the slider is 1.4 mm/s, and the speed is increased by 17% compared with the traditional wave. When a 200 g load is applied to the slider, the linear actuator operates at a stable speed of $1.2~\mu $ m/s, 33% faster than conventional sawtooth waves. As a result, the designed actuator has superior performance at higher loads.}, archive = {J_TASE}, author = {Yuxi Wu and Yiling Yang and Gaohua Wu and Yuguo Cui and Yanding Wei}, doi = {10.1109/TASE.2024.3417828}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5226--5237}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A new stick-slip piezoelectric actuator using dual-arch bridge-type compliant mechanisms}, volume = {22}, year = {2025}, }
• (2025). A streamlined heuristic for the problem of min-time coverage in constricted environments. TASE, 22, 5207--5225. (www)

  @article{ , abstract = {The problem of min-time coverage in constricted environments concerns the employment of robotic fleets to support routine inspection and service operations within well-structured but constricted environments. In our previous work we have provided a detailed definition of this problem, specifying the objectives and the constraints involved, a Mixed Integer Programming (MIP) formulation for it, a formal analysis of its worst-case computational complexity, and additional structural properties of the optimal solutions that enable a partial relaxation of the original MIP formulation which preserves optimal performance. We have further employed these structural results towards the development of a construction heuristic for this problem. But while the worst-case computational complexity of the construction heuristic is polynomial with respect to the size of the problem-defining elements, its practical scalability has been limited by the requirement to formulate and solve a large number of linear programming formulations. In order to address this issue, this work presents a modified version of the heuristic that significantly reduces the computational times involved. Furthermore, we develop a local search method that further improves the solution obtained from the modified heuristic.Note to Practitioners—This paper concerns the application of the current and the emergent robotic technologies in the inspection and monitoring of remote and difficult-to access facilities, like underground utility networks and oil and gas pipeline networks. The constricted and remote nature of these environments necessitates the careful preservation of a wireless ad hoc communication network among the deployed robots and a command-&-control center that supervises the entire operation, and this need gives rise to some novel, very interesting and very challenging coordination and scheduling problems. We have undertaken the investigation of these problems in a recent series of papers, providing a systematic formal characterization of them, and some analytical results that have identified important underlying structure and have also led to the development of a pertinent heuristic approach. This work complements and extends those earlier developments by enhancing significantly the computational expediency of the aforementioned heuristic method, and further augmenting the resulting computational capability to a systematic search method that can iteratively improve any available initial solution.}, archive = {J_TASE}, author = {Young-In Kim and Spyros Reveliotis}, doi = {10.1109/TASE.2024.3417820}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5207--5225}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A streamlined heuristic for the problem of min-time coverage in constricted environments}, volume = {22}, year = {2025}, }
• (2025). Control of multiple identical mobile microrobots for collaborative tasks using external distributed magnetic fields. TASE, 22, 5193--5206. (www)

  @article{ , abstract = {The collaboration of microrobot teams has attracted considerable attention, particularly in the field of micro/nano manipulation. Achieving independent control and motion planning of multiple magnetic microrobots for coordinated movements is one of the most important tasks that is still unsolved. In this paper, a $12\times 12$ coil array system is developed to generate a series of localized magnetic fields that enable simultaneous control of multiple identical magnetic microrobots, allowing teams of microrobots to collaborate in parallel for micromanipulation tasks. First, the structure of the microcoil is optimized based on the finite element model to increase the strength and gradient of the magnetic field, which in turn enhances the driving performance of the system. Meanwhile, an improved multi-target tracking algorithm that utilizes kernel correlation filtering (KCF) and image contour detection (ICD) techniques is proposed to improve the tracking accuracy of microrobots. In addition, collaborative planning for multiple magnetic microrobots is also achieved with the combination of the conflict-based search (CBS) algorithm. Finally, the developed system is tested with extensive physical experiments. Especially, experiments on magnetic droplet transport with two microrobots are also conducted. The results impressively demonstrated the effectiveness of the devised system and the proposed methods. Note to Practitioners—This article is motivated by the recent wide interest in magnetic microrobots. Actuated by external magnetic field, magnetic microrobots can wirelessly perform targeted delivery/therapy and other micro-assembly tasks. To facilitate collaboration between microrobots, independent control of each microrobot is desirable. However, due to the interaction between magnetic microrobots and the global magnetic field, the collaboration of multiple microrobots presents great challenges. Therefore, several coil-array-based systems have been developed. In this paper, we develop a magnetic actuation system from both hardware and software aspects for the collaborative motion of multiple magnetic microrobots. The coil structure is optimized to enhance the driving performance of the devised system, and a fused multi-target tracking algorithm is proposed to improve the tracking accuracy. In combination with the CBS algorithm, collision-free paths are planned for multiple identical microrobots. The experimental results show that the constructed system and proposed methods can realize coordinated motion of multiple identical magnetic microrobots, which has enormous potential for some biomedical applications.}, archive = {J_TASE}, author = {Qigao Fan and Guangming Cui and Juntian Qu and Yueyue Liu and Xinyu Liu}, doi = {10.1109/TASE.2024.3417443}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5193--5206}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Control of multiple identical mobile microrobots for collaborative tasks using external distributed magnetic fields}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance tracking control for nonlinear multiagent systems with distributed observation errors compensation. TASE, 22, 5182--5192. (www)

  @article{ , abstract = {This paper investigates the prescribed performance formation tracking control problem for nonlinear multiagent systems (MASs) under state constraints. Compared with the existing results using predefined-time distributed observers, an observation error compensation term is designed to prevent the tracking error caused by the estimation error from violating the prescribed performance. Different from the existing related literature, this paper introduces a novel approach by combining a state-dependent transformation method with a time-varying constraint-based prescribed performance control (PPC) method. This approach effectively eliminates the feasibility conditions of state constraints and the initial condition dependence restriction of PPC. Moreover, by designing an appropriate PPC-related time-vary function, the proposed method can accurately map the preset control performance before and after the transformation process. Therefore, the loss of state feasible sets in the existing relevant results can be tackled. The simulation example verifies the feasibility of the control method. Note to Practitioners—This paper studies the prescribed performance formation tracking control problem for nonlinear MASs, which has application value in autonomous driving, smart factories, environmental exploration, unmanned delivery and other fields. By using a state-dependent transformation method and a time-varying function, a PPC strategy, under state constraints, is devised for MASs formation. Moreover, a compensation term is designed to handle the tracking errors caused by the observation errors before the predefined observation time.}, archive = {J_TASE}, author = {Xuefei Wang and Dan Ye and Lili Zhang and Xingang Zhao}, doi = {10.1109/TASE.2024.3417387}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5182--5192}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance tracking control for nonlinear multiagent systems with distributed observation errors compensation}, volume = {22}, year = {2025}, }
• (2025). Robotic perforated patch clamp based on active release control of perforating materials. TASE, 22, 5170--5181. (www)

  @article{ , abstract = {In recent years, the perforated patch clamp technique has been widely applied in cellular electrophysiology research due to its low mechanical disturbance and almost no loss of cellular content in cell membrane perforation process. However, the current passive release process of perforating materials prolongs the perforation process and easily disturbs the gigaseal formation process, significantly lowering the efficiency of perforated patch clamp operation. Addressing this, a robotic perforated patch clamp system was developed based on the active release control of perforating materials in this paper. First, a novel holding module of the patch micropipette integrated with an independently driven transmission channel of the perforating materials was developed for the first time. Then, through release tests of the perforating materials, the appropriate drive mode of transmission channel was determined to be the hydraulic mode with a faster response and higher stability. Further, the concentration gradient field of the perforating materials at the opening of the channel was modeled according to Fick’s law to prevent the false release of them in gigaseal formation process. Furthermore, a cell circuit model was developed to detect perforation degree online for feedback control of the perforating materials release. Experimental results on pyramidal neurons in mouse brain slices demonstrated that, in comparison to the traditional method with passive releases of perforating materials, the proposed system was capable of perforating cell membrane at an almost doubled throughput and with a 57% improvement in the success rate. In comparison to the traditional non-perforated whole-cell patch clamp method, the signal recording duration of neurons operated by our method was doubled due to its fewer negative influences on gigaseal and almost no cellular content loss. Note to Practitioners—The perforated patch clamp technique, utilizes the cell membrane-perforating molecules to drill subnanometer-sized conductive pores in the cell membrane aspirated into a micropipette for the measurement of cellular electrophysiological signals. Unfortunately, the poor controllability of the current passive release of perforating materials in the patch clamp operation easily leads to a long drug diffusion process, and also, disturbs the gigaseal formation between the aspirated cell membrane and micropipette, which is required for the measurement of the extremely weak cellular electrophysiological signals. For the first time, an active release control method of perforating materials was developed in this paper based on the self-developed novel three-channel holding device. With active release control of perforating materials, the proposed method was capable of perforating cell membranes at a doubled speed with a significantly higher success rate, and doubled recording duration in comparison to the traditional perforated patch clamp methods and non-perforated whole-cell patch clamp method, respectively, due to its fewer negative influences on gigaseal and almost no cellular content loss. With the above advantages, our robotic perforated patch clamp method may be applied in cellular electrophysiology research in the future.}, archive = {J_TASE}, author = {Chaoyu Cui and Ruimin Li and Shaojie Fu and Shiyu Xu and Biting Ma and Minghui Li and Jinyu Qiu and Ripeng Zhu and Xin Zhao and Qili Zhao}, doi = {10.1109/TASE.2024.3417380}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5170--5181}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic perforated patch clamp based on active release control of perforating materials}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy fixed-time control for nonlinear systems with input and output quantization. TASE, 22, 5158--5169. (www)

  @article{ , abstract = {The most existing studies on fixed-time control are restricted to nonlinear systems without the quantized signals. To remove this restriction, for a class of strict-feedback nonlinear systems (SFNSs) with input and output quantization, a new adaptive fuzzy fixed-time control scheme is presented in this article. Firstly, since the functions in SFNSs are unknown, fuzzy logic systems (FLSs) are applied to approximate them. Based on the quantized input and output signals, a novel fuzzy state observer is devised to evaluate the unmeasurable states. Secondly, in the backstepping design process, because of the discontinuity of output quantization, the partial derivatives of the virtual control signals are difficult to obtain. To get over this difficulty, a command filtering technology is adopted. Furthermore, an actual quantized fixed-time controller and novel parameter adaptive laws are established. Thirdly, to compensate the effects of the quantization errors, Lemmas 10 and 11 are presented. It is indicated that the presented control scheme can ensure the practical fixed-time stability of control systems. Finally, an example is given to demonstrate the effectiveness of the presented scheme.Note to Practitioners—In practical remote control systems, because of the limitation of communication channel capacity, signals usually need to be quantized before transmission. However, the most existing fixed-time control schemes for SFNSs do not study the signals quantization. This means that the control processes still rely on the continuous states/output and control input signals, which are not practical in engineering application. Therefore, based on FLSs and command filtering technology, this paper focuses on developing a novel adaptive fuzzy quantized fixed-time control scheme. This scheme ensures that SFNSs under the input and output quantization is still practically fixed-time stable.}, archive = {J_TASE}, author = {Xinyi Lu and Fang Wang and Zhi Liu and Ruiyi Zhu and C. L. Philip Chen}, doi = {10.1109/TASE.2024.3417312}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5158--5169}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy fixed-time control for nonlinear systems with input and output quantization}, volume = {22}, year = {2025}, }
• (2025). Semi-supervised robust hidden markov regression for large-scale time-series industrial data analytics and its applications to soft sensing. TASE, 22, 5143--5157. (www)

  @article{ , abstract = {Hidden Markov models (HMMs) for time-series data analysis are attracting wide interests in industries due to their ability to model the extensively existing dynamics and non-Gaussianities. In this paper, with the focus on industrial soft sensor applications, a semi-supervised robust hidden Markov regression (SsRHMR) model is first proposed to improve the performance of the HMMs in two challenging industrial scenarios, i.e., the scarcity of labeled samples and outlying data, which may prevent the HMMs from learning well-suited parameters. Furthermore, a distributed learning algorithm for the SsRHMR (termed D-SsRHMR) is developed to overcome the limitations of the HMMs in modeling large-scale time-series data, namely computational complexity and inability of handling long-period missing values. Performance evaluations of both the SsRHMR and D-SsRHMR are presented using a synthetic case and an actual process, based on which the effectiveness and feasibility of the proposed models and learning algorithms in improving the prediction accuracy and in accelerating the training speed have been demonstrated. Note to Practitioners—Before applying the SsRHMR to industrial soft sensing, we advise to first select features based on the process mechanisms and expert knowledge. That is, to carefully select the secondary variables so as to reduce the dimensionality of the input space. This is because, in general the lower the dimensionality of the secondary variables, the more accurate the estimated distributions of the secondary variables and the more efficient the training process for the SsRHMR. In addition, the D-SsRHMR would benefit from equal-sized subsets, since the efficiency of the distributed learning algorithm depends on the most computationally demanding slave computer, such as the one processing the largest number of data. Therefore, practically it is preferable for the D-SsRHMR to partition the entire time-series dataset with as equal size as possible.}, archive = {J_TASE}, author = {Weiming Shao and Wenxue Han and Chuanfa Xiao and Lei Chen and Meng-Qin Yu and Junghui Chen}, doi = {10.1109/TASE.2024.3417019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5143--5157}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semi-supervised robust hidden markov regression for large-scale time-series industrial data analytics and its applications to soft sensing}, volume = {22}, year = {2025}, }
• (2025). Multi-scenario cellular KPI prediction based on spatiotemporal graph neural network. TASE, 22, 5131--5142. (www)

  @article{ , abstract = {With the increasing demand for high-quality telecommunication services, cellular KPI prediction becomes crucial for telecommunication network monitoring and management. In this work, we propose a novel framework for cellular KPI prediction, which considers its distribution discrepancy under different network operation scenarios. In particular, three specific predictors for normal, target alarm, and neighbor alarm scenarios are proposed based on spatiotemporal graph neural networks and unified through transfer learning. Temporal convolution and attention mechanism are embedded to model the impact of anomalies on KPIs and its propagation across neighboring cells according to the cellular network topology. An experiment on a real cellular KPI dataset shows the effectiveness of the proposed method compared to the state-of-the-arts. Note to Practitioners—Cellular network KPI prediction under scenarios of network alarms is crucial to evaluate the impact of alarms on network services and guides cellular network maintenance policies. This problem is similar to a general multivariate time series prediction problem with data multimodality. However, the first challenge in our case is that, under different scenarios, i.e., normal, target alarm, and neighbor alarm, the effective information and spatiotemporal dependencies among KPIs are different. The second challenge is the imbalanced or sparse sample size for specific scenarios, deteriorating the model performance. This paper proposes a cellular KPI prediction framework consisting of three scenario-specific predictors with similar but different modules to process different scenario-specific data. To address the dataset imbalance across scenarios, we adopt a transfer learning strategy to unify the training and prediction of three predictors. The experiment results on a real cellular KPI dataset demonstrate that the proposed framework is more feasible and effective than the state-of-the-art models for multivariate time series prediction. Future research can consider developing maintenance policies such that the cost caused by abnormal KPIs can be minimized.}, archive = {J_TASE}, author = {Junpeng Lin and Tian Lan and Bo Zhang and Ke Lin and Dandan Miao and Huiru He and Jiantao Ye and Chen Zhang and Yan-Fu Li}, doi = {10.1109/TASE.2024.3416952}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5131--5142}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-scenario cellular KPI prediction based on spatiotemporal graph neural network}, volume = {22}, year = {2025}, }
• (2025). Learning freehand ultrasound through multimodal representation and skill adaptation. TASE, 22, 5117--5130. (www)

  @article{ , abstract = {With medical ultrasound becoming one of the most prevalent examination methods, robotic ultrasound systems offer the potential to simplify the scanning process and relieve professional sonographers from repetitive and tedious tasks. Despite recent advances, enabling robots to autonomously perform ultrasound examinations remains a challenge, mainly due to the difficulty in representing and generalizing professional ultrasound skills. In this paper, we present a comprehensive framework for learning autonomous ultrasound skills from freehand demonstrations in clinical settings. Our proposed framework consists of two key stages: offline learning and online adaptation. During the offline learning stage, ultrasound skills are encapsulated into a low-dimensional probabilistic model using a self-supervised architecture. The multimodal signals include ultrasound images, probe orientations, and contact forces. During the online adaptation stage, the model predicts the optimal actions either by direct regression or by using local exploration schemes. We perform clinical demonstrations with 24 volunteers and collect 120 experiences. Our benchmark includes 5 different tasks, including intra-patient, inter-patient, inter-sex, inter-age, and inter-obesity tasks. Both one-step and sequence-based predictions are achieved by using different variants of our framework. Customized and generic representation learning backbones are tested and analyzed. In conclusion, our autonomous ultrasound framework is flexible and robust, and potentially enriches the options for freehand/robotic ultrasound applications. Note to Practitioners—This paper is motivated by the problem of learning multimodal manipulation skills from human demonstrations, with a specific focus on freehand ultrasound skills. Our multimodal fusion framework is effective and compatible with some popular image representation backbones. Our adaptive methods and the variants have satisfactory prediction accuracy, with flexibility achieved by adjusting a few factors. We collect high-quality freehand demonstrations from ultrasound examinations in clinical settings. The data is openly available to facilitate the reproducibility of our work and to support the development of autonomous ultrasound strategies based on imitation learning.}, archive = {J_TASE}, author = {Xutian Deng and Junnan Jiang and Wen Cheng and Chenguang Yang and Miao Li}, doi = {10.1109/TASE.2024.3416827}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5117--5130}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning freehand ultrasound through multimodal representation and skill adaptation}, volume = {22}, year = {2025}, }
• (2025). Security and diagnosability of finite state machines under cyber-attacks. TASE, 22, 5108--5116. (www)

  @article{ , abstract = {In many applications, security is a serious issue due to the high risk of cyber-attacks. An adversary can cause severe damage by providing wrong information about the system and consequently leading the controller to perform incorrectly. Detecting any malicious activity is necessary to cover up its negative effects and make the system operate reliably. In this paper, we propose a new approach to analyze security and diagnosability of a Finite-state machine (FSM) under multiple attacks. Different kinds of attacks are modeled by an FSM and the composition of the nominal and attack model can express all the effects of possible attacks on the given system. We define different concepts of security and give conditions under which detectability of the attacks is possible. Moreover, diagnosability of an FSM affected by multiple attacks is addressed, and the special case of critical observability under attack is characterized. Note to Practitioners—Nowadays, cyber-physical systems (CPSs) are being widely used in industry and the extensive use of communication networks by CPSs raises the concern of vulnerability to malicious attacks. Therefore, it is a major challenge to detect the attack specifically when multiple attacks might launch on different sensors or communication channels. These facts motivate us to investigate the attack detectability properties of a system modeled by Finite State Machines (FSMs). To this end, we consider one of the well-known types of cyber-attacks which can inject, replace or remove output information in the communication network, called the man-in-the-middle attack. This may also be modeled by an FSM. In this paper, we provide a good understanding of the security level of a system under this powerful kind of attack, which can provide better insight into the weaknesses and strengths of the system before designing a supervisor. In some applications e.g. in air traffic control, the designer needs to detect if a state belongs to a “critical set”, i.e. a set of dangerous or unsafe states. As a second important contribution of the paper, we investigate under which conditions this is possible even if the system is under attack. An extension to the more general property of diagnosability is illustrated.}, archive = {J_TASE}, author = {Naeimeh Fakhr Shamloo and Elena De Santis and Maria Domenica Di Benedetto}, doi = {10.1109/TASE.2024.3416821}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5108--5116}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Security and diagnosability of finite state machines under cyber-attacks}, volume = {22}, year = {2025}, }
• (2025). MVoxTi-DNeRF: Explicit multi-scale voxel interpolation and temporal encoding network for efficient dynamic neural radiance field. TASE, 22, 5096--5107. (www)

  @article{ , abstract = {Neural radiance fields have revolutionized the field of novel view synthesis, achieving remarkable results. However, traditional approaches based on implicit representations, particularly those built upon NeRF, suffer from slow rendering speeds due to the need for numerous MLP evaluations. Recently, there has been a promising shift towards explicit representations using voxel grids, which has significantly improved reconstruction times for static scenes. Nonetheless, extending these methods from static scenes to dynamic scenes is a non-trivial task as it requires accounting for the changing geometry and appearance of the scene over time. In this paper, we propose an efficient dynamic neural radiance field with multi-scale explicit voxel interpolation and temporal encoding. We leverage an explicit voxel structure to store the 3D dynamic features, while employing a lightweight MLP to estimate the displacement, thereby significantly enhancing the reconstruction speed. In our canonical module, we incorporated temporal information encoding in density estimation and color estimation to rectify the error estimation of the displacement in deformation module. In addition, a multi-scale voxel interpolation is designed to accommodate large-scale motions while meticulously capturing intricate details in small-scale motions in the density estimation module. In experiment, we evaluate our MVoxTi-DNeRF method on both synthetic and real scenes, where it achieves superior or comparable rendering quality compared to state of the art methods, while remaining computationally efficient (more than $60\times $ faster than the original DNeRF). More experiment results and test code are available at https://github.com/CHenYYff/MVoxTi-DNeRF Note to Practitioners—Neural Radiance Fields(NeRF) can create highly detailed and realistic 3D models from 2D images via a neural network. It is particularly popular for its capacity to generate novel views of a scene, enabling it to synthesize images from previously unobserved viewpoints. NeRF has found applications in a wide range of fields, including virtual reality, augmented reality, gaming, and so on. In this study, we introduces an efficient dynamic NeRF method. First, our network leverages an optimized explicit voxel grid to store 3D dynamic features and employs a lightweight MLP to decode these deformation features, significantly accelerating the training process. Second, to correct the error estimation related to deformation displacement, we introduce encoding of temporal information into density and color estimation in our canonical module, which fortifies the canonical field’s perception of temporal information. Third, we utilize multi-scale voxel interpolation to capture different-scale motion in the density estimation module, where minor motions are modeled using nearby voxels, while motion within a broader range is captured through more distant voxels. The consideration of multi-scale voxel features diminishes the detrimental impact caused by inaccurate displacement estimation.}, archive = {J_TASE}, author = {Weiqing Yan and Yanshun Chen and Wujie Zhou and Runmin Cong}, doi = {10.1109/TASE.2024.3416533}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5096--5107}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MVoxTi-DNeRF: Explicit multi-scale voxel interpolation and temporal encoding network for efficient dynamic neural radiance field}, volume = {22}, year = {2025}, }
• (2025). Toward practical autonomous flight simulation for flapping wing biomimetic robots with experimental validation. TASE, 22, 5084--5095. (www)

  @article{ , abstract = {The utilization of a well-established flapping wing robot simulation holds significant importance in the advancement of flapping wing mechanisms and algorithms. This research paper introduces a pioneering application-oriented flapping wing robot simulation platform that exhibits high compatibility with diverse mechanical designs and adaptability to various robotic tasks. Initially, the paper presents the blade element theory and the quasi-steady model as computational approaches for determining the aerodynamics of flapping wings based on their kinematics. The computation incorporates translational lift, translational drag, rotational lift, added mass force, and clap-and-fling mechanism. The simulation platform is validated through flight control tasks, providing a comprehensive assessment of its performance. Moreover, this study addresses the challenges of attitude tracking and trajectory tracking control for a specially designed flapping wing robot. The proposed control strategies are evaluated through real flight experiments, offering practical insights into the robot flight capabilities. Note to Practitioners—One of the primary contributions of this research is the introduction of a novel and robust flapping wing flight controller that effectively addresses the challenges associated with attitude tracking control and positional trajectory tracking. The proposed controller demonstrates superior performance compared to existing algorithms, as evidenced by comparative simulations conducted on the developed simulation platform. Furthermore, real flight experiments are performed on a self-made flapping wing robot using the same control algorithm and parameters employed in the simulations. Another highlight lies in the transferability from simulation to reality. The utilization of a high-fidelity simulation platform enables in-depth exploration and scrutiny of complex behaviors manifested in diverse flight tasks. This, in turn, allows for meticulous analysis and examination to gain valuable insights into the practical implementation of flapping wing robot flight.}, archive = {J_TASE}, author = {Chen Qian and Yongchun Fang and Fan Jia and Jifu Yan and Yiming Liang and Tiefeng Li}, doi = {10.1109/TASE.2024.3415715}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5084--5095}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward practical autonomous flight simulation for flapping wing biomimetic robots with experimental validation}, volume = {22}, year = {2025}, }
• (2025). A physical Human–Robot interaction framework for trajectory adaptation based on human motion prediction and adaptive impedance control. TASE, 22, 5072--5083. (www)

  @article{ , abstract = {Physical human-robot interaction (pHRI) plays an important role in robotic. In order for a human operator to be able to easily adapt to interact with a robot, a minimal interaction force in pHRI should be achieved. In this paper, a pHRI framework is proposed to allow the robot to regulate its trajectory adaptively for minimizing the interaction force with small position-tracking errors. The trajectory of the robot is first adjusted by the interaction force which is updated by the performance evaluation index. Then, the human hand motion is predicted based on the autoregressive (AR) model to further adapt the trajectory. Thirdly, an adaptive impedance control method is developed to update the stiffness in the robot impedance controller using surface electromyography (sEMG) signals for robot compliant interaction with the environment. This method allows the human operator to interact with the robot by the interaction force, the hand motion and muscle contraction. By investigating the performance of the proposed method, the interaction force is decreased and a good position tracking accuracy is achieved. Comparative experiments demonstrate the enhanced performance of the proposed method. Note to Practitioners—This paper focuses on developing a novel method that can allow the robot to compliantly interact with the human operator while simultaneously taking into account the trajectory-tracking accuracy and the interaction force in pHRI scenarios. The proposed method has a large application potential in a variety of pHRI tasks, such as human-robot collaborative transporting, curing, assembly, cutting, and so on. In addition, the proposed method can allow the human operator to physically interact with the robot in an easier and more intuitive manner, by taking advantage of human motion prediction and adaptive impedance control. Therefore, it is also potentially utilized for rehabilitation and assistive robots, and robot learning skills from human physical demonstration.}, archive = {J_TASE}, author = {Jing Luo and Chaoyi Zhang and Weiyong Si and Yiming Jiang and Chenguang Yang and Chao Zeng}, doi = {10.1109/TASE.2024.3415650}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5072--5083}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A physical Human–Robot interaction framework for trajectory adaptation based on human motion prediction and adaptive impedance control}, volume = {22}, year = {2025}, }
• (2025). KG-planner: Knowledge-informed graph neural planning for collaborative manipulators. TASE, 22, 5061--5071. (www)

  @article{ , abstract = {This paper presents a novel knowledge-informed graph neural planner (KG-Planner) to address the challenge of efficiently planning collision-free motions for robots in high-dimensional spaces, considering both static and dynamic environments involving humans. Unlike traditional motion planners that struggle with finding a balance between efficiency and optimality, the KG-Planner takes a different approach. Instead of relying solely on a neural network or imitating the motions of an oracle planner, our KG-Planner integrates explicit physical knowledge from the workspace. The integration of knowledge has two key aspects: 1) We present an approach to design a graph that can comprehensively model the workspace’s compositional structure. The designed graph explicitly incorporates critical elements such as robot joints, obstacles, and their interconnections. This representation allows us to capture the intricate relationships between these elements; 2) We train a Graph Neural Network (GNN) that excels at generating nearly optimal robot motions. In particular, the GNN employs a layer-wise propagation rule to facilitate the exchange and update of information among workspace elements based on their connections. This propagation emphasizes the influence of these elements throughout the planning process. To validate the efficacy and efficiency of our KG-Planner, we conduct extensive experiments in both static and dynamic environments. These experiments include scenarios with and without human workers. The results of our approach are compared against existing methods, showcasing the superior performance of the KG-Planner. A short video introduction of this work is available via this link. Note to Practitioners—This paper was motivated by the problem of human-robot collaboratively working on remanufacturing processes such as disassembly that require human operators and collaborative robots to work closely with each other. The robots need to plan their trajectories efficiently enough to avoid collision with humans and the trajectories need to be short enough to reduce the cycle time. Traditional motion planners usually struggle with finding a balance between efficiency and optimality, which limits wide applications of collaborative robots in remanufacturing systems that are usually less structured than manufacturing systems. This paper suggests a new planning approach that integrates the workspace’s physical information into a graph and leverages deep learning to obtain safe and near-optimal solutions quickly. Experimental studies and observations demonstrated some advantages of this approach including learning capability, efficiency, and optimality, which makes it a great potential approach to be applied to real remanufacturing processes.}, archive = {J_TASE}, author = {Wansong Liu and Kareem Eltouny and Sibo Tian and Xiao Liang and Minghui Zheng}, doi = {10.1109/TASE.2024.3415497}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5061--5071}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {KG-planner: Knowledge-informed graph neural planning for collaborative manipulators}, volume = {22}, year = {2025}, }
• (2025). Quantum inference fuzzy rules network model free adaptive control for discontinuous derivative discrete-time systems. TASE, 22, 5051--5060. (www)

  @article{ , abstract = {The problem of discontinuous derivatives with respect to the control effort is observed in the prototype of preproduction DC motor torque-control. This behavior violates the basic requirement of conventional adaptive controllers. To overcome this issue, a quantum-inference fuzzy rules emulated network (QFREN) is developed, enabling the crossing state of discontinuity through a coherent superposition of its qbit membership functions. The adjustable parameters, which encompass both the linear parameters of QFREN and the nonlinear parameters of the rotation gate and quantum-controlled gate, are adjusted using the derived learning laws. Moreover, theoretical results are provided to ensure the convergence of the tracking error based on the selection of the designed variables in a practical context. Experimental validation is conducted to verify the effectiveness of the proposed controller. Additionally, comparative results with similar schemes are presented to highlight the advantages of the proposed approach. Note to Practitioners—The presence of discontinuous derivatives and nonlinearities, such as dead-zone, can compromise the closed-loop performance of control systems. Utilizing adaptive controllers based on quantum computation and neural networks (QNN) appears promising in mitigating these challenges. However, QNN approaches have predominantly remained in theoretical and numerical simulation contexts. In this work, an application of QFREN as a controller is introduced in a practical control engineering scenario. QFREN demonstrates its effectiveness in addressing the issue of discontinuous derivatives and compensating for unknown nonlinearities.}, archive = {J_TASE}, author = {C. Treesatayapun}, doi = {10.1109/TASE.2024.3415488}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5051--5060}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quantum inference fuzzy rules network model free adaptive control for discontinuous derivative discrete-time systems}, volume = {22}, year = {2025}, }
• (2025). Dynamic error-triggered adaptive control method and its industrial applications. TASE, 22, 5038--5050. (www)

  @article{ , abstract = {Industrial systems often undergo dynamic changes during operation, which presents challenges for traditional identification and control methods. These challenges arise in two aspects: variations in model structure and parameters, and differences in control objectives across diverse operating conditions. Traditional static predictive control methods face challenges in meeting the high-precision, real-time requirements in practice. In addition, control schemes with fixed parameters encounter difficulties in adapting to varying control objectives, resulting in suboptimal control performance. To address these problems, this article proposes a dynamic error-triggered adaptive control method, which can identify the operating conditions and objectives in real-time. Specifically, a dynamic error-triggered model updating mechanism is first established to detect changes in operating conditions and update the prediction model. To overcome the model mismatch during the transition process, a novel enhanced transition control (ETC) method is proposed, which designs a transition error corrector to decline prediction error and a high informative pseudo-random binary sequence (HIPRBS) input to enhance the excitation level. Considering the differences in control objectives under varying operating conditions, a fuzzy weight-adaptive method is proposed to balance heterogeneous indicators in different conditions. Two types of systems, high-speed and high-stability, are designed to validate the superiority of the proposed method. Extensive experimental results demonstrate that, compared to some state-of-the-art methods, the proposed method can efficiently and accurately identify emerging operating conditions, dynamically adjust optimization objectives, and achieve real-time control effects under varying operation conditions.Note to Practitioners—The motivation of this paper is to develop a high-precision and real-time control method for industrial systems that operate under frequently changing conditions. The proposed method can adapt to changes in model parameters and control objectives in multiple operating conditions processes. Compared with some state-of-the-art methods, this method significantly enhances the control performance in the transition process of mode switching, meeting the long-term stable operation requirements of industrial systems.}, archive = {J_TASE}, author = {Keke Huang and Pei Li and Dehao Wu and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2024.3415379}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5038--5050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic error-triggered adaptive control method and its industrial applications}, volume = {22}, year = {2025}, }
• (2025). Enhancing motion reconstruction from sparse tracking inputs with kinematic constraints. TASE, 22, 5029--5037. (www)

  @article{ , abstract = {In virtual reality, there is a growing demand for reconstructing accurate full-body 3D avatars from the sparse motion captured through head-mounted displays and hand-held controllers. However, due to the limited information from sparse inputs, precisely reconstructing full body poses is an ill-posed and challenging task. Existing methods often exhibit notable errors in lower body poses, which results in unrealistic poses and occasional floor penetration artifacts. To address the above issue, a MLP-based model with Kinematic Constraints and Temporal Diversity (KCTD) was proposed for full body poses reconstruction, which incorporates Kinematic Constraints Hierarchical Decoder with Temporal Diversity Awareness Module and a Generative Feedback Module to further improve the accuracy of the reconstruction of the full body poses. Specifically, the potential constraints of human kinematic chain are incorporated into the model through a hierarchical decoder, which elevates overall precision through the interaction of the human kinematic chain. Then, a temporal diversity awareness module is integrated into the hierarchical decoder to help the model capture information at different frequency in the time domain. In addition, a generative feedback module is imposed on leg poses reconstruction to further improve its accuracy without increasing the model’s inference time. Test results on the AMASS dataset demonstrate that, the proposed model effectively improves the reconstruction accuracy of the full-body poses with the mean per joint rotation error and position error of of 2.60 and 3.62 respectively, which surpasses the state-of-the-art methods. Particularly, the proposed model can alleviate irrational poses in the lower body and reduce the floor penetration artifacts. Note to Practitioners—The motivation behind this manuscript stems from the necessity for precise full pose reconstruction in certain virtual reality applications. While existing methods have significantly contributed to the reconstruction of full-body poses based on sparse inputs, they have not taken tailored measures to improve accuracy in the lower body. Consequently, this paper introduces a full human body pose reconstruction model grounded in constraints from the human kinematic chain, featuring a temporal diversity awareness module. This model excels in reconstructing more accurate complete human body poses using sparse inputs from the head and hands, notably addressing issues such as irrational lower-body poses and floor-penetrating artifacts. Benefiting from state-of-the-art precision, the proposed approach has the potential to significantly elevate user experience and immersion in virtual reality applications, particularly in scenarios such as automated task training and simulation of hazardous conditions.}, archive = {J_TASE}, author = {Xiaokun Dai and Xinkang Zhang and Shiman Li and Xinrong Chen}, doi = {10.1109/TASE.2024.3415151}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5029--5037}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Enhancing motion reconstruction from sparse tracking inputs with kinematic constraints}, volume = {22}, year = {2025}, }
• (2025). Solving a real-life stochastic car batching and sequencing problem with dynamic programming approaches. TASE, 22, 5012--5028. (www)

  @article{ , abstract = {This paper addresses a real-life stochastic car batching and sequencing problem in the body shop of a vehicle plant. Unlike previous research on similar problems in the paint shop or assembly shop, our problem primarily focuses on production planning in the body shop, with specific constraints on the production sequence (e.g., the bodies of each model cannot be produced individually but must be produced in batches of a specific size) and considers the impact of sampling inspection. Solving this large-scale car batching and sequencing problem in the body shop within an acceptable computation time is challenging. In this paper, several efficient dynamic programming-based algorithms are designed to solve the problem. First, a dynamic programming model is established for the deterministic version of the problem, and the optimal solution can be obtained by the dynamic programming approach. Furthermore, faced with the uncertainty introduced by sampling inspection, a more difficult and practical stochastic car batching and sequencing problem is modeled as a discrete-time Markov decision process. A rollout method-based approximate dynamic programming algorithm is designed to solve this complex problem. Finally, the proposed algorithms’ effectiveness is examined using real-life production data. Note to Practitioners—This paper is motivated by our collaboration with a vehicle production plant in Shanghai, China. The plant mainly produces $3\sim 4$ electrified models, with an annual output of up to 300,000 units/year. The plant consists of three workshops: the body shop, the paint shop, and the assembly shop. A multi-model flexible production line has been set up in the body shop. In practice, as multiple models are produced on the production line simultaneously, the plant adopts a mode of batch production to batch and sequence the bodies to be produced. Due to the specialty of the batch production mode and the uncertainty brought by the sampling inspection, the stochastic car batching and sequencing problem in the body shop presents a challenge to the plant. This paper proposes dynamic programming and approximate dynamic programming algorithms to effectively solve this stochastic optimization problem. The numerical experiments show the proposed algorithms can solve large-scale problems with near-optimal solutions within a reasonable time span. The algorithm can assist decision-makers in determining the batch size and the production sequence in automobile manufacturing enterprises.}, archive = {J_TASE}, author = {Ran Liu and Yue Zhang and Bo Zhou}, doi = {10.1109/TASE.2024.3415075}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5012--5028}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Solving a real-life stochastic car batching and sequencing problem with dynamic programming approaches}, volume = {22}, year = {2025}, }
• (2025). Efficient scheduling of residency-time-constrained dual-armed cluster tools with condition-based chamber cleaning operations. TASE, 22, 5000--5011. (www)

  @article{ , abstract = {Cleaning operations are required to clear particles and residual chemical gases in processing chambers of cluster tools in semiconductor manufacturing. To clean the chambers, condition-based chamber cleaning operations considering the actual state of chambers have been developed in practice. For single-armed cluster tools with such operations and wafer residency time constraints, the scheduling problem has been tackled. However, for dual-armed ones, this problem remains open and it is difficult to solve since their scheduling strategies are different from and more complex than single clusters’. To address this challenging problem, this work presents an efficient scheduling strategy to find the schedules that meet wafer residency time constraints and maximize their productivity. Two examples are given to show the effectiveness of the developed strategies. Note to Practitioners—Dual-armed cluster tools are extensively employed in semiconductor manufacturing processes. For residency-time-constrained dual-armed cluster tools with condition-based chamber cleaning operations, this work derives efficient scheduling strategies and algorithms for different cleaning cases. They can be used to realize feasible schedules that meet the wafer residency time constraints and restore a cluster tool to its steady-state as soon as possible. Hence, this work is significant and promising in helping semiconductor manufacturers enhance their fab’s throughput.}, archive = {J_TASE}, author = {Lu Zhen and BinQi Zhang and FaJun Yang and MengChu Zhou}, doi = {10.1109/TASE.2024.3414843}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {5000--5011}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient scheduling of residency-time-constrained dual-armed cluster tools with condition-based chamber cleaning operations}, volume = {22}, year = {2025}, }
• (2025). Robust LPV system identification with skewed and asymmetric measurement noise. TASE, 22, 4988--4999. (www)

  @article{ , abstract = {In this study, the skewed and asymmetric measurement noise is considered and solved in the identification of linear parameter varying (LPV) systems and a new robust global identification framework is established based on the shifted asymmetric Laplace (SAL) measurement distribution. The skewness and tails of the SAL distribution can be adaptively adjusted by the hyperparameters, that means the statistical property of the SAL distribution is governed by the hyperparameters which makes the SAL distribution flexible to resist various types of outliers including the skewed and asymmetric noise. The mathematical formulations of the identification problem are realized by the expectation maximization (EM) algorithm and the maximum likelihood estimates of the parameters are produced. It is realized that both the model parameters and the hyperparameters are extracted directly from the collected identification data. Compared with the existing robust methods, the advantages and disadvantages of the current work are revealed through the designed verification tests performed on the numerical example and the three-tank system, and the main results of this paper are also summarized.Note to Practitioners—The LPV system is widely applied in industrial processes due to its flexible capability of describing the complex nonlinear dynamics. For the probability-based identification of LPV systems, the Gaussian, Laplace and Student’s t distributions are commonly used to describe the output noise. But all of them exhibit symmetric statistical properties which may limit their applications in practical industrial settings, will them keep effective for the skewed and asymmetric measurement noise? Motivated by this question, this paper solves the robust identification of LPV systems with skewed and asymmetric measurement noise and a new robust global identification approach is introduced based on the SAL distribution. In this paper, it is proved that the common Laplace distribution can be seen as a special case of the SAL distribution. That means the proposed method is robust not only for the skewed and asymmetric measurement noise but also for the outliers, which could extend its applications in practical industrial processes. The tests performed on the numerical example and the three-tank system verify the proposed approach.}, archive = {J_TASE}, author = {Xin Liu and Yang Hai and Wei Dai}, doi = {10.1109/TASE.2024.3414500}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4988--4999}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust LPV system identification with skewed and asymmetric measurement noise}, volume = {22}, year = {2025}, }
• (2025). Optimizing multi-row layouts with the through-aisle structure: A hybrid approach of teaching-learning-based optimization and linear programming. TASE, 22, 4972--4987. (www)

  @article{ , abstract = {The multi-row facility layout problem is a prevalent and significant planning challenge in manufacturing workshops. This problem requires distributing facilities with pairwise transport weights among several rows to attain a layout with minimal logistics costs. However, the significance of aisles in multi-row facility layout has frequently been overlooked. An efficient aisle structure can result in a smooth transportation path and reduced material-handling costs. This paper contributes to the existing literature by introducing a new multi-row facility layout problem that considers long-straight aisles. First, mathematical formulas for the actual transportation distance between facilities through aisles are defined, and a mixed-integer programming model is constructed. Second, a hybrid algorithm based on an intelligent algorithm and a mathematical model is proposed. This method utilizes an improved teaching-learning-based optimization algorithm as a framework for optimizing the discrete facility sequence, and two decoding methods based on linear programming are designed to obtain the facility locations and transportation paths. Experimental results demonstrate that the two decoding strategies have their own advantages in terms of solution quality, efficiency, and area utilization. Moreover, improvement strategies for teaching-learning-based optimization algorithms are observed to be effective. Finally, we present two actual workshop examples of multi-row layout designs. The comparison of different algorithms reveals that the proposed algorithm has significant advantages in terms of solution quality and stability. Note to Practitioners—In multi-row layouts within manufacturing workshops, the strategic placement of facilities is crucial for reducing logistics costs. Essential factors such as the through-aisle structure and precise facility location are significant considerations. To more accurately reflect the actual material flow in such layouts, this paper proposes a new multi-row layout that takes into account the interaction of logistics aisles and the exact positioning of facilities. This problem is addressed by developing a mathematical model and proposing a novel hybrid method that combines teaching-learning-based optimization and linear programming to tackle real-life scenarios. The experimental findings suggest that the hybrid method is effective in generating superior solutions for the problem, and the optimization of the continuous and precise locations of facilities within the multi-row layout problem can significantly reduce material-handling costs. Our approach is shown to be feasible in accurately reflecting the logistics cost and layout scheme efficiency, thereby helping to achieve better production efficiency and cost reduction.}, archive = {J_TASE}, author = {Zeqiang Zhang and Yu Zhang and Feng Chen and Junqi Liu and Can Wang}, doi = {10.1109/TASE.2024.3414133}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4972--4987}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimizing multi-row layouts with the through-aisle structure: A hybrid approach of teaching-learning-based optimization and linear programming}, volume = {22}, year = {2025}, }
• (2025). Data-driven control with prescribed-time convergence for discrete-time nonlinear systems. TASE, 22, 4961--4971. (www)

  @article{ , abstract = {In this paper, the issue of data-driven prescribed-time convergence control for certain types of discrete-time nonlinear systems is investigated. The dynamic linearization technique and the discrete prescribed-time gain-scheduled feedback function have been combined to create a model-free adaptive prescribed time control strategy. To keep the steady-state error within a bespoke steady-state error band, the controller is switched to a steady-state controller once the system has completed its prescribed time control task. Theoretically, it is demonstrated that the error dynamics can gradually compress from the initial error to the steady-state error band during the initial and prescribed time periods. Furthermore, all of the rule designs in this study are data-driven and solely rely on system input and output data. Manipulator experiments and numerical modeling demonstrate the viability of the proposed strategy.Note to Practitioners—The motivation for this paper stems from the need to design a prescribed-time control method for the industrial robot trajectory tracking control tasks. Prescribed-time control method offers a potential to improve both the speed and coordination of industrial robot trajectory tracking. However, due to the increasing complexity of the robot structure and operating environment, it has brought great difficulties to the use of model-based prescribed-time control. To address this issue, a data-driven prescribed-time control approach is proposed, the controller design and convergence proof through mathematical derivations are also provided. Finally, the effectiveness of the proposed control scheme is validated through numerical simulations and industrial robotic manipulator experiments. This research contributes a viable and efficient time-optimal control method for trajectory-tracking tasks in the realm of industrial robotics.}, archive = {J_TASE}, author = {Yijia Li and Xianhua Ou and Peng Chen and Xiongxiong He and Qianru Jiang}, doi = {10.1109/TASE.2024.3413991}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4961--4971}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven control with prescribed-time convergence for discrete-time nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Artificial lateral line sensor for robotic fish speed measurement based on surface flow field detection and turbulence noise suppression. TASE, 22, 4947--4960. (www)

  @article{ , abstract = {Compared with traditional underwater vehicles, robotic fish have been receiving increasing attention in recent years due to their excellent maneuverability. However, the characteristics of fishlike undulatory motions and complex underwater working environment have posed significant challenges to robotic fish speed measurement, limiting their autonomy. To overcome these challenges, an artificial lateral line sensor (ALLS) was developed, drawing inspiration from the tactile system of fish. It captured the real-time speed of robotic fish through assessing the deformation of the stressed component under laminar flow impact. To mitigate turbulence disturbances near the ALLS, three flow control components, fairing, flow conditioner, and flow collector, were proposed to attenuate turbulence noise under the viscous effect. Furthermore, a physics-informed calibration method was presented to establish the nonlinear model of ALLS. Specifically, a physical model embedding algorithm based on data resampling was used to mitigate the risk of overfitting by the multilayer perceptron, considering the influence of turbulence disturbance and fishlike undulatory noise. Compared with the classical calibration method based on physical model fitting, the calibration method proposed in this paper reduced the error by 36.0%. Our ALLS’s final mean absolute error was 0.016 m/s with a linearity (R2) of 0.956. The experimental results indicated that the significant changes in the motion state of robotic fish reduced the accuracy of ALLS. The fusion with other sensors is expected to enhance the robustness of ALLS in the future. Note to Practitioners—The motivation of this paper is to design an artificial lateral line sensor based on surface flow field detection and turbulence noise suppression, providing a small-sized and high-precision solution to the speed measurement problem of bionic robotic fish. Most existing ALLS research focused on developing new types of sensors based on different measurement principles, without suppressing the noise caused by fishlike motions, and most experiments were conducted in environments with excessive controls rather than free-swimming robotic fish. To this end, we developed an ALLS based on deformation measurement and proposed three flow control components to make the measured flow more stable. Furthermore, a physics-informed overfitting suppression method was used for the calibration task of the ALLS. A series of simulations and experiments demonstrated that the proposed turbulence noise suppression and calibration method were practical and effective. Hopefully, our methods can provide theoretical and technical guidance to marine engineers for underwater vehicle speed measurement and flow sensing. The recommended flow control component is applicable for conditioning surface fluids in pneumatic control systems. Furthermore, the proposed biomimetic tactile sensor is poised to inspire tactile-based human-machine interaction methods.}, archive = {J_TASE}, author = {Zhuoliang Zhang and Chao Zhou and Long Cheng and Junfeng Fan and Min Tan}, doi = {10.1109/TASE.2024.3413776}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4947--4960}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Artificial lateral line sensor for robotic fish speed measurement based on surface flow field detection and turbulence noise suppression}, volume = {22}, year = {2025}, }
• (2025). Fast finite time stability in probability of p-norm stochastic nonlinear systems with mismatched uncertainties and dead-zone. TASE, 22, 4937--4946. (www)

  @article{ , abstract = {This brief describes a fast finite time adaptive control strategy for p-norm stochastic nonlinear systems (SNSs) using adding a power integrator approach. In terms of convergence speed, finite time control (FTC) offers a substantial advantage in the vicinity of the equilibrium point, but it may exhibit notably slower convergence compared to exponential convergence when the initial state is far from the origin. To overcome this problem, the Lyapunov function is skillfully constructed in this study. A method is devised based on this function, which incorporates the characteristic of adding a power integrator technique and the symbolic function to effectively tackle challenges posed by complex system structures. Then, adaptive control is utilized to deal with the mismatched uncertain nonlinear function. Subsequently, a novel fast FTC strategy is proposed for p-norm SNSs with dead-zone via the back-stepping framework, which ensures the transient performance of the closed-loop system. In comparison to existing results, this controller effectively achieves performance control for p-norm SNSs with dead-zone and mismatched uncertainties functions. Finally, the superiority of the scheme is illustrated by comparative simulations. Note to Practitioners—The FTC problem is a prominent subject in the field of control, playing a crucial role in practical applications. This is especially notable in p-norm nonlinear systems, where the dynamic behavior exhibits uncontrollable linearization characteristics near the origin, introducing complexity to control analysis. Another challenging aspect in nonlinear control is the impact from external disturbances. In practical systems, the presence of random disturbances is inevitable and frequently results in system instability. A fundamental technical barrier is the involvement of Brownian motion integral terms in stochastic nonlinear systems, along with Itô differential, which introduces not only gradients in Lyapunov analysis but also the Hessian of the Lyapunov function. Furthermore, it is observed that mismatched uncertainties and dead-zone phenomena exist in real systems. Designing fast finite-time controllers has become a focal point of controller research. To address these issues, this paper proposes a fast FTC algorithm tailored for a class of p-norm SNSs with mismatched uncertainties. The proposed approach ensures that the system state can converge to the desired region near the origin in an almost fast finite time.}, archive = {J_TASE}, author = {Yixuan Yuan and Liping Xie and Junsheng Zhao and Kanjian Zhang and Xiangpeng Xie}, doi = {10.1109/TASE.2024.3412734}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4937--4946}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast finite time stability in probability of p-norm stochastic nonlinear systems with mismatched uncertainties and dead-zone}, volume = {22}, year = {2025}, }
• (2025). Reagent addition control for zinc first rougher with a dual FP tree-based feature setpoint generator and knowledge core-based reagent fine presetting. TASE, 22, 4926--4936. (www)

  @article{ , abstract = {Reagent addition control performance of zinc first rougher significantly affects the quality of the final product. It is common to control visual features to satisfactory feature setpoints, in order to realize satisfactory reagent addition control. However, existing methods usually use the historical data under the desired production state to generate the feature setpoint, and the historical data under the undesired production state are ignored. Moreover, the number of feed grade categories for feature setpoint generation is settled, and thus the richness of corresponding query rules for presetting the reagent addition rate is limited. In this situation, a reagent addition control strategy with a dual frequent pattern tree (dual FP tree)-based feature setpoint generator and knowledge core-based reagent fine presetting is proposed for zinc first rougher. First, a dual FP tree-based feature setpoint generator with the complementary input is constructed to fully utilize the dataset, where the main tree is designed to construct the knowledge repository and the secondary tree is established for optimization. Second, the optimized knowledge rules are taken as cores to construct an optimal operational pattern base for reagent fine presetting. Then, a designed feedback controller based on a fuzzy logic system with a three-way decision mechanism is explored to adjust reagent addition rates and try to make visual features track prescribed feature setpoints. The ablation results show effectiveness of the dual FP tree-based feature setpoint generator, reagent fine presetting and feedback controller. The comparative experimental results demonstrate the potential of the proposed control strategy. Note to Practitioners—This article is motivated by the problem of reagent addition control in the zinc flotation process. This article focuses on making visual features track prescribed feature setpoints by adjusting reagent addition rates, aiming to achieve the goal of controlling the zinc concentrate grade within its acceptable range. However, two challenges exist for control: the historical data under the desired production state are usually used during feature setpoint generation, while latent process knowledge in the historical data under the undesired production state is ignored; the number of feed grade categories for feature setpoint generation is settled during reagent presetting, which means the richness of corresponding query rules for presetting the reagent addition rate is limited. Therefore, this article suggests a reagent addition control strategy with a dual FP tree-based feature setpoint generator and knowledge core-based reagent fine presetting to address those problems. A suitable feedback controller is designed for outputting reagent addition adjustment values to qualify the concentrate grade, which is based on feature setpoint generation and reagent fine presetting. The experiments are carried out for evaluation. The evaluation shows the potential and superiority of the proposed approach in terms of decision-making and automation control in froth flotation.}, archive = {J_TASE}, author = {Xiaoliang Gao and Zhaohui Tang and Yongfang Xie and Hu Zhang and Nongzhang Ding and Weihua Gui}, doi = {10.1109/TASE.2024.3412682}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4926--4936}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Reagent addition control for zinc first rougher with a dual FP tree-based feature setpoint generator and knowledge core-based reagent fine presetting}, volume = {22}, year = {2025}, }
• (2025). Probabilistic data-driven modeling of a melt pool in laser powder bed fusion additive manufacturing. TASE, 22, 4908--4925. (www)

  @article{ , abstract = {The widespread adoption of laser powder bed fusion (LPBF) additive manufacturing is hampered by process unreliability problems. Modeling the melt pool behavior in LPBF is crucial to develop process control methods. While data-driven models linking melt pool dynamics to specific process parameters have shown appreciable advancements, existing models often oversimplify these relationships as deterministic, failing to account for the inherent instability of LPBF processes. Such simplifications can lead to overconfident and unreliable predictions, potentially resulting in erroneous process decisions. To address this critical issue, we propose a probabilistic data-driven approach to melt pool modeling that incorporates process noise and uncertainty. Our framework formulates a problem that includes distribution approximation and uncertainty quantification. Specifically, the Gaussian distribution with higher order priors, aided with variational inference and importance sampling, is used to approximate the probability distribution of melt pool characteristics. The uncertainty inherent in both LPBF process data and the modeling approach itself are then decomposed and approximated by using Monte Carlo sampling. The melt pool model is improved further by using a novel grid-based representation for the neighborhood of a fusion point, and a neural network architecture designed for effective feature fusion. This approach not only refines the accuracy of the model but also quantifies the uncertainty of the predictions, thereby enabling more informed decision-making with reduced risk. Two potential applications, including LPBF process planning and anomaly detection, are discussed. The implementation of our model is available at https://github.com/qihangGH/probabilistic_melt_pool_model. Note to Practitioners—Modeling the melt pool behavior in laser powder bed fusion (LPBF) processes is pivotal for enhancing its quality control. However, a problem is that most existing data-driven melt pool models learn melt pool behavior with a deterministic function, which predicts the same outputs if its inputs are the same. This deviates from the reality and neglects the uncertainty in LPBF processes. As a consequence, the quality control methods based on such melt pool models lack required reliability. In response to these challenges, this work proposes to model melt pool behavior by using probability distributions with deep learning techniques, which can quantify the uncertainty in both LPBF process data and data-driven models. Aided with an elegantly designed representation for the neighborhood of a fusion point as model input, and a neural network architecture that fuses multi-modal data, the proposed model achieves accurate melt pool size prediction results. More importantly, this work quantifies and decomposes the prediction uncertainty. By accounting for noise and parameter variations, the probabilistic modeling models developed herein offer a more robust foundation for LPBF quality control than the existing ones. They can be readily applied by practitioners to perform improved process planning, defect prognosis, and real-time anomaly detection tasks.}, archive = {J_TASE}, author = {Qihang Fang and Gang Xiong and Meihua Zhao and Tariku Sinshaw Tamir and Zhen Shen and Chao-Bo Yan and Fei-Yue Wang}, doi = {10.1109/TASE.2024.3412431}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4908--4925}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Probabilistic data-driven modeling of a melt pool in laser powder bed fusion additive manufacturing}, volume = {22}, year = {2025}, }
• (2025). Activation control of multiple piecewise linear neural networks. TASE, 22, 4895--4907. (www)

  @article{ , abstract = {Piecewise linear neural networks (PLNNs) are proven universal approximators for continuous functions on the compact domain. For multiple PLNNs (mPLNNs) differing from each other in suffering different approximation errors (AEs) when approximating the same continuous function, activating all the PLNNs to approximate the continuous function, and then, picking up the minimum AE (MAE) from all the AEs seems to be a practical way to arrive at such MAE. Activating PLNNs has to consume energy, and more activated PLNNs provide more AEs to consider, which can maximize the probability of guaranteeing the MAE, while also needing to harvest more energy. Therefore, how to make the optimal tradeoff between energy harvested and approximation accuracy for mPLNNs arises as an interesting issue. To address this problem, this paper first deduces the objective function, with the accumulative probability of obtaining the MAE as the objective and the accumulative energy harvested as the constraint, then reveals the optimal probability that each hidden neuron (HN) in the activated PLNN should be activated, and finally uncovers the optimal activation probability for each PLNN. An algorithm based on our discovered theoretical results is proposed for mPLNNs to enjoy the maximum probability of achieving the MAE at the acceptable level of accumulative energy harvested. Theoretical analysis and experiments verify its performance. Note to Practitioners—This paper addresses the interesting issue of how to make mPLNNs approximate the continuous function with the highest approximation accuracy under the energy constraint. Through the insightful discovery of the optimal activation probability for each HN in the activated PLNN as well as the optimal activation probability for each PLNN, this paper facilitates mPLNNs to suffer the MAE in their approximations to continuous functions, with the maximum probability while maintaining the energy harvested within a given range. Because PLNNs can model the nonlinear and complex system with the arbitrary AE, our proposed approach can be effectively applied by such system to operate in an energy-efficient and universal-approximation manner, which we believe could push the development of the low-power multiple neural networks.}, archive = {J_TASE}, author = {Chen Hou}, doi = {10.1109/TASE.2024.3412239}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4895--4907}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Activation control of multiple piecewise linear neural networks}, volume = {22}, year = {2025}, }
• (2025). Energy-efficient control of multiple smart sensors with uncertain controllability. TASE, 22, 4881--4894. (www)

  @article{ , abstract = {For multiple smart sensors (MSSs) operating with limited energy, a smart sensor is said to be controllable if there exists proper external signal (ES) enabling it to report the sensing data, reflecting the physical world with tolerable sensing errors. In order to improve energy-efficiency while meeting the required sensing accuracy, the controllable smart sensors are expected to be activated while the uncontrollable ones are not. However, due to the exterior noises or interior interferences (ENII) of MSSs, whether a smart sensor is or not controllable is often uncertain, which may wrongly activate uncontrollable sensors or deactivate controllable ones, not only wasting energy but also deteriorating sensing errors. Therefore, addressing the problem of assuring MSSs with uncertain controllability to report tolerable sensing errors at the minimum cost of energy emerges as an interesting and challenging issue. To address this problem, this paper first formulates MSSs as the probabilistic multiple input multiple output (PMIMO) system, with all the ESs and sensing errors as the input and output, respectively. It then discloses the optimal activation probability for each smart sensor and reveals the optimal ES corresponding to the minimum energy for MSSs experiencing uncertain controllability to report the desired sensing errors. An algorithm is proposed based on our discovered results to implement such ES. The theoretical analysis and field experiments verify its performance. Note to Practitioners—This paper addresses the challenging issue of how to guarantee MSSs experiencing uncertain controllability to report tolerable sensing errors at the minimum cost of energy. Though the insightful discovery of the optimal probability that each smart sensor is activated, and further the optimal ES corresponding to the minimum energy for enabling MSSs experiencing uncertain controllability to report the desired sensing errors, this paper facilitates MSSs with uncertain controllability to operate in an energy-efficient as well as high-accurate manner, which we believe could improve the power usage and strengthen the sensing capacity of the Internet of Things (IoT) or Cyber-Physical Systems (CPS) that depend on MSSs settings to sense the physical world. Theoretical analysis and experiments both verify that the proposed solution outperforms existing ones.}, archive = {J_TASE}, author = {Chen Hou}, doi = {10.1109/TASE.2024.3412229}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4881--4894}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Energy-efficient control of multiple smart sensors with uncertain controllability}, volume = {22}, year = {2025}, }
• (2025). Optimal robust formation of multi-agent systems as adversarial graphical apprentice games with inverse reinforcement learning. TASE, 22, 4867--4880. (www)

  @article{ , abstract = {This paper introduces a novel approach to solving robust optimal formation control problems in heterogeneous multi-agent systems (MASs) with disturbances. Our approach frames this problem as an adversarial graphical apprentice game problem, using the game concept and an inverse reinforcement learning (IRL) algorithm. This work innovates the derivation of unknown reward functions through demonstrations to achieve the desired formation. In addition, we define learner, estimator, and expert MASs as separate entities. This aim is achieved by introducing a model-based IRL algorithm with three stages of learning. After determining the optimal control of estimators in the first stage, the optimal control of learners is found in the second stage by using them. In the third stage, the reward functions of the estimator and learner MASs are updated in the inverse optimal control (IOC) stage simultaneously. Following that, we present an algorithm for model-free IRL that does not require knowledge of learner agents’ dynamics and can reconstruct reward functions after observing online trajectories. As subproblems, both proposed IRL algorithms address optimal control and IOC. We analyze these approaches for stability and convergence and demonstrate that state reward weights are non-unique. Simulation results show the effectiveness of the introduced algorithms for a group of unmanned aerial vehicles (UAVs). Note to Practitioners—Designing an optimal controller is challenging in MAS, especially where we only have access to the optimal reference path. The challenge is compounded if we do not know the agents’ dynamics and the appropriate reward function needed to achieve the desired path. This article proposes two IRL algorithms with the motivation to solve this problem. This study considers the optimal reference path for agents as expert MAS. Also, the MAS, we want to follow the desired path with the minor error, is represented as a learner MAS. The proposed approach is suitable for a large class of linear MAS since we have modeled the learner MAS with disturbance. Therefore, this method suits tracking, path-planning, and optimal formation problems in MASs. It requires minimal information but yields impressive results.}, archive = {J_TASE}, author = {Fatemeh Mahdavi Golmisheh and Saeed Shamaghdari}, doi = {10.1109/TASE.2024.3412188}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4867--4880}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal robust formation of multi-agent systems as adversarial graphical apprentice games with inverse reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). A hierarchical MPC for end-effector tracking control of legged mobile manipulators. TASE, 22, 4855--4866. (www)

  @article{ , abstract = {This article presents a hierarchical model predictive control (MPC) framework for the end-effector tracking control problem of a legged mobile manipulator. In the high-level part, a kinematic MPC over a long-time horizon computes both base and joint trajectories, then a quadratic program (QP) based optimization solves ground-reaction-forces (GRFs) satisfying the robot’s centroidal dynamics and the friction cone constraints. In the low-level part, a kinodynamic MPC over a short-time horizon tracks command end-effector trajectories and outputs from the high-level part while satisfying all nonlinear dynamics constraints. Due to the complexity of MPC formulations and high real-time requirements, traditional MPC for legged mobile manipulators can only generate short-time horizon solutions for tracking tasks over longer time horizons, which may lead to the optimization falling into bad local minima. In our method, the long-term trajectories from the high-level part can guide the optimization of the short-term kinodynamic MPC to generate a better solution. We validate the effectiveness of our method through several simulation and hardware experiments. In comparison to traditional MPC, the proposed method improves the trajectory tracking accuracy of the robot’s end-effector while reducing the violations of the system’s physical limit constraints and environment-collision avoidance constraints.Note to Practitioners—Legged mobile manipulators have received increasing research attention in recent years due to their potential applications. In this paper, we particularly focus on the high-efficiency end-effector tracking control of legged mobile manipulators. We propose a hierarchical MPC framework incorporating a high-level part generating long-term whole-body motions and a low-level kinodynamic MPC replanning at a high frequency. Simulation and hardware experiments show that the proposed method improves the trajectory tracking accuracy while reducing the violations of the system’s physical limit constraints and environment-collision avoidance constraints in comparison to traditional MPC. In future research, we will address the end-effector tracking control problem when the end-effector contacts the environment.}, archive = {J_TASE}, author = {Dongqi Wang and Jiyu Yu and Shuangpeng Wu and Zhang Li and Chao Li and Rong Xiong and Shaoxing Qu and Yue Wang}, doi = {10.1109/TASE.2024.3412111}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4855--4866}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical MPC for end-effector tracking control of legged mobile manipulators}, volume = {22}, year = {2025}, }
• (2025). Composite disturbance filtering for onboard UWB-based relative localization of tiny UAVs in unknown confined spaces. TASE, 22, 4840--4854. (www)

  @article{ , abstract = {Due to its small size and light weight, the tiny unmanned aerial vehicle (UAV) is especially suitable for tasks that involve confined indoor space exploration. However, autonomous localization of the tiny UAV has become a great challenge in the global navigation satellite system (GNSS)-denied, unstructured, and resource-constrained environments. In this paper, an onboard ultra-wideband (UWB)-based relative localization scheme of the tiny UAV is presented, where an anchor-UAV equipped with the UWB anchors is employed to provide position reference for the tiny UAV. As the key in the proposed scheme, the state estimation algorithm should be able to overcome the joint effect of the inertial sensor bias, the UWB measurement noises, and the limited computational resources. To this end, a composite disturbance filtering (CDF) method is proposed which consists of the disturbance observer for real-time bias compensation, an improved particle filter to deal with the non-Gaussian noises, and a particle size adaptation procedure to release the computation burden. The proposed CDF method represents a refined treatment of the multi-source heterogeneous uncertainties in the relative localization system and strikes a balance between localization accuracy and computational efficiency. The effectiveness of the proposed localization method is validated via both simulation tests and flight experiments.Note to Practitioners—The tiny UAVs are capable of dexterous flight in confined indoor spaces, which makes them especially suitable for the indoor operation tasks such as environment exploration, disaster relief, gas seeking, and meteorological measurement. In these tasks, accurate localization of the tiny UAV is of vital importance. To achieve autonomous localization of the tiny UAV in the GNSS-denied, infrastructure-free, and unstructured indoor environment, an onboard inertial measurement unit (IMU)/UWB fusion-based relative localization scheme is presented. In the presented scheme, a large UAV equipped with UWB anchors is employed to provide position reference for the tiny UAV. To address the multi-source heterogeneous disturbances (the dynamic IMU biases and the non-Gaussian UWB noises) in the localization model and the limited computational resource onboard the tiny UAV, a CDF method with an adaptive sample size is proposed. As demonstrated by the simulation and experiments, the positioning accuracy and computation efficiency of the tiny UAV have been effectively enhanced with the proposed scheme.}, archive = {J_TASE}, author = {Jingting Jia and Bo Tian and Wenshuo Li and Dadong Fan and Kexin Guo and Xiang Yu and Lei Guo}, doi = {10.1109/TASE.2024.3412068}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4840--4854}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Composite disturbance filtering for onboard UWB-based relative localization of tiny UAVs in unknown confined spaces}, volume = {22}, year = {2025}, }
• (2025). Distributed privacy-preserving estimation for multi-area power systems with multichannel event-triggered mechanisms. TASE, 22, 4831--4839. (www)

  @article{ , abstract = {In power systems with multi-area power exchanges, distributed state estimation not only depends on sensor measurements but also requires area-interaction. Plentiful transmitted information puts pressure on data privacy and communication resources. To preserve the privacy of sensor and interaction data, a switching encryption, whose minimum dwell-time is less than the decryption time of eavesdroppers, is proposed. Based on encrypted data, multichannel event-triggered mechanisms are established to save the communication resources of sensor-to-estimator and area-interaction channels, simultaneously. By designing the appropriate estimator gains and encryption-compensating mechanisms, the asymptotic stability of estimation error dynamics is guaranteed. Compared with the existing privacy-preserving technologies, the switching encryption improves the estimation accuracy of power system states while avoiding the decryption of eavesdroppers. The effectiveness of distributed privacy-preserving estimation is illustrated by a three-area power system. Note to Practitioners—Distributed estimation of multi-area power systems produces lots of communication data, which challenges data privacy and communication resources. Since the estimation of power system states generally requires high accuracy, the existing privacy-preserving technologies, which sacrifice estimation accuracy, are inappropriate for power systems. Therefore, a switching encryption scheme combined with the encryption-compensating mechanisms is proposed to preserve data privacy while the asymptotic stability of estimation error dynamics is guaranteed. Furthermore, the simultaneous transmissions of sensor measurement and area-interaction data put pressure on the communication resource of distributed estimation. Therefore, the multichannel event-triggered mechanisms are proposed to reduce the transmission frequencies of encrypted data.}, archive = {J_TASE}, author = {Tian-Yu Zhang and Guang-Hong Yang and Dan Ye}, doi = {10.1109/TASE.2024.3411803}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4831--4839}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed privacy-preserving estimation for multi-area power systems with multichannel event-triggered mechanisms}, volume = {22}, year = {2025}, }
• (2025). Class-aware cartilage segmentation for autonomous US-CT registration in robotic intercostal ultrasound imaging. TASE, 22, 4818--4830. (www)

  @article{ , abstract = {Ultrasound imaging has been widely used in clinical examinations owing to the advantages of being portable, real-time, and radiation-free. Considering the potential of extensive deployment of autonomous examination systems in hospitals, robotic US imaging has attracted increased attention. However, due to the inter-patient variations, it is still challenging to have an optimal path for each patient, particularly for thoracic applications with limited acoustic windows, e.g., intercostal liver imaging. To address this problem, a class-aware cartilage bone segmentation network with geometry-constraint post-processing is presented to capture patient-specific rib skeletons. Then, a dense skeleton graph-based non-rigid registration is presented to map the intercostal scanning path from a generic template to individual patients. By explicitly considering the high-acoustic impedance bone structures, the transferred scanning path can be precisely located in the intercostal space, enhancing the visibility of internal organs by reducing the acoustic shadow. To evaluate the proposed approach, the final path mapping performance is validated on five distinct CTs and two volunteer US data, resulting in ten pairs of CT-US combinations. Results demonstrate that the proposed graph-based registration method can robustly and precisely map the path from CT template to individual patients (Euclidean error: $2.21\pm 1.11~mm$ ). Note to Practitioners—The precise mapping of trajectories has been a bottleneck in developing autonomous intercostal intervention within limited acoustic space. Existing methods, based on external features such as the skin surface or passive markers, fail to capture the acoustic properties of local tissues, leading to significant shadowing when ribs are involved. The proposed method begins by utilizing distinctive anatomical features to extract cartilage bones and stiff ribs through a class-aware segmentation network. To ensure the segmentation accuracy of the shape of the anatomy of interest, a VAE-based boundary-constraint post-processing in manifold space is developed. Subsequently, a dense skeleton graph-based registration is developed to explicitly consider the subcutaneous bone structure, allowing for the precise mapping of intercostal paths from generic templates to individual patients. Results from ten randomly paired CT and US datasets show that the proposed method accurately maps the intercostal path from the template to individual patients, significantly improving accuracy and robustness over previous methods. We believe that the proposed method can further pave the way for autonomous robotic US imaging.}, archive = {J_TASE}, author = {Zhongliang Jiang and Yunfeng Kang and Yuan Bi and Xuesong Li and Chenyang Li and Nassir Navab}, doi = {10.1109/TASE.2024.3411784}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4818--4830}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Class-aware cartilage segmentation for autonomous US-CT registration in robotic intercostal ultrasound imaging}, volume = {22}, year = {2025}, }
• (2025). Online monitoring of heterogeneous partially observable data streams based on Q-learning. TASE, 22, 4802--4817. (www)

  @article{ , abstract = {With the rapid advances in Internet of Things (IoT) technology and computational infrastructure, heterogeneous data streams are becoming common in various manufacturing applications. Meanwhile, the resource constraints often restrict the full observability of data streams due to limited budget for deploying or turning on every sensor at a time, as well as limited transmission and processing time for collecting high frequency information from all data streams, which poses significant challenges for multivariate statistical process control (SPC) and quality improvement. In this article, diverging from conventional heuristic approaches, we propose a new algorithm based on Q-learning to online monitor and quickly detect mean shifts occurring to heterogeneous data streams in the context of limited resources, where only a subset of observations is available at each acquisition time. In particular, we integrate Q-learning with a nonparametric cumulative sum (CUSUM) procedure to effectively detect a wide range of possible mean shifts when data streams follow arbitrary distributions. Both simulations and a case study are thoroughly conducted to evaluate the performance and demonstrate the superiority of the proposed method.Note to Practitioners—This paper is motivated by the practical issue of online process monitoring and anomaly detection with resource limitations. In particular, we can only select a subset of data streams to monitor at each time epoch, and the challenges are to dynamically choose which ones to observe and when to raise an alarm. Unlike the existing methodologies which are heuristic and only consider short-term rewards from the dynamic sampling, this paper proposes a novel monitoring and sampling strategy that allows the practitioners to cost-effectively monitor heterogeneous data streams by considering the long-term rewards. Three main steps are involved in the proposed method: (i) construct a local nonparametric monitoring statistic for each data stream; (ii) train the proposed reinforcement learning framework; and (iii) at each time epoch, determine the most informative data stream to observe according to the Q-table and decide whether to raise the alarm. Experimental results through simulations and a case study have shown that the proposed method has better performance than the existing methods in reducing detection delay.}, archive = {J_TASE}, author = {Haoqian Li and Honghan Ye and Jing-Ru C. Cheng and Kaibo Liu}, doi = {10.1109/TASE.2024.3411770}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4802--4817}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online monitoring of heterogeneous partially observable data streams based on Q-learning}, volume = {22}, year = {2025}, }
• (2025). A probabilistic quality-relevant monitoring method with gaussian mixture model. TASE, 22, 4790--4801. (www)

  @article{ , abstract = {Process uncertainty, which is usually caused by various factors, is generally subject to unknown complex distribution. However, many existing monitoring methods are established with a single distribution, and thus they may not accurately reflect the uncertainty within process systems. In this study, a probabilistic quality- relevant monitoring (PQM-GMM) is proposed with the Gaussian mixture model to address the aforementioned issue. Different from conventional monitoring methods, the proposed method measures the process uncertainty using multiple Gaussian distributions, which can be used to approximate any unknown complex distribution. Then, the optimization problem of the proposed PQM-GMM model is solved using the expectation maximization (EM) algorithm, which includes an augmented Lagrange multiplier in the M-step for model parameter estimation. Using the obtained results, a quality-relevant monitoring model is established with three statistics. It is noted that the proposed model can also be extended to many existing methods since they share a similar structure. Besides, the detailed information such as initial value selection, missing data problem, computation complexity is discussed. The effectiveness and superiority of the proposed method are tested using a numerical simulation example and a real low-pressure heater application. In comparison with some commonly used quality-relevant methods, the proposed model can be robustly established in the presence of corrupted data, and has a better detection sensitivity for the process anomalies in both process and quality variables. Note to Practitioners—A quality-relevant monitoring method is proposed in this study with Gaussian mixture model (GMM) for detecting the abnormal conditions of industrial processes under harsh environment. Since GMM can be used to approximate any unknown complex distribution, the process uncertainty within the collected data can be meticulously measured using the proposed PQM-GMM model. Besides, the quality-independent faults and quality-related faults can also be effectively distinguished using the designed monitoring statistics.}, archive = {J_TASE}, author = {Wanke Yu and Chunhui Zhao and Biao Huang and Hui Yang}, doi = {10.1109/TASE.2024.3411471}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4790--4801}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A probabilistic quality-relevant monitoring method with gaussian mixture model}, volume = {22}, year = {2025}, }
• (2025). Development of deep reinforcement learning co-simulation platforms for power system control. TASE, 22, 4780--4789. (www)

  @article{ , abstract = {This paper introduces four co-simulation platforms for testing deep reinforcement learning (DRL)-based control solutions in power systems. The first one is to connect the off-the-shelf Matlab DRL toolbox with the developed Matlab Simulink power system models (platform 1). The second is to utilize the Python-C interface to integrate the algorithm design in Python and the model development in Matlab, wherein the developed model is first generated to C-code and then compiled into a shared library (platform 2). The third and fourth platforms are based on the real-time simulator, whereas these two employ two different communication protocols, i.e., TCP/IP (platform 3) and EtherCAT (platform 4). The Opal-RT real-time simulators utilized by platforms 3 and 4 ensure that both platforms can run the models in a real-time manner. Specifically, platform 4 possesses capabilities for conducting real-time DRL training and control owing to the real-time feature of both communication and simulation. The detailed procedures regarding implementing these different DRL co-simulation platforms are provided in this paper, with the pros and cons of each one commented on, which can help researchers speed up the preliminary design of DRL-based control solutions for dynamic power systems. Note to Practitioners—This article is motivated by the emerging deep reinforcement learning-based control solutions for dynamic power systems. There is an urgent need for co-simulation platforms that enable a high-fidelity and real-time simulation of power systems and a thorough evaluation of various DRL control algorithms. Since most existing DRL testing tools are developed using Python with simplification adapted to large time-scales or specific functionalities, limiting their applicability in small-time-scale online control. The model with higher granularity is necessary for small-time-scale online control problems for power system applications that involve fast dynamics. The developed platforms allow for direct integration of the dynamic Matlab Simulink power system model and run complex dynamic models in a real-time manner via the Opal-RT simulator. A classic optimal generation control problem for a test power system is studied in this paper to validate the effectiveness of the developed online and real-time DRL platforms. For other power system applications, one only needs to follow the procedures provided in this paper to customize the corresponding dynamic models or environments. The comprehensive real-time and online tests of different DRL control algorithms can be accomplished on the developed platforms with slight effort.}, archive = {J_TASE}, author = {Zhenghong Tu and Zhen Fan and Wei Zhang and Wenxin Liu}, doi = {10.1109/TASE.2024.3411305}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4780--4789}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development of deep reinforcement learning co-simulation platforms for power system control}, volume = {22}, year = {2025}, }
• (2025). Observer-based sampled-data adaptive tracking control for heterogeneous nonlinear multi-agent systems under denial-of-service attacks. TASE, 22, 4771--4779. (www)

  @article{ , abstract = {This paper tackles the security cooperative tracking control problem for uncertain heterogeneous nonlinear multi-agent systems (MASs) subject to denial-of-service (DoS) attacks. Since the communication between agents can only transmit digital signals, a periodic sampling mechanism is used to reduce bandwidth pressure. Based on these digital signals, a novel distributed switching observer is designed to estimate the state of the reference system for each follower in the case of DoS attacks on aperiodic intermittently blocked communication channels. Then, an attack-parameter-dependent Lyapunov functional is established to analyze the exponential convergence of the error signals and a design scheme of observer gains is proposed. Further, with the help of the sampling observer outputs, follower’s states and adaptive signals, a sampled-data neural network controller is designed to guarantee the boundedness of the tracking errors by backstepping method and impulsive system analysis theory. Finally, a simulation example is performed to showcase the effectiveness of the developed strategy. Note to Practitioners—In industrial systems, since it is difficult for a single plant to complete complex control tasks, the cooperative control of MASs has gradually become a hot research issue. For example, in industrial automated production, a single robotic arm cannot achieve mass production of products, while multiple robotic arms can achieve this task and improve production efficiency. In this context, the study on cooperative control of MASs is of great scientific and economic value. The wireless communication network is the medium of information transfer between the agents and the controllers, while giving attacker an opportunity to attack. When network interaction information is affected by attacks, the performance and stability of the system are easily damaged. Moreover, under the condition of ensuring the performance of the system, the number of packet transmission is as small as possible, which will be conducive to the applications with limited bandwidth or congested network. Taking these problems into consideration, this paper studies sampled-data tracking control for uncertain heterogeneous nonlinear MASs under DoS attacks. This result provides a reference for the design of digital security controller, and can also be applied to intelligent engineering systems.}, archive = {J_TASE}, author = {Ning Zhao and Huiyan Zhang and Peng Shi}, doi = {10.1109/TASE.2024.3411074}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4771--4779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based sampled-data adaptive tracking control for heterogeneous nonlinear multi-agent systems under denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). Asynchronously delayed intermittent control for highly nonlinear stochastic delayed large-scale networks. TASE, 22, 4760--4770. (www)

  @article{ , abstract = {This paper investigates the stabilization of highly nonlinear stochastic delayed large-scale networks (HSDLN) based on asynchronously delayed intermittent decentralized control (ADIDC). By incorporating highly nonlinearity into the networks, practical problems can be more accurately modeled, and the nonlinear coupling relationships among nodes can be effectively captured, vastly expanding the practical application scope. ADIDC effectively integrates the advantages of asynchronously intermittent decentralized control and delayed intermittent control, enabling the networks to achieve the desired behavior. Subsequently, we set up an auxiliary timer and then define a newfangled Lyapunov functional. In particular, an innovative Lyapunov-based analysis, not utilizing Halanay-type differential inequality, is introduced to obtain a stabilization criterion for HSDLN, which effectively copes with asynchronously intermittent decentralized control and highly nonlinear, stochastic, and delayed elements. Meanwhile, a graph-theoretic methodology is employed to address the cross-terms arising in an asynchronous sense. Ultimately, as a practical application of the theoretical findings, a class of FitzHugh-Nagumo models is considered, and the stabilization is validated through numerical simulations. Note to Practitioners—This paper is motivated by existing results on various intermittent controls for the dynamics of complex networks. The existing results mainly require intermittent controls to be continuously/discretely synchronous sampled-data each node of complex networks, which limits the application scope of intermittent controls. This paper designs a new hybrid control called asynchronously delayed intermittent decentralized control. Particularly, asynchronously intermittent sampled-data control is a special case of control in this paper. Moreover, some lower conservative sufficient conditions are derived, such as local Lipschitz condition and polynomial growth condition instead of global Lipschitz condition and linear growth condition for coefficients of complex networks, and the average work time ratio instead of the maximum rest time ratio for intermittent control. The proposed results provide a novel idea for the stabilization study of FitzHugh-Nagumo models, and it is expected that the proposed methods can be extended to more actual physical engineering systems.}, archive = {J_TASE}, author = {Hui Zhou and Zhenyu Ye and Ju H. Park and Wenxue Li}, doi = {10.1109/TASE.2024.3410979}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4760--4770}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronously delayed intermittent control for highly nonlinear stochastic delayed large-scale networks}, volume = {22}, year = {2025}, }
• (2025). Semi-supervised image domain adaption for aerial refueling drogue detection on embedded chip under foggy conditions. TASE, 22, 4748--4759. (www)

  @article{ , abstract = {The application of aerial refueling technology to UAVs can reduce the dependence on the pilot’s operation, which has unique advantages in carrying out battlefield reconnaissance, monitoring suspicious targets and collecting intelligence through all-weather work. The existing vision-based drogue detection methods are assumed to be carried out under daily lighting conditions, but special weather, such as fog, makes it difficult to identify the characteristics of the drogue, which will greatly degrade the model performance or even fail. Moreover, the traditional computing architecture is difficult to be directly applied to real airborne equipment, so it is necessary to adopt AI processor module with faster and better computing power and supporting parallel computing to meet the requirements of low delay and high security in aerial refueling. Therefore, this work proposes a robust detection network based on image domain adaption. Firstly, an end-to-end image defogging module is designed to deal with foggy image enhancement under weak supervision. Then, knowledge distillation with the teacher-student network is applied to guide the student model to obtain the instance-level features of the unlabeled target domain. In addition, the detection model is compiled and transplanted on the system-on-chip chip of JFMQL100TAI. The comprehensive experimental results on public datasets and real refueling datasets validate the effectiveness and feasibility of the proposed work, which effectively complements the drogue detection of special autonomous aerial refueling tasks. Note to Practitioners—As a widely used refueling technology in the field of national defense, probe-and-drogue refueling has developed from manual control docking to monitoring auxiliary docking. However, it is difficult for pilots to accurately and quickly obtain the relative position of the refueling drogue through visual perception. In this work, a semi-supervised drogue detection network for special aerial refueling task is designed. The proposed work has good application potential in refueling scenes, which can provide fast and accurate drogue positioning under foggy conditions.}, archive = {J_TASE}, author = {Wei Tong and Ai Gu and Xingying Wu and Xiangyang Deng and Yandong Cai and Ya Duan and Yuhong Hou}, doi = {10.1109/TASE.2024.3448255}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4748--4759}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semi-supervised image domain adaption for aerial refueling drogue detection on embedded chip under foggy conditions}, volume = {22}, year = {2025}, }
• (2025). Robust drogue positioning system based on detection and tracking for autonomous aerial refueling of UAVs. TASE, 22, 4737--4747. (www)

  @article{ , abstract = {In modern war, endurance mileage and combat radius are important factors for the combat effectiveness of military aircraft. However, the existing aerial refueling schemes mainly use manual docking operations and complex conditions require pilots to have high operating skills. Therefore, this work proposes a visual positioning system for autonomous aerial refueling without adding cooperation marks, which mainly includes a dynamic graph convolution module for drogue detection and a kernel correlation filter for drogue tracking. Firstly, a dynamic GCN module is designed to generate the correlation matrix to aggregate adjacent high-order features and fuse them with global features extracted from the CNN stream to achieve accurate drogue detection. Then, the multi-scale features extracted from the drogue detector network and the HOG features are input to the filter learning module together, and weighted response maps are fused to alleviate the occlusion and scale change problems in the tracking process. In addition, a visual positioning scheme combining a drogue detector and tracker is introduced to output an accurate drogue ROI area. The effectiveness and robustness of the proposed work are verified by comparison with the mainstream methods on COCO detection datasets and real aerial refueling datasets. Note to Practitioners—For probe-and-drogue refueling, the combination of autonomous aerial refueling technology and machine vision technology can improve the real-time and robustness of autonomous refueling docking system in complex aerial visual scenes. Therefore, this work studies the accurate recognition of drogue detection and real-time drogue tracking. In addition, a drogue ROI positioning system is also designed. Experiments on public datasets and real refueling scenarios validate the feasibility and effectiveness of the proposed work, which has good application potential in refueling scenes and can provide decision-making for autonomous refueling and manual docking refueling of UAVs.}, archive = {J_TASE}, author = {Kevin W. Tong and Jie Wu and Yu-Hong Hou}, doi = {10.1109/TASE.2023.3308230}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4737--4747}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust drogue positioning system based on detection and tracking for autonomous aerial refueling of UAVs}, volume = {22}, year = {2025}, }
• (2025). Multi-sensory visual-auditory fusion of wearable navigation assistance for people with impaired vision. TASE, 22, 4724--4736. (www)

  @article{ , abstract = {Visually impaired individuals face limited mobility and restricted independent navigation in complex environments. Improving mobility and independent navigation for people with visual impairments is essential. In this paper, we introduce wearable electronic glasses (E-Glasses) that utilize a target detection network to fuse visual and auditory information for searching desired targets. Integrated with the electronic glasses, we present a neural path planning that combines spiking neural and convolutional neural networks. Several participants took part in experiments that showcased the remarkable capabilities of the developed system in target detection and navigation. The experimental results revealed an impressive success rate of 95.46% for the target detection network, providing participants with more accurate target information. Additionally, the neural path planning network achieved a success rate of 92.60%, demonstrating a significant speed advantage compared to the enhanced $A^{\ast }$ algorithm. Note to Practitioners—This article aims to address the mobility and navigation challenges faced by visually impaired individuals in complex environments. Many solutions have been proposed to assist with wearable navigation by utilizing sensors to perceive the environment and provide path information to users. In this article, we introduce wearable electronic glasses (E-Glasses) with advanced target detection and path planning capabilities to help visually impaired individuals navigate more effectively in complex environments. We have developed a target detection network that integrates visual and auditory information to effectively search for desired targets. Additionally, we have developed a neural path planning algorithm that combines spiking neural networks and convolutional neural networks. Furthermore, experiments conducted in indoor navigation challenges demonstrate the feasibility of this approach. In future research, we will focus on further exploring the target detection network and refining the neural path planning algorithms to improve the overall performance of the system.}, archive = {J_TASE}, author = {Yang Song and Zhijun Li and Guoxin Li and Binglu Wang and Min Zhu and Peng Shi}, doi = {10.1109/TASE.2023.3340335}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4724--4736}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-sensory visual-auditory fusion of wearable navigation assistance for people with impaired vision}, volume = {22}, year = {2025}, }
• (2025). Multi-modal autonomous ultrasound scanning for efficient Human–Machine fusion interaction. TASE, 22, 4712--4723. (www)

  @article{ , abstract = {Robotic autonomous ultrasound imaging is a challenging task as robots require strong analytical capabilities to make sound decisions in complex spatial relationships. In this paper, we integrate visual and tactile information into the ultrasound robotic system drawing inspiration from the process of human doctors conducting ultrasound scans, and explore the impact of different modalities of information on our task. The proposed multimodal deep reinforcement learning (DRL) framework can integrate real-time visual feedback and tactile perception, and directly output 6D pose decisions to control the ultrasound probe, thereby achieving fully autonomous ultrasound imaging of soft, movable, and unmarked targets. We demonstrate the feasibility of our method on a simulation platform and propose an effective model transfer learning method. Subsequently, we conducted further evaluations of the approach in a real-world environment. The results indicate that our approach effectively enhances the performance of autonomous ultrasound scanning and manual adjustments further optimize the outcomes. Note to Practitioners—This work is motivated by the increasing demand for intelligent human-machine interaction in medical applications. By improving the automation of traditional medical scanning procedures such as ultrasound scanning, the efficiency of medical scanning can be greatly improved. In this work, we propose a multi-modal autonomous ultrasound scanning system based on DRL, which can be applied to improve the efficiency of human-machine interaction in medical environments to execute daily health screening or used in emergency situations.}, archive = {J_TASE}, author = {Chengwen Luo and Yuhao Chen and Haozheng Cao and Mustafa A. Al Sibahee and Weitao Xu and Jin Zhang}, doi = {10.1109/TASE.2024.3370728}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4712--4723}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-modal autonomous ultrasound scanning for efficient Human–Machine fusion interaction}, volume = {22}, year = {2025}, }
• (2025). Limited information based emergency response control for underwater vehicle systems with disturbances and DoS attacks. TASE, 22, 4700--4711. (www)

  @article{ , abstract = {This paper focuses on the issue of limited information based emergency response control for a specific category of underwater cyber-physical systems (UCPSs) confronted with multiple emergencies. The focus is on a typical underwater vehicle system (UVS), whose communication with the control center may be compromised by Denial-of-Service (DoS) attacks. The occurrence of dual DoS attacks leads to significant information loss, intensifying the complexity of decision-making and emergency response during critical situations. Moreover, abrupt ocean current disturbances or faults will also seriously affect the performance of UVSs. Drawing upon distinct DoS attack scenarios, this study introduces a novel emergency response decision mechanism. A limited information-based disturbance observer (DO) is then proposed to effectively handle various unknown disturbances, ensuring favorable disturbance estimation performance. Subsequently, as for different attack channels, two innovative emergency response controllers are respectively proposed, considering the constraints of limited information. Furthermore, distinct stability criteria are derived by using Lyapunov stability and stochastic analysis techniques to guarantee the stabilization of UVSs. Finally, a series of numerical results is presented to illstrate the efficiency of the proposed algorithm under various emergency scenarios. Note to Practitioners—This study presents an novel framework for emergency automatic control and decision-making in UVSs suffering from different emergencies. While most of previous research primarily concentrated on emergency situations in the physical layer, this article delves into both the information layer and the physical layer to tackle decision-making and response challenges within information-constrained environments. Specifically, a limited information-based DO is designed to dynamically model disturbances, such as ocean currents and actuator faults. The DO adapts its structure based on valuable information obtained from emergency monitoring, ensuring accurate disturbance estimation upon successful transmission. The proposed strategy holds significant practical applicability in UVSs for handling emergencies, and further experimental validations are planned to be conducted on actual underwater vehicles.}, archive = {J_TASE}, author = {Weixin Wang and Yang Yi and Qin Wang and Mouquan Shen and Guangyu Zhu and Jun Yang}, doi = {10.1109/TASE.2024.3352071}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4700--4711}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Limited information based emergency response control for underwater vehicle systems with disturbances and DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Large-scale person re-identification for crowd monitoring in emergency. TASE, 22, 4691--4699. (www)

  @article{ , abstract = {The task of associating photographs/videos of an individual obtained from the same camera on various occasions or across cameras is called Person Re-identification (PRId). Computer-aided monitoring of persons of interest is an active research area in automated visual surveillance. It becomes more substantial in emergencies like natural disasters, contrived incidents, and public health crises. Part-level features of a pedestrian image hold significant importance in person retrieval. Traditionally, part-based PRId tasks required pose estimators or body part detectors for the hard partition of the pedestrian image. However, such approaches attract additional issues due to their dependency on external cues. This article emphasized employing the convolutional partition of body parts to learn discriminative part features. We focus on two significant contributions: (I) A parallel architecture called Convolutional Part Refine (CPR) and (II) Three different convolutional part refine strategies of outliers to handle the existing inconsistencies of uniform partition. The experiments confirm that CPR achieves competitive performance with state-of-the-art methods. Note to Practitioners—This work is motivated by the need to quickly recognize a target person captured across multiple camera views in a crowded environment. Presently, there is no ideal person re-identification solution. This article highlights the future requirements of smart visual surveillance through person detection and re-identification (re-id). The live feed of CCTV cameras can simultaneously detect and re-identify the target person to proactively handle the surveillance issues. The method detects and identifies a person’s identity captured in a CCTV by searching across an extensive database of images known as a gallery set. However, it is difficult to automatically recognize an individual across multiple camera views due to challenging scenarios such as low resolution, occlusion, background clutter, viewpoint, and illumination variations. Thanks to the deep learning-based body-part partition strategies that facilitate learning discriminative features of the target person. Traditionally, hard and soft partition strategies were used to partition the body parts. However, the proposed method focuses on a recent convolutional part partition strategy. Most part-based approaches assume that all the pixels in each part partition are homogeneous. However, the proposed method highlights the within-part-inconsistency problem. Against this background, this paper provides researchers and practitioners with a short review of the part-based body partition strategies. The proposed method demonstrates the effectiveness of convolutional part-partition over the hard and soft partition of body parts over three publicly available benchmark datasets. The proposed process also introduces a refine strategy to reduce the within-part-inconsistency issues in the part partitions.}, archive = {J_TASE}, author = {Nayan Kumar Subhashis Behera and Pankaj Kumar Sa and Khan Muhammad and Sambit Bakshi}, doi = {10.1109/TASE.2023.3318007}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4691--4699}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Large-scale person re-identification for crowd monitoring in emergency}, volume = {22}, year = {2025}, }
• (2025). End-to-end video captioning based on multiview semantic alignment for Human–Machine fusion. TASE, 22, 4682--4690. (www)

  @article{ , abstract = {Video captioning can understand videos, provide decision-makers with user-friendly natural language narration, alleviate the gap between man and machine, and promote human-machine interaction. Therefore, it has good application prospects in emergency response scenarios, such as aerial refueling and assisted driving. However, there are two problems with the current video captioning methods: 1) they are mainly oriented to general domains, and there are few studies on industrial applications; 2) the methods only interact with the semantics of video and text from a single view (tokens or sentences). For the above problems, this paper proposes a multi-view end-to-end video caption (MVVC) method for human-machine fusion. Compared with the previous video captioning methods, 1) the MVVC model is an end-to-end model which directly takes video frames as input without object detection for each frame; 2) we perform cross-modal interaction of video and text from both local and global views. So the model can simultaneously understand video content and generate text at two granularities(tokens to sentences). In order to verify the performance of the new model, we conducted a series of comparative and ablation experiments on MVVC on the two data sets of aerial refueling and automatic driving. The experiments show that our new method has a stronger video understanding ability and can generate more accurate video descriptions. At the same time, it also verified that the video captioning task could promote human-machine fusion and assist decision-making in emergency scenarios. Note to Practitioners—The motivation of this paper is to convert the video into natural language so that the autonomous system can automatically understand the observed scene, describe it to relevant stakeholders, and promote human-machine fusion. However, the traditional method needs to process the video offline and has an insufficient understanding of the video content. Therefore, this paper proposes an end-to-end video capture method based on multi-view semantic alignment, which can understand the video content directly from the original video pixels in real time and improve captioning accuracy. It can meet the application requirements of the industrial field and has practical application value.}, archive = {J_TASE}, author = {Shuai Wu and Yubing Gao and Weidong Yang and Hongkai Li and Guangyu Zhu}, doi = {10.1109/TASE.2024.3352903}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4682--4690}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {End-to-end video captioning based on multiview semantic alignment for Human–Machine fusion}, volume = {22}, year = {2025}, }
• (2025). An early warning approach for pilots’ cognitive tipping points based multi-modal signals. TASE, 22, 4670--4681. (www)

  @article{ , abstract = {When executing complex missions in emergency scenarios, pilots’ cognitive state may deteriorate, posing significant challenges to flight safety and mission execution. This paper proposes a cross-subject and cross-session early warning approach based on small-sample and multi-modal signals for predicting cognitive collapse state. We designs an experimental paradigm that has been demonstrated to induce cognitive collapse in 87% of trials by analyzing questionnaire scores, physiological signals, and task performance. The extracted multi-modal features are fused and selected to construct the optimal feature set. Further, a two-step early warning method is introduced to identify critical slowing down and to predict tipping points by detecting the high cognitive workload state and classifying the corresponding signals into early warning signal (EWS) and normal state. The early warning of pilot’s cognitive tipping point obtains the result that recall is 77.76%, false alarm rate is 28.43%, and the AUC of the warning method can reach 0.83. Our method shows better early warning performance compared with other classification models and has strong generalizability on other datasets. Note to Practitioners—The pilots’ cognitive state is crucial for mission completion and flight safety. Previous studies mainly focused on assessing current cognitive state, and there is little research on early warning of cognitive collapse that occurs in emergency scenarios (e.g., faults of system and the sudden increase in flight missions). In particular, identification is not effective for small-sample and instable data. Therefore, this paper presents an early warning method for cognitive tipping points, which can provide timely warning of pilots’ cognitive collapse and reduce human-caused risks in emergency scenarios.}, archive = {J_TASE}, author = {Si Wang and Yadong Liu and Dewen Hu}, doi = {10.1109/TASE.2024.3384504}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4670--4681}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An early warning approach for pilots’ cognitive tipping points based multi-modal signals}, volume = {22}, year = {2025}, }
• (2025). “Jumpingly” perceive time series: Image generation approach to modeling functional brain activation. TASE, 22, 4661--4669. (www)

  @article{ , abstract = {This paper presents a novel Linear Mapping Field (LMF) to map time series into two-dimensional images. The LMF extracts deeper features of fNIRS signals, which makes fNIRS less reliant on some prior. The developed convolution neural networks detect more prominent features than the state-of-the-art methods. The experimental results indicate that different from RNNs which can only perceive the time series in a “sequential” manner, LMF’s characteristic of “jumpingly” perception is the key to achieving excellent results. Note to Practitioners—As an optical and non-invasive technique to obtain the changes of oxyhemoglobin (O2Hb) and deoxyhemoglobin (HHb), fNIRS can be used to measure the changes of cerebral hemodynamics related to brain activities. This work proposes a linear mapping field and other mapping fields to map fNIRS signals to two-dimensional images, and the deep features of these generated images can be further extracted by convolutional neural networks, establishing an end-to-end bridge to detect the activation of brain functions under different tasks. Compared with mainstream methods, this work can be used as an effective mapping for fNIRS with low computational complexity and good performance.}, archive = {J_TASE}, author = {Yuxuan Yao and Xinying Wu and Kevin W. Tong and Miaomiao Zhang and Zhiyi Shi and Jin Ma and Yu-Hong Hou}, doi = {10.1109/TASE.2023.3347756}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4661--4669}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {“Jumpingly” perceive time series: Image generation approach to modeling functional brain activation}, volume = {22}, year = {2025}, }
• (2025). Human-in-the-loop assisted trafficability prediction for planetary rover on soft dangerous terrains. TASE, 22, 4651--4660. (www)

  @article{ , abstract = {It needs to predict the soft terrain trafficability of the Mars rover to avoid hazards in emergencies. The current prediction models of Mars terrain trafficability simply transfer information between humans and machines, with lower levels of human-machine interaction fusion, which hinders the creation of effective emergency handling measures. In this paper, a human-in-the-loop multi-indices fusion decision framework called HITL-FD is proposed for Mars rover trafficability prediction. The framework deals with the fuzziness of the trafficability boundaries by fuzzy theory, achieving the fusion processing of multiple evaluation indices and adopting a fuzzy decision tree algorithm to predict trafficability. Additionally, human-assisted decision is incorporated into the decision loop, and a reward function based on decision error is proposed to generate a closed-loop decision framework. The experimental results show that the framework has excellent trafficability prediction performance and reduces 5.1% of MAYBE terrain uncertainty compared to machine-based decision. Note to Practitioners—The trafficability prediction is crucial for Mars rovers traversing soft terrain to avoid emergencies including sinkage. Conventional trafficability prediction models have a low degree of human-machine fusion, which is not beneficial for applications of planetary rovers in practical engineering. In this paper, a multi-indices human-machine fusion decision framework is proposed, specifically designed for predicting the trafficability of Mars rover. In this framework, fuzzy theory is used to address the ambiguity associated with trafficability boundaries. It provides an overall evaluation of trafficability by fusing a single index and the use of a fuzzy decision tree algorithm to make predictions regarding trafficability. Furthermore, the framework incorporates human-assisted decision-making within the closed decision loop, and proposes a reward function based on decision error. Experimental results show that the framework can increase the correctness of trafficability prediction.}, archive = {J_TASE}, author = {Sui Wang and Lihang Feng and Di Xiao and Yong Hu}, doi = {10.1109/TASE.2024.3353815}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4651--4660}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human-in-the-loop assisted trafficability prediction for planetary rover on soft dangerous terrains}, volume = {22}, year = {2025}, }
• (2025). Semantic encoding algorithm for classification and retrieval of aviation safety reports. TASE, 22, 4643--4650. (www)

  @article{ , abstract = {Automated analysis of aviation safety reports is helpful in effectively preventing future accidents and improving emergency response capabilities. To date, there are no publicly available large-scale aviation text similarity datasets, which hinders the successful application of NLP techniques in the aviation domain. We present an automatically created aviation text similarity dataset consisting of more than 500,000 pairs for fine-tuning pretrained language models. Since technical terms have specialized meanings that differ from everyday language, we propose an efficient semantic encoding algorithm to improve the ability of embeddings to adequately represent aviation terms. We provide new solutions and revised evaluation metrics for the classification and the retrieval of safety reports, confirming the reliability of our dataset and the superiority of our algorithm. Note to Practitioners—Text representation is an essential task in natural language processing(NLP). A crucial step towards the successful application of NLP in safety reports analysis is to ensure that aviation texts are adequately encoded. Aiming at the problem of poor ability of current embeddings to represent technical terms, we automatically create an aviation text similarity dataset and propose a semantic encoding algorithm for aviation terms. It is clear that the proposed method has great potential in representation of technical terms, thus providing assistance for downstream tasks such as text classification, information retrieval and question answering.}, archive = {J_TASE}, author = {Yubing Gao and GuangYu Zhu and Ya Duan and Jianfeng Mao}, doi = {10.1109/TASE.2024.3359356}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4643--4650}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Semantic encoding algorithm for classification and retrieval of aviation safety reports}, volume = {22}, year = {2025}, }
• (2025). Service level and operational resilience assessment of urban rail transit under disruption conditions based on multimodal information perception. TASE, 22, 4633--4642. (www)

  @article{ , abstract = {Urban Rail Transit (URT) system is often disturbed by various types of security incidents during operation. Therefore, effective utilization of passenger in-transit information to dynamically assess the URT system’s abilities to maintain and resume normal operation under disturbances (operational resilience) is the key and prerequisite for URT safety management and emergency response. Based on typical operational disruption scenario, this paper study the multimodal data-driven URT operational resilience and its evaluation methods. The main contributions of this research includes: capturing key information about passengers in the travel process based on historical data, and quantifying the system’s service capabilities; proposing the concept of operational resilience oriented towards URT emergency management based on service level; employing the resilience curve to analyze the evolutionary patterns of URT service levels during safety events, and emphasizing the relativity and dynamism of operational resilience; constructing a two-stage “disruption-recovery” URT operational resilience evaluation model and taking Hangzhou URT as an example to verify the effectiveness of the model through experiments. The results provide a good reference for URT safety status assessment and emergency response. Note to Practitioners—With the expansion of urban rail transit scale, the passenger flow it undertakes is also increasing. Frequent security incidents can seriously disrupt the operation of urban rail transit, causing significant economic losses and even threatening the safety of passengers. If the ability of urban rail transit to resist safety incidents and restore normal operation can be understood in advance, it will help operation managers make reasonable judgments and emergency decisions. Therefore, this paper proposed an evaluation method for operational resilience with the service level of urban rail transit as the key.}, archive = {J_TASE}, author = {Guangyu Zhu and Yiru Wang and Liang Mu and Wenjie Duan and Rengkui Liu and Bo Wu and Rob Law}, doi = {10.1109/TASE.2024.3357476}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4633--4642}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Service level and operational resilience assessment of urban rail transit under disruption conditions based on multimodal information perception}, volume = {22}, year = {2025}, }
• (2025). Emergency control method of multi-modal passenger flow in urban rail transit. TASE, 22, 4622--4632. (www)

  @article{ , abstract = {Emergencies often lead to multi-modal and unbalanced distribution of urban rail transit passenger flows. Aiming at this problem, a emergency control method of multi-modal passenger flow is proposed under the premise of comprehensively considering the operation cost and passenger travel efficiency of urban rail transit(URTS).The main work includes: 1) From the perspective of train scheduling, designing and developing a contingency strategy for multi-modal passenger flow by combining the train operation plan based on full-length and short-turn routing, station passenger flow restrictions, and dynamic train departure intervals; 2) From the perspective of train operation and passenger travel synergy, the emergency control model of passenger flow is established with the objective of minimizing the total train operation time and average passenger waiting time; 3) A multi-agent deep reinforcement learning(MDRL) method with a new action selection, reward and double loop mechanism-ARDQMIX is proposed to realize emergency autonomous perception and control of passenger flow in urban rail transit. The simulation results show that the emergency control method of multi-modal proposed in this study has a good utility for improving passenger travel efficiency and reducing operation cost. Note to Practitioners—Urban rail transit system(URTS) is punctual, fast, safe and large capacity, and has gradually become the preferred means of transportation for passengers. However, due to the rapid growth of travel demand, the URTS in the morning and evening peak hours traffic problems are particularly serious, will produce a multi-modal passenger travel demand, such as can not be quickly alleviated, the traffic problem will be caught in a vicious circle, leading to the urban rail transit system service capacity decline, and even lead to the potential safety problems. Aiming at this problem, a emergency control method of multi-modal passenger flow is proposed under the premise of comprehensively considering the operation cost and passenger travel efficiency of URTS.}, archive = {J_TASE}, author = {Guangyu Zhu and Liang Mu and Ranran Sun and Nuo Zhang and Bo Wu and Peng Zhang and Rob Law}, doi = {10.1109/TASE.2023.3322031}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4622--4632}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Emergency control method of multi-modal passenger flow in urban rail transit}, volume = {22}, year = {2025}, }
• (2025). Multi-objective optimization of a lower limb prosthesis for metabolically efficient walking assistance. TASE, 22, 4606--4621. (www)

  @article{ , abstract = {Enhancing energy efficiency is pivotal for advancing robotic lower limb prostheses. However, the conventional control strategy in powered prostheses displays notable variance in energy economy across terrains during walking. This challenge could be managed by developing control optimization methods to effectively reduce the subject’s metabolic cost and the device’s energy expenditure, thereby improving the walking economy in diverse environments. In this study, we investigate a multi-objective optimization problem (MOP) for controlling above-knee prostheses, intending to reduce prosthetic power consumption and the subject’s metabolic costs during walking activities. To bolster real-time performance in the optimization loop, we introduce a knee-guided evolutionary algorithm (KGEA) for MOP, which efficiently reduces exploration space and time complexity, thus enabling the practical implementation of a limited number of solutions in each iteration. The effectiveness of the proposed optimization-based strategy was evaluated through experiments involving two lower limb amputees engaging in slope, stair, and flat walking. Our results demonstrated that the optimization-based strategy significantly reduced the subjects’ metabolic consumption, with an average reduction of 16.53% for subject 1 and 10.23% for subject 2 across different terrains, compared to using the prosthesis without the optimization strategy. This promising outcome marks a substantial advancement in the development of energy-efficient, multi-degree-of-freedom power prostheses. Note to Practitioners—Lower Limb amputees using powered prostheses experience energy efficiency challenges due to limited battery life, particularly with multi-degree-of-freedom prostheses over varied terrains. Additionally, compared to passive alternatives, the increased weight of powered prosthetics poses a challenge in managing users’ metabolic costs, requiring suitable optimization strategies to offset the additional burden and optimize body economy during locomotion. The proposed KGEA-based multi-objective optimization method, designed explicitly for above-knee prostheses, targets the reduction of prosthetic power consumption and subject’s metabolic costs during walking activities. This innovation will likely benefit individuals with lower limb amputations, offering them improved mobility and comfort across various terrains. By addressing variance in varied terrains and inefficiencies in energy consumption caused by traditional control strategies, this work can assist practitioners in creating more effective and user-friendly prosthetic devices. Furthermore, we explore continuous refinement of the multi-objective optimization model and KGEA to accommodate a broader range of user-profiles and walking scenarios, ensuring its applicability in real-world situations.}, archive = {J_TASE}, author = {Guoxin Li and Jiandong Wang and Zhijun Li and Gary G. Yen and Qi Qi and Jinqiu Xing}, doi = {10.1109/TASE.2024.3370615}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4606--4621}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-objective optimization of a lower limb prosthesis for metabolically efficient walking assistance}, volume = {22}, year = {2025}, }
• (2025). Trust-aware Human–Robot fusion decision-making for emergency indoor patrolling. TASE, 22, 4596--4605. (www)

  @article{ , abstract = {Trust plays a crucial role in decision-making during human-robot collaboration, particularly in emergency scenarios where it becomes more susceptible due to dynamic factors. Misalignment of human-robot trust significantly hampers the efficiency of the collaboration. Therefore, it is imperative to establish effective human-robot interaction and decision support mechanisms that mitigate biases in human confidence levels regarding the robot’s capabilities in dynamic environments. Additionally, online correction of human-robot trust based on human behavioral feedback is vital. This paper focuses on a specific type of emergency task scenario, specifically, indoor human-robot collaborative patrolling in the event of sudden power outages. We propose a trust model based on linear Gaussian and sparse Gaussian processes (sparse GP). We also employ Monte Carlo Tree Search (MCTS) method to determine the robot’s optimal fusion decision-making. Through VR-based human-robot collaborative experiments, we ascertain that the robot prioritizes enhancing human-robot trust in emergency scenarios to mitigate the long-term costs of human-robot collaboration. Note to Practitioners—The primary motivation of this paper lies in tackling the issue of decreased collaboration efficiency in human-robot cooperation, stemming from irrational human decision-making, a problem exacerbated in emergency scenarios where comprehending all available information proves challenging. Grounded in the realm of human-robot trust, this study evaluates the temporal and substantive aspects of human-robot interactions contingent upon the degree of trust. Moreover, it adapts the fusion decision-making process within the human-robot team in accordance with the decision-making strategies adopted by humans at varying trust levels. The optimized decisions of the human-robot ensemble are conveyed to the human participant via decision support. This approach ensures the maintenance of trust levels, facilitating the acceptance of decision support that may seem intuitively irrational but holds objective superiority. Furthermore, it mitigates redundant human-robot interactions once a sufficient level of trust is established.}, archive = {J_TASE}, author = {Yang Li and Jiaxin Xu and Di Guo and Huaping Liu}, doi = {10.1109/TASE.2024.3350639}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4596--4605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trust-aware Human–Robot fusion decision-making for emergency indoor patrolling}, volume = {22}, year = {2025}, }
• (2025). Rat-UAV navigation: Cyborg rat autonomous navigation with the guidance of a UAV. TASE, 22, 4583--4595. (www)

  @article{ , abstract = {Controlling the cyborg rat with the guidance of an unmanned aerial vehicle (UAV) is important yet challenging. UAV can provide a mobile bird’s eye view (BEV) with multiple shooting angles to expand the motion space of the cyborg rat. However, uncertainty is a critical problem arising from the obscure and blurry imaging from UAV and the locomotion willingness of cyborg rats. To solve this problem, we propose the collaborative Rat-UAV navigation (RUN) paradigm. The uncertainty in RUN is formulated using the partially observed Markov decision processes (POMDP) in a perceptual-control framework. First, perceptual uncertainties in rat pose estimation and environmental modeling are reduced. Second, control policies are delivered to maximize rewards and minimize control uncertainties until the cyborg rat reaches the target. To achieve this, RUN can accurately detect tiny rats with fewer parameters, dynamically plan paths with partial observations from the UAV, and stably control the rat using a locomotion willingness transition graph in a large practical arena. Our framework is tested in real-world experiments. For the first time, a cyborg rat successfully completed a navigation task with the guidance of a UAV.Note to Practitioners—This paper studies, for the first time, the problem of navigating a cyborg rat with the guidance of an unmanned aerial vehicle (UAV). Most existing researches on the autonomous navigation of cyborg rats are conducted in fully observed environments with fixed cameras, disregarding uncertainties in complex environmental perception and rat control. In this paper, we propose a collaborative Rat-UAV navigation (RUN) paradigm, which formulates uncertainties using the partially observed Markov decision processes (POMDP). Specifically, RUN enhances the cyborg rat’s interaction with the UAV in the navigation environment. First, it narrows the gap between the rat’s perceived and actual states by utilizing keypoint detection, dynamic environment modeling, and locomotion willingness estimation from the view of the UAV. Secondly, RUN constructs a policy generation module to maximize the reward of controlling the cyborg rat, based on a locomotion willingness transition graph. Experimental results demonstrate that our framework effectively guides the cyborg rat through a large maze, followed by six targets, with the assistance of the UAV. In future researches, we will investigate the real-time prediction of rat locomotion willingness using multimodal neuroethological recordings, towards a bidirectional brain computer interface between rats and UAVs.}, archive = {J_TASE}, author = {Nenggan Zheng and Han Zhang and Le Han and Chao Liu and Qiqi Li and Pengfei Wang and Guangyu Zhu}, doi = {10.1109/TASE.2024.3378673}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4583--4595}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rat-UAV navigation: Cyborg rat autonomous navigation with the guidance of a UAV}, volume = {22}, year = {2025}, }
• (2025). Physical interaction oriented aerial manipulators: Contact force control and implementation. TASE, 22, 4570--4582. (www)

  @article{ , abstract = {Aerial manipulator (AM) systems are significantly more effective than both conventional manipulators and flying robots, especially in disaster rescue settings, because they can perform human-like interaction tasks in flight. However, controlling the contact force of an AM is difficult due to the coupling between its flying platform and manipulator. This paper proposes a contact force control framework to address this problem. First, the UAV/AM position response under an external force is studied, and it is theoretically shown that a closed-loop UAV/AM behaves as a spring-mass-damper system. Second, a contact force controller is designed using an inverse-dynamics method. Third, an attitude feed-forward approach is employed to improve the force tracking performance. Then, the real-time contact position is introduced into the control system to achieve interaction with an actively moving environment. Finally, an innovative AM is developed and subjected to flight experiments, validating the proposed framework. An AM’s characteristics in interaction operations are summarized, and general conclusions are drawn. This study is novel in that 1) the contact force control is implemented without relying on force sensors, 2) the whole framework is applicable for controlling a constant/variable contact force with high closed-loop performance, and 3) it can also perform reliable interaction with a dynamic and unknown environment. Note to Practitioners—This study is motivated by the contact force control problem of an aerial manipulator (AM) during interaction operations. Previous approaches have explored the feasibility of controlling contact forces to some extent, but they lack universality and assume a static environment. Ensuring sufficient safety and providing reliable solutions in real-world applications remain challenging. This paper aims to investigate the position response of a closed-loop aircraft system under external forces, without disrupting its existing steady flight. The proposed method transforms contact force control into position control, eliminating the need for a force sensor. A series of aerial experiments were conducted to validate the effectiveness and applicability of the method in various scenarios. Our ongoing work will focus on migrating such an AM system from a laboratory scenario to the real world.}, archive = {J_TASE}, author = {Xiangdong Meng and Yuqing He and Jianda Han and Aiguo Song}, doi = {10.1109/TASE.2024.3354996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4570--4582}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Physical interaction oriented aerial manipulators: Contact force control and implementation}, volume = {22}, year = {2025}, }
• (2025). Magnetothermal analysis of chitosan microrobots for cancer therapy. TASE, 22, 4559--4569. (www)

  @article{ , abstract = {Cancer starkly jeopardizes the security of human life as an intractable challenge in the realm of emergency medical care. Existing therapeutic modalities have thus far proven inadequate in effecting a complete eradication of cancer. This study introduces a microrobot-assisted approach to cancer treatment through the conceptualization of a magnetothermal biocompatible chitosan microrobot. The designed biocompatible chitosan microrobot possessed excellent thermal stability and magnetothermal performance. Leveraging computational modeling and simulation within the COMSOL framework, it was ascertained that the microrobot possessed a capacity for precision-controlled temperature elevation through the augmentation of coil current intensity and magnetic field frequency. Furthermore, the applicability of the microrobot for cancer therapy was confirmed through the establishment of a liver cancer model, which proved the microrobot’s adeptness in swiftly affecting temperature elevation and effectually ablating cancer tissue within a short time. The accomplishments of this study provide foundational support for the landscape of cancer treatment in emergency medical scenarios, thereby offering valuable insights into the magnetothermal cancer treatment process of microrobots. Note to Practitioners—The motivation of this paper is to develop a magnetothermal biocompatible chitosan microrobot and investigate its precise magnetothermal control properties for application in emergency medical scenarios of cancer treatment. Magnetothermal therapy avoids the damage to normal tissues caused by the use of chemical agents in cancer treatment, and its precise temperature control is crucial for the therapeutic effect. Therefore, a magnetothermal microrobot was designed based on biocompatible chitosan material and its magnetothermal controllable performance was studied by simulation. The magnetothermal chitosan microrobot possessed good thermal properties, and the temperature rise behaviors of the microrobot could be effectively controlled by changing the current intensity and magnetic field frequency. Furthermore, the ability of the magnetothermal microrobot to rapidly and efficiently increase the temperature to kill cancer tissues was verified in a liver cancer model. Therefore, the magnetothermal chitosan microrobot has a rapid, precise, and controllable temperature rise effect and is suitable for emergency medical applications in cancer treatment.}, archive = {J_TASE}, author = {Jingxi Wang and Baoyu Liu and Kun Li and Junwei Xu and Ping Li}, doi = {10.1109/TASE.2023.3334029}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4559--4569}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Magnetothermal analysis of chitosan microrobots for cancer therapy}, volume = {22}, year = {2025}, }
• (2025). Magnetic actuation analysis of chitosan hydrogel robots for emergency medical applications. TASE, 22, 4548--4558. (www)

  @article{ , abstract = {Extensive medical cases have substantiated the imperative need for researching rapid, precise, and convenient soft robots to assist traditional surgical methods and ensure the safety of emergency medical care required to save the lives of patients, particularly in scenarios involving unexpected mutations within the body. Hence, this paper presents the design of an emergency medical assistance robot based on biocompatible magnetic chitosan hydrogels and magnetic actuation. Firstly, a description of the designed magnetic chitosan hydrogel robot was demonstrated. Subsequently, a mechanistic analysis of the magnetic response characteristics of the magnetic chitosan hydrogel robot was conducted. The COMSOL modeling and simulation results demonstrated the feasibility of employing a rectangular permanent magnet with a gradient magnetic field distribution for robot actuation. Furthermore, precise and controllable deformation behaviors of the designed robot could be achieved by regulating the strength of the driving magnetic field, the magnetic properties of the robot, and the distance between the robot and the permanent magnet. Moreover, simulation results validated the capability of the robots to execute tapping motion, biomimetic flight, and thrombus removal through advanced structural design. The accomplishments of this study provide foundational support for the development of magnetically actuated devices in emergency medical scenarios and offer valuable insights into the deformation motion processes of magnetic actuation. Note to Practitioners—The motivation of the article is to develop a biocompatible magnetic chitosan hydrogel robot and investigate its precise control methods for emergency medical applications. In emergency medical situations, there is a great demand for fast and accurate medical interventions that are simple and easy to operate. Therefore, a magnetic chitosan hydrogel robot is designed in the article, and its magnetic driving force characteristics are studied through experiments and simulations. The biocompatible chitosan hydrogel robot possessed excellent magnetic responsiveness and remarkable controllable deformation behaviors under different magnetic field driving conditions, as proved by COMSOL simulations. Furthermore, the effects of driving magnetic field strength, the magnetic properties of the robots, and the distance between the robot and the permanent magnet on the magnetic-driven deformation behaviors were investigated, which was of significant importance in guiding the control of robot deformation. Besides, various motions could be achieved through the structure design of robots. Consequently, the magnetic chitosan hydrogel robot possesses the characteristics of convenience, speed, and precision through magnetic actuation, making it highly suitable for emergency medical applications.}, archive = {J_TASE}, author = {Jingxi Wang and Baoyu Liu and Kun Li and Jin Ma and Ping Li}, doi = {10.1109/TASE.2023.3315482}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4548--4558}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Magnetic actuation analysis of chitosan hydrogel robots for emergency medical applications}, volume = {22}, year = {2025}, }
• (2025). Integral line of sight guidance scheme-based tracking method for snake robots. TASE, 22, 4537--4547. (www)

  @article{ , abstract = {This study investigates the trajectory tracking strategy of a snake robot with sideslip disturbance and unknown model parameters. To guide the robot to track the ideal trajectory faster and more accurately, an adaptive anti-sideslip strategy for a snake robot with the Integral Line-of-Sight (ILOS) function is reported. This technique eliminates direction sideslip and error fluctuation by using auxiliary integral terms and shortens the convergence time of state variables. Following the position and angle control objectives, the proposed controller considers the negative effects caused by the uncertainty and time variability of environmental parameters and compensates for the joint input using the adaptive update laws. The environment adaptability and tracking efficiency are improved. The stability analysis indicates that the state errors converge to the origin. The simulation and experiment data verifies the effectiveness and strength of the work. Note to Practitioners—This article was motivated by the problem of robust trajectory tracking for a snake robot in an environment with sideslip disturbance and unknown model parameters. In this environment, information of the motion space (for example, the coefficient of ground friction) cannot be obtained. In addition, there may be other system limitations (for example, motion sideslip limitations) and other operational limitations. These limitations are caused by the requirements of various common trajectory tracking objectives. These cases should also be considered in the control strategy. However, based on the existing methods of tracking control for snake robots, there is still a lack of a complete and reliable autonomous control scheme that can consider the above problems. On this basis, we present a reliable control strategy, which considers the above problems and the dynamic uncertainty of the model. In the future, we will extend the proposed method to the field of formation tracking control for multiple robots.}, archive = {J_TASE}, author = {Dongfang Li and Jiechao Zhou and Yanwei Huang and Dali Zhang and Ping Li and Aiguo Song}, doi = {10.1109/TASE.2023.3327958}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4537--4547}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integral line of sight guidance scheme-based tracking method for snake robots}, volume = {22}, year = {2025}, }
• (2025). Hierarchical task-oriented whole-body locomotion of a walking exoskeleton using adaptive dynamic motion primitive for cart pushing. TASE, 22, 4525--4536. (www)

  @article{ , abstract = {This paper proposes a hierarchical task-oriented whole-body locomotion framework for the exoskeleton walking-cart pushing (EWCP) task, which includes a straight gait and a bypassing gait, allowing the exoskeleton robot to avoid obstacles during walking. In this framework, the core components are gait planning and phase estimation for locomotion in unstructured environments. Notably, our mobile redundancy exoskeleton system can provide more flexibility and versatility in manipulation when performing complex tasks. For the hierarchical task-oriented whole-body locomotion, the detour gait consists of straight lines and turning shapes so that the EWCP system can avoid obstacles on the ground. For gait planning, we use the dynamic motion primitives to learn the joint motion trajectory of whole-body locomotion, which has good generalization ability and adaptability with respect to the gait. For phase estimation, the current gait phase can be estimated from the joint angles. Additionally, we design an task switching mechanism, where the exoskeleton system can switch different configurations flexibly for different scenarios, such as track switching only with both feet supported. And the phase estimation and gait switching strategy ensure the stability of task switching. The experimental results show that the exoskeleton can effectively accomplish EWCP tasks in an environment with obstacles. Our work has also shown that even some challenging motion tasks can be implemented with relatively simple controllers, which greatly simplifies the design of control systems. Note to Practitioners—This paper is motivated by issues of hierarchical tasks of the lower limb exoskeleton. Traditional exoskeletons cannot achieve obstacle avoidance in complex scenes because they do not have enough degrees of freedom or they do not use hierarchical locomotion. In this paper, a hierarchical task-oriented whole-body locomotion is proposed, which includes a straight gait and a bypassing gait, allowing the exoskeleton robot to avoid obstacles during walking. For gait planning, we use DMP to generate the gait trajectory, which can ensure the smoothness of trajectory. For phase estimation, the current gait phase can be estimated from the joint angles. Additionally, we design an task switching mechanism, where the exoskeleton system can switch different configurations flexibly for different scenarios, such as track switching only with both feet supported. And the phase estimation and gait switching strategy ensure the stability of task switching. The proposed hierarchical task-oriented whole-body locomotion framework is expected to be applied to exoskeletons to assist patients in rehabilitation training and users in mobility in daily life. Additionally, the proposed framework uses a series of motion primitives to learn and reproduce the trajectory and update it online, which requires a heavy computation load calculation to ensure the update speed of the trajectory. Therefore, we expect to reduce computation complexity for a walking exoskeleton in the future.}, archive = {J_TASE}, author = {Xu Hao and Zhijun Li and Pengbo Huang and Peng Shi and Guoxin Li}, doi = {10.1109/TASE.2023.3345919}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4525--4536}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical task-oriented whole-body locomotion of a walking exoskeleton using adaptive dynamic motion primitive for cart pushing}, volume = {22}, year = {2025}, }
• (2025). Collaborative line-of-sight guidance-based robust formation control for a multi-snake robot. TASE, 22, 4514--4524. (www)

  @article{ , abstract = {For the collaborative formation control problem of a multi-snake robot with uncertain sideslip and lumped unknowns, this study proposed collaborative line-of-sight (LOS) guidance and consensus-based path variable, which enables snake robots to track multiple parametric paths in a fixed formation. Meanwhile, the improved finite-time sideslip observer effectively compensates for the kinematic deviation, thereby improving the accuracy of the multi-snake robot’s path tracking. A robust adaptive controller based on an integral sliding mode manifold is developed to eliminate external disturbances. The model uncertainty and time-varying disturbances are comprehensively involved and incorporated as lumped unknowns in dynamics. The prediction-based RBF neural network estimators are derived to approximate unknowns and compensate controllers. The Lyapunov theory is adopted to prove the stability of formation errors for snake robots in the closed-loop system. Simulation and experiment results show the practicability and superiority of this work. Note to Practitioners—This paper is mainly motivated to study a cooperative formation control framework suitable for snake robots’ dynamics, especially for the application of resisting model uncertainty and external interference. In practice, the collaborative line of sight guidance law is employed to generate and adjust the reference direction and speed of robots in real time, and then the method is applied to the joint torque controller to guide robots to track the desired parameterized path. Meanwhile, the improved finite time sideslip observer corrects the deviation of the guidance law, ensuring the observation accuracy while reducing the computational time. During the control process, for highly coupled and complex snake robot models, solving external disturbances and model uncertainties remains a challenge. A neural network based state estimator is used to handle these unknowns to pre-compensate for underrotation/overrotation of the joint steering gear when disturbed. Simulation and experimental results validate that the proposed scheme can enable multiple snake robots to synergistically perform tasks, expressing high potential in unknown environments. In the future, we plan to design a three-dimensional spiral gait instead of the two-dimensional one used in this study. With this change, we will further improve the usability of snake robots in real-world scenarios.}, archive = {J_TASE}, author = {Yang Xiu and Yilong Zhang and Hongbin Deng and Hongkai Li and Yuanqing Xu}, doi = {10.1109/TASE.2023.3348469}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4514--4524}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Collaborative line-of-sight guidance-based robust formation control for a multi-snake robot}, volume = {22}, year = {2025}, }
• (2025). A bio-inspired decision-making memristive circuit based on classical and operant conditioning. TASE, 22, 4501--4513. (www)

  @article{ , abstract = {This work proposes a bio-inspired decision-making memristive circuit drawing on Hull’s secondary learning system. This circuit can not only mimic the decision-making initiated by secondary drive stimuli and shaped and guided by secondary reinforcers via integrating classical conditioning (CC) and operant conditioning (OC), but also consider the factors that influence decision-making, such as demand states, incentive motivation, and habit strength. These bionic functions have not yet been implemented by existing memristive circuits. Our circuit primarily includes CC module, drive regulation module, habit memory module, incentive generation module, and winner-takes-all module, which is designed through a modular hierarchical circuit design method. Memristors play a core role in our circuit and enable the circuit to perform brain-like online learning in an in-memory computing way, which has power and area advantages. The PSPICE-based simulations in various scenarios show that our circuit has a strong adaptive decision-making ability since more bionic features are considered. The proposed circuit can be applied to a bionic intelligent robot, enabling the robot capable of autonomous associative learning abilities to perform complex tasks such as detection and rescue. Note to Practitioners—This work is motivated by the problem of neuromorphic circuit design for bio-inspired learning and decision-making. To realize brain-like online in-situ learning in an in-memory computing way and enhance the adaptability of the circuit in dynamic environment, a memristive circuit integrating CC and OC is proposed. Referring to Hull’s secondary learning system, the proposed circuit takes into account factors that affect decision-making, such as demand states, incentive motivation, and habit strength, as well as the fact that decision-making processes can be evoked by secondary drive stimuli and shaped by secondary reinforcers, which is unaddressed by existing memristor-based works. The bionic foraging simulations in various scenarios show that our circuit can make favorable adaptive decisions based on various cues categorized by associative memories when engaging with their environment. Such a bio-inspired memristive circuit system can be applied to bionic robots or rescue detection robots through large-scale integration to achieve adaptive learning and decision-making of complex tasks with low power consumption.}, archive = {J_TASE}, author = {Chao Yang and Xiaoping Wang and Zhanfei Chen and Zilu Wang and Sen Zhang and Zhigang Zeng}, doi = {10.1109/TASE.2023.3326461}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4501--4513}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A bio-inspired decision-making memristive circuit based on classical and operant conditioning}, volume = {22}, year = {2025}, }
• (2025). Robust adaptive tracking control for aerial transporting a cable-suspended payload using backstepping sliding mode techniques. TASE, 22, 4490--4500. (www)

  @article{ , abstract = {Aerial transportation technology is the lifeline of air disaster rescue. In this article, a robust adaptive tracking control scheme using backstepping sliding mode techniques is proposed for a quadrotor-based aerial transportation system with a cable-suspended payload in disaster rescue, where the payload is ensured to be driven to predefined trajectories in the presence of strong coupling, uncertainties, and external disturbances. The quadrotor and the payload are modeled as a rigid body and a point mass, respectively, and the two coupling terms between the virtual input of the payload position loop and the payload attitude error as well as between the input force and the quadrotor attitude error are analyzed owing to the underactuated of the quadrotor-based transportation system. Then, adaptive backstepping sliding mode control strategies are designed for the position and swing dynamics of the payload to guarantee payload trajectory tracking, and an observer-based geometric attitude control method is presented for the quadrotor attitude dynamics to ensure the global attitude stability of the system, where prior information about disturbances is not required. The closed-loop stability of the whole system is strictly proven. Finally, real-world experiments are conducted to verify the feasibility and robustness of the proposed control scheme. Note to Practitioners—The motivation of this article is to investigate a robust and adaptive control tracking scheme for aerial transportation systems with a cable-suspended payload in disaster rescue. In most of the existing aerial transportation control schemes with a cable-suspended payload, the payload is driven to follow a desired trajectory while only considering the coupling effect between the aerial platform and the payload. However, in practical disaster rescue applications, the aerial transportation system is inevitably affected by strong coupling, uncertainties, and external disturbances. Meanwhile, to the authors’ best knowledge, there exist few studies that investigate payload following issues while considering strong coupling, uncertainties, and external disturbances simultaneously. Therefore, this article proposes a robust and adaptive tracking control scheme using backstepping sliding mode techniques for a quadrotor-based aerial transportation system with a cable-suspended payload to ensure the stable and accurate payload following control under strong coupling, uncertainties, and external disturbances, where prior information about disturbances is not required under the proposed scheme. The closed-loop stability of the whole system is strictly and mathematically analyzed as well as real-world experiments provide promising results. Moreover, the proposed scheme provides a more realistic setup for autonomous aerial transportation with cable-suspended supplies in disaster rescue.}, archive = {J_TASE}, author = {Jiacheng Liang and Yaonan Wang and Hang Zhong and Yanjie Chen and Hongwen Li and Hean Hua and Wei Wang}, doi = {10.1109/TASE.2024.3355225}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4490--4500}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust adaptive tracking control for aerial transporting a cable-suspended payload using backstepping sliding mode techniques}, volume = {22}, year = {2025}, }
• (2025). Multi-agent Q-net enhanced coevolutionary algorithm for resource allocation in emergency human-machine fusion UAV-MEC system. TASE, 22, 4473--4489. (www)

  @article{ , abstract = {Unmanned aerial vehicle (UAV) assisted communication has emerged as a powerful technology for reliable and flexible emergency communications (e.g., earthquakes, hurricanes and floods), especially when the mobile infrastructure is seriously damaged. UAV assisted mobile edge computing (UAV-MEC) system can be deployed in the natural disaster area as communication relay or air mobile base stations to resume communication and provide computing resources for the users in disaster areas. However, the optimized resource allocation performance of UAV-MEC system can be further guaranteed with the human-fusion decision making. In this paper, we construct an emergency human-machine fusion UAV-MEC system consisting of multiple UAVs equipped with computing resources, and the human-machine decision makings are fused for UAV deployment. In order to solve the resource allocation problem of human-machine fusion UAV-MEC system, we establish an human-machine deep integration model for UAV-MEC system, and the UAVs are dispatched reasonably through human-machine fusion decision makings to maintain efficient communication in emergency communication areas. To minimize task latency and improve the computation efficiency in emergency human-machine fusion UAV-MEC system, we consider the number of dispatched UAVs, deployment plans, flight plans, and simultaneously optimize the task allocation scheme, priority order, and task offloading ratio. We propose a reinforcement learning framework combined with evolutionary algorithms, which is named as multi-agent Q-net enhanced cooperative genetic algorithm (MQCGA), for resource allocation of UAV. Based on neural network forecasts, the greedy rate during training processing can be dynamically controlled, and the learning ability of different agents can be strengthened. Simulation experiments are conducted to evaluate the proposed framework, and the results show that our proposed MQCGA algorithm is significantly superior to other algorithms in terms of latency and energy consumption. Note to Practitioners—With the development of MEC and the popularity of UAVs, the potential of UAV-assisted MEC draw much attention from industrial field. Considering the lack of communication capabilities in a certain area in an unexpected situation, UAVs can be quickly deployed to corresponding locations and provide computing services. In this paper, an UAV-MEC system that integrates human-machine decision-making for emergency communication situations, named human-machine fusion UAV-MEC system, is considered. The system divides the scene into regions, models users and UAVs, and provides detailed deployment schemes, maximizing the practicality and applicability of the scene. In order to improve the communication efficiency in the case of emergency communication, this paper proposes a new resource scheduling algorithm and adds human-machine decision-making to enable UAVs to continuously provide efficient services for a certain area. The experimental results provide practitioners with a theoretical basis, such as the task completion time, UAV energy consumption and computing resource scheduling. Applying the system to actual scenarios also requires two preconditions of the system, one is the information collected by the large UAV, and the other is the communication among the UAVs. These two preconditions facilitate the human-machine fusion UAV-MEC system deployment in practical applications.}, archive = {J_TASE}, author = {Lu Sun and Ziqian Liu and Zhaolong Ning and Jie Wang and Xianping Fu}, doi = {10.1109/TASE.2024.3409551}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4473--4489}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-agent Q-net enhanced coevolutionary algorithm for resource allocation in emergency human-machine fusion UAV-MEC system}, volume = {22}, year = {2025}, }
• (2025). Human-UAV interaction assisted heterogeneous UAV swarm scheduling for target searching in communication denial environment. TASE, 22, 4457--4472. (www)

  @article{ , abstract = {Unmanned aerial vehicle (UAV) swarm shows great potential as an effective tool for target tracking through completing complex tasks by collaboration of heterogeneous UAVs. However, UAV swarm scheduling faces challenges with poor quality communication and obstacles, especially in communication denial environment with multiple obstacles. To overcome these challenges, first, this paper proposes a scheduling slot model which divides the scheduling process into multiple time slots, allowing UAVs to communicate in communication slots while predicting instead of communication in communication denial slots. In communication denial slots, this model utilizes route fitting and two-stage Kalman filtering for UAV location prediction and optimizes UAV scheduling to align with predicted positions. In enabled slots, this model corrects position deviations to obtain precise UAV locations manually. Then, we propose an obstacle avoidance strategy to facilitate swarm scheduling for target searching under communication constraints. The obstacle avoidance strategy simplifies obstacles as regular hexagons and facilitates the determination of UAV avoidance routes by introducing intermediary points. Finally, to optimize UAV scheduling strategy, we propose a region co-evolution algorithm (RCEA), which emphasizes the collaboration among diverse individuals or populations. RCEA adopts area evaluation and Pareto strategy to enhance scheduling efficiency with following three steps. RCEA divides the overall scheduling region into multiple sub-regions, generates the foundational solution pool through the implementation of the area evaluation or Pareto strategy, and then proceeds to execute the region cooperation process base on the foundational solution pool. Simulation experiments are conducted to validate the performance of human-UAV interaction scheduling model with proposed scheduling methods and obstacle avoidance strategy. The simulation results demonstrate that RCEA outperforms other scheduling algorithms for UAV swarm in communication denial environment with multiple obstacles. Note to Practitioners—This paper addresses challenges inherent in real-world application scenarios, and the proposed algorithm has the potential to bring many benefits to practitioners. Firstly, the scheduling slot model can be applied not only to UAV swarm for target searching but can also be extended to other swarm devices for complex tasks with collaboration relying on communication support while facing poor quality communication or obstacles. Secondly, the proposed RCEA focuses on collaboration and region partitioning, the algorithm demonstrates remarkable scalability, effectively tackling challenges across diverse scales and complexities. Thirdly, the experimental scenarios can serve as a validation dataset for other peer researchers, and although the simulation experiment is based on a 2D movement model, this study still offers theoretical support applicable to a 3D movement model.}, archive = {J_TASE}, author = {Lu Sun and Jiashuai Wang and Liangtian Wan and Kuixian Li and Xiaojie Wang and Yun Lin}, doi = {10.1109/TASE.2024.3412005}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4457--4472}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human-UAV interaction assisted heterogeneous UAV swarm scheduling for target searching in communication denial environment}, volume = {22}, year = {2025}, }
• (2025). Adaptive robust trajectory tracking controller for a quadrotor UAV with uncertain environment parameters based on backstepping sliding mode method. TASE, 22, 4446--4456. (www)

  @article{ , abstract = {In order to improve the trajectory tracking accuracy and environmental adaptability of a quadrotor unmanned aerial vehicle (UAV), an adaptive robust trajectory tracking controller for a quadrotor UAV with uncertain environment parameters based on Backstepping sliding mode method is proposed. This controller uses adaptive estimates to approximate the unknown parameters in the system model and designs an anti-interference link to counteract the negative impact of the external environment, which compensates the control input of the UAV rotor and improves the accuracy and stability of the quadrotor UAV trajectory tracking. Firstly, the mathematical model of the quadrotor UAV in the interference environment is established. Secondly, the tracking target of the quadrotor UAV is divided into attitude target and position target. Then, the adaptive method and Backstepping sliding mode method are used to design the update laws of the estimates and the rotor control input equations. Meanwhile, the appropriate Lyapunov function is established to verify the asymptotic stability of the UAV attitude system and position system. Finally, simulation experiments are carried out to compare the trajectory tracking performance of the quadrotor UAV with various controllers, and the effectiveness and superiority of the proposed controller are verified. Note to Practitioners—This paper mainly focuses on studying a trajectory tracking controller suitable for the dynamics of quadrotor UAVs, especially its application in resisting unknown environmental parameters and disturbances. In practice, adaptive predictive laws are used to estimate model parameter perturbations, and this controller is applied to the rotor control in yaw, pitch, and roll directions, enabling the UAV to track the established reference trajectory. Meanwhile, our backstepping sliding mode controller can effectively improve the robustness of the UAV, ensuring flight accuracy while reducing interference from environmental parameters. In flight control, it is worth studying to solve the external wind field interference and modeling deviation. Parameter prediction based on adaptive methods can effectively handle these unknown variables and prevent UAVs from oversteer in complex environments. The simulation results have verified that the proposed controller can enable the UAV to track the artificially set spiral ascent trajectory, demonstrating high application value in interference environments. In the future, we plan to design an intelligent control method that does not rely on physical modeling, rather than precise models. With this change, we will further improve the application realizability of quadrotor UAV in real world scenarios.}, archive = {J_TASE}, author = {Yuhong Hou and Dengkai Chen and Shuming Yang}, doi = {10.1109/TASE.2023.3324434}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4446--4456}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive robust trajectory tracking controller for a quadrotor UAV with uncertain environment parameters based on backstepping sliding mode method}, volume = {22}, year = {2025}, }
• (2025). Emergency UAV landing on unknown field using depth-enhanced graph structure. TASE, 22, 4434--4445. (www)

  @article{ , abstract = {With the expanding use of Unmanned Aerial Vehicles (UAVs) in a variety of applications, their safety has become a critical concern. UAVs are confronted with a variety of unforeseen circumstances during missions; in these instances, the UAVs need to autonomously locate a suitable landing site, plan flight routes, and avoid obstacles in unstructured environments. Due to the limitations of computing power and sensors, it is challenging to attain the goal. The aim of research study is to investigate the monocular emergency autonomous landing algorithm. This work concentrates specifically on extracting depth and vision information. A topology information extractor is designed to transform images into graphs and assess the connectivity of terrain. In addition, a depth information extractor is designed to compute the slope and roughness of the ground. A 3D topology optimizer is designed to optimize the graph based on depth information and evaluate the landing suitability using a heuristic strategy. For action decision making, a 3D topology decision method based on Depth-Enhanced Graph Structure (DEGS) is proposed. In order to demonstrate the efficacy of DEGS, this study constructed a simulated scenario based on an actual scene. The results of the experiment indicate that DEGS outperforms its counterparts in terms of the accuracy of action prediction and landing success rate. Note to Practitioners—Using DEGS, this study proposes a novel method for emergency UAV landing on unknown fields. The proposed method is founded on the following fundamental concepts: First, the UAV first generates a DEGS of the unknown field. Second, the UAV then evaluates the landing risk and guides the UAVs in planning a safe landing trajectory. Third, the UAV implements the landing trajectory and lands safely on the unknown field using monocular aerial vision. Frame Sequential and Self-supervision Network (FSSN), a multi-scale vision-based Monocular Depth Estimation (MDE) network, is proposed to estimate the depth map for real-time UAV flight phases. This method has been evaluated using simulations and real-world dataset of simulation images of monocular continuous frames and has shown to be effective in landing UAVs on unknown fields. A human-in-the-loop learning approach is proposed for updating DEGS with dynamic terrain classification that made the procedure feasible for unknown fields. In terms of landing success rate and action prediction accuracy, the results demonstrate that the proposed method is capable of landing a UAV on unknown terrain in a safe and efficient manner, and it is particularly useful in emergency situations where the UAV does not have prior knowledge of the field.}, archive = {J_TASE}, author = {Jie Chen and Weiming Du and Junmou Lin and Uddin Md. Borhan and Yanning Lin and Bingqing Du and Xing Li and Jianqiang Li}, doi = {10.1109/TASE.2024.3391017}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4434--4445}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Emergency UAV landing on unknown field using depth-enhanced graph structure}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Special issue on human-machine fusion decision-making for emergency handling. TASE, 22, 4427--4433. (www)

  @article{ , archive = {J_TASE}, author = {Edmond Q. Wu and Jianqiang Li and Guimin Chen and Mehmet R. Yuce and Shafiqul Islam and Javier Del Ser and Hui Yu and Peter Xiaoping Liu}, doi = {10.1109/TASE.2025.3531572}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4427--4433}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Special issue on human-machine fusion decision-making for emergency handling}, volume = {22}, year = {2025}, }
• (2025). Multi-zone HVAC control with model-based deep reinforcement learning. TASE, 22, 4408--4426. (www)

  @article{ , abstract = {The application of reinforcement learning in controlling Heating, Ventilation, and Air Conditioning (HVAC) systems has been extensively researched. Existing studies primarily focus on Model-Free Reinforcement Learning (MFRL), which involves trial-and-error interactions with real buildings to train the agent. However, MFRL encounters a significant challenge: it requires a large amount of training data to achieve satisfactory performance. While simulation models have been used to generate training data and expedite the training process, they necessitate high-fidelity building models that are difficult to calibrate. As a result, Model-Based Reinforcement Learning (MBRL) has been employed for HVAC control. Although MBRL demonstrates remarkable sample efficiency, it often falls short in terms of asymptotic control performance, particularly in achieving substantial energy savings while ensuring occupants’ thermal comfort. In this study, we conduct experiments to analyze the limitations of current MBRL-based HVAC control methods, focusing on model uncertainty and controller effectiveness. Leveraging the insights gained from these experiments, we develop MB2C, an innovative MBRL-based HVAC control system that combines high control performance with exceptional sample efficiency. MB2C learns the dynamics of the building by employing an ensemble of environment-conditioned neural networks and utilizes a novel control method called Model Predictive Path Integral (MPPI) for HVAC control. MPPI generates candidate action sequences using an importance sampling weighted algorithm, which is well-suited for multi-zone buildings with high state and action dimensions. We evaluate MB2C using EnergyPlus simulations in a five-zone office building, and the results demonstrate that MB2C achieves 8.23% higher energy savings compared to the state-of-the-art MBRL solution while maintaining comparable thermal comfort. Moreover, MB2C significantly reduces the required training data set by an order of magnitude ( $10.52\times $ ) while delivering performance on par with MFRL approaches. Note to Practitioners—Our research addresses a critical challenge in HVAC control, offering an innovative solution to enhance the data efficiency of HVAC systems while optimizing energy usage. Traditional approaches, such as Model-Free Reinforcement Learning, often require a large volume of real-world data. Our primary focus is improving the effectiveness of HVAC control, a vital aspect of building management that directly affects energy consumption and occupant well-being. We introduce MB2C, a Model-Based Reinforcement Learning system designed to significantly improve energy savings while maintaining thermal comfort. MB2C achieves remarkable results, offering exceptional sample efficiency and substantially reducing the required training data. Our research leverages an ensemble of environment-conditioned neural networks and employs Model Predictive Path Integral in HVAC control. While MB2C presents notable benefits, it also has limitations. Further research and development are required to optimize its performance across different building environments and specific use cases. Future directions should focus on addressing the safety challenges associated with real-world deployment. Beyond HVAC control, the principles and methods explored in this research have potential applications in various automation domains, such as robotics, industrial automation, and manufacturing processes.}, archive = {J_TASE}, author = {Xianzhong Ding and Alberto Cerpa and Wan Du}, doi = {10.1109/TASE.2024.3410951}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4408--4426}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-zone HVAC control with model-based deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). DGNR: Density-guided neural point rendering of large driving scenes. TASE, 22, 4394--4407. (www)

  @article{ , abstract = {Despite the recent success of Neural Radiance Field (NeRF), it is still challenging to render large-scale driving scenes with long trajectories, particularly when the rendering quality and efficiency are in high demand. Existing methods for such scenes usually involve with spatial warping, geometric supervision from zero-shot normal or depth estimation, or scene division strategies, where the synthesized views are often blurry or fail to meet the requirement of efficient rendering. To address the above challenges, this paper presents a novel framework that learns a density space from the scenes to guide the construction of a point-based renderer, dubbed as DGNR (Density-Guided Neural Rendering). In DGNR, geometric priors are no longer needed, which can be intrinsically learned from the density space through volumetric rendering. Specifically, we make use of a differentiable renderer to synthesize images from the neural density features obtained from the learned density space. A density-based fusion module and geometric regularization are proposed to optimize the density space. By conducting experiments on a widely used autonomous driving dataset, we have validated the effectiveness of DGNR in synthesizing photorealistic driving scenes and achieving real-time capable rendering. Our project page is available at https://github.com/JOP-Lee/DGNR-Rendering. Note to Practitioners—While Neural Radiance Field (NeRF) has been gaining attraction, it is still challenging to create highly detailed, efficient renderings of large driving scenes. Current methods often resort to spatial warping, geometric guidance from tools like zero-shot normal or depth estimates, or dividing the scene into smaller parts. Unfortunately, these techniques can result in blurred images or fail to meet efficiency needs. To solve these challenges, we introduce a learned density space to build a point-based renderer, termed Density-Guided Neural Rendering (DGNR). With DGNR, we no longer need geometric priors because the density space can inherently learn them through volume rendering. Specifically, we use a flexible renderer to create images from the neural density features derived from the learned density space. We have also proposed a density-based fusion module and geometric regularization to optimize the density space. We evaluated DGNR on a popular autonomous driving dataset and found it to be effective in creating realistic driving scenes and capable of real-time rendering. Project page: https://github.com/JOP-Lee/DGNR-Rendering.}, archive = {J_TASE}, author = {Zhuopeng Li and Chenming Wu and Liangjun Zhang and Jianke Zhu}, doi = {10.1109/TASE.2024.3410891}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4394--4407}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DGNR: Density-guided neural point rendering of large driving scenes}, volume = {22}, year = {2025}, }
• (2025). Adaptive fault-tolerant optimized platoon cloud tracking control for heterogeneous vehicles via dual learning mechanism. TASE, 22, 4382--4393. (www)

  @article{ , abstract = {This work investigates an adaptive fault-tolerant cloud tracking control scheme for heterogeneous vehicular platoon systems (HVPSs) based on a dual learning mechanism (DLM). A human-vehicle-road-cloud traffic scenario is built, in which the cloud control platform operates HVPSs by sending the command signal to a non-autonomous leader vehicle, which generates the required trajectory. Moreover, a DLM consisting of an improved reinforcement learning mechanism and a composite learning mechanism is proposed to optimize the control performance of the entire vehicular platoon. Furthermore, unlike existing fault-tolerant control methods, the proposed DLM can effectively reduce the adverse effects caused by actuator fault. Via the DLM, and cloud control platform, an adaptive fault-tolerant optimized platoon technology is developed to guarantee the bistability. Finally, simulation results certify the effectiveness and superiority of the developed strategy.Note to Practitioners—This paper is inspired by the problem of vehicular platoon control, but the proposed approach is not only applicable to transportation systems, but also to a variety of scenarios such as smart grids, supply chain management, and fire rescue. The intelligent transportation system aims to realize V2P, V2V, and V2I. However, there still exist several problems, such as the lack of obstacle avoidance/collision mechanism, the occurrence of platoon actuator fault, and the excessive consumption of vehicle energy, making it difficult to apply in complex traffic. In this paper, the proposed method can effectively solve the above problems, and simulation experiments have shown that this method is feasible and superior. However, vehicular platoon path programming and external network attacks have not yet been considered. To approach the real traffic environment more closely also prompts us to conduct research in the future.}, archive = {J_TASE}, author = {Jiaxin An and Liang Cao and Yingxun Wang and Ahmer Khan Jadoon and Shuhui Wang}, doi = {10.1109/TASE.2024.3410642}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4382--4393}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fault-tolerant optimized platoon cloud tracking control for heterogeneous vehicles via dual learning mechanism}, volume = {22}, year = {2025}, }
• (2025). Trilateral shared control of a dual-user haptic tele-training system for a hexapod robot with adaptive authority adjustment. TASE, 22, 4366--4381. (www)

  @article{ , abstract = {Considering the increasingly widespread use of multilegged robots in various fields, the design of their bilateral or multilateral teleoperation system is facing new challenges, such as foot slippage, tele-driving skills, and operators experience, which can induce activity of its teleoperation and poor maneuverability. With the assistance of an experienced trainer and the development of a multi-user cooperative teleoperation system, it is more conducive to tele-driving the multilegged robot in real field environment. Accordingly, in this study, a novel trilateral shared control architecture based on the velocity and force coordination of a dual-user haptic tele-training system for a hexapod robot subjected to soft terrain is proposed. Its methodology simultaneously considers the online transfer strategy of the task authority adjustment between dual users and the dynamic compensation approach for the active exogenous disturbance. In our system, the dominance factor adjusts the control authority of dual users according to the velocity and force command-tracking performance of the trilateral system. The closed-loop stability of the system is established by its passivity, and the force applied by the environment termination is determined to be approximately equal to the sum of the forces felt by two operators. The experiments of the proposed dual-user haptic training system for the hexapod robot demonstrate that it results in a stable trilateral teleoperation with a satisfactory tracking performance. Note to Practitioners—Multi-legged mobile robots have been widely used in daily and scientific scenarios, including industrial manufacturing, field exploration, and domestic service. However, as tasks and requirements become larger and more complex, the difficulty of development increases. To overcome these issues, this paper proposes a trilateral tele-training architecture for teleoperating the movement of a hexapod robot on deformable terrain, with complications such as foot slippage. The shared control method incorporates primarily two novel technical contributions, which refer to an adaptive authority adjustment method to allocate control between trainer and trainee operators, and a passivity controller (an adaptive damper) informed by observed loss in the remote robot’s velocity (both of which can be executed simultaneously). The former is implemented with a varied dominance term. Two different shared control methods are proposed, one which essentially relies on the errors in provided position inputs between the two leaders and the follower (position-error-based, PEB), and one that examines torques applied to the operators compared with environmental interaction force (direct force reflection, DFR). In addition, the practitioners who research in fields where efficient control of cooperative teleoperation for multi-legged robots subject to outdoor environment can benefit from these proposed methods. In future research, we will further address the dynamic teleoperation of hexapod robot and time delays.}, archive = {J_TASE}, author = {Jiayu Li and Chenguang Yang and Liang Ding and Bo You and Weihua Li and Xiyang Zhang and Haibo Gao}, doi = {10.1109/TASE.2024.3410522}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4366--4381}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trilateral shared control of a dual-user haptic tele-training system for a hexapod robot with adaptive authority adjustment}, volume = {22}, year = {2025}, }
• (2025). Sim2Real rope cutting with a surgical robot using vision-based reinforcement learning. TASE, 22, 4354--4365. (www)

  @article{ , abstract = {Cutting is a challenging area in the field of autonomous robotics but is especially interesting for applications such as surgery. One large challenge is the lack of simulations for cutting with surgical robots that can transfer to the real robot. In this work, we create a surgical robotic simulation of rope cutting with realistic visual and physics behavior using the da Vinci Research Kit (dVRK). We learn a cutting policy purely from simulation and sim2real transfer our learned models to real experiments by leveraging Domain Randomization. We find that cutting with surgical instruments such as the EndoWrist Round Tip Scissors comes with certain challenges such as deformations, cutting forces along the jaw, fine positioning, and tool occlusion. We overcome these challenges by designing a reward function that promotes successful cutting behavior through fine positioning of the jaws directly from image inputs. Policies are transferred using a custom sim2real pipeline based on a modular teleoperation framework for identical execution in simulation and the real robot. We achieve a 97.5% success rate in real cutting experiments with our 2D model and a 90% success rate in 3D after the sim2real transfer of our model. We showcase the need for Domain Randomization and a specialized reward function to achieve successful cutting behavior across different material conditions through optimal fine positioning. Our experiments cover varying rope thicknesses and tension levels and show that our final policy can successfully cut the rope across different scenarios by learning entirely from our simulation. Further project information is available at https://medcvr.utm.utoronto.ca/tase2024-cutrope.htmlNote to Practitioners—Cutting is one of many repetitive tasks during a surgical procedure that can be automated to reduce a surgeon’s fatigue. This paper presents an autonomous approach to cutting that can be learned from simulation. The approach builds a surgical robotic rope environment in simulation that mimics the visuals and physics of real-life rope cutting. An autonomous agent, trained with learning-based methods, uses images to position the surgical scissors and controls the jaw to perform a cut on the rope. The agent creates quality cuts by learning to finely position the jaws optimally near the scissor joint, which is learned through rewards in simulation. Agents from the simulation are transferred to the real robot setup using a custom modular ROS teleoperation framework, treating deep learning-based autonomous agents as input into a teleoperation scheme. Experiments were conducted to examine the success of the method with different material conditions and the effect of fine positioning on cut success.}, archive = {J_TASE}, author = {Mustafa Haiderbhai and Radian Gondokaryono and Andrew Wu and Lueder A. Kahrs}, doi = {10.1109/TASE.2024.3410297}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4354--4365}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sim2Real rope cutting with a surgical robot using vision-based reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). MSNet: Multiple strategy network with bidirectional fusion for detecting salient objects in RGB-D images. TASE, 22, 4341--4353. (www)

  @article{ , abstract = {Various salient object detection (SOD) approaches have been developed to identify visually attractive objects in scenes captured in RGB-D (RGB and depth) images. High-level features often provide abstract semantics, and low-level features include more details such as textures and spatial structures. Hence, effectively fusing multimodal information from different levels has become a major area of development. We propose a multiple-strategy network (MSNet) with bidirectional fusion for RGB-D SOD that incorporates multilevel feature fusion and cross-modal aggregation into a multisupervised framework. We first use a multiple-strategy fusion module to transmit high-level semantic features along a top-down progressive pathway to generate a series of appearance features. Thereafter, a self-refinement module further refines and optimizes the saliency map. Furthermore, a depth optimization module strengthens depth information extraction, especially from low-quality depth maps. Extensive experimental results on seven benchmark datasets reveal the superiority and efficacy of the proposed MSNet, compared with state-of-the-art RGB-D SOD approaches.Note to Practitioners—This study introduces a RGB-D SOD network known as Multiple-Strategy Network (MSNet) with bidirectional fusion. Initially, we employ a multiple-strategy fusion module to transmit high-level semantic features in a top-down progressive manner, generating a sequence of appearance features. Subsequently, we apply a self-refinement module to further enhance and optimize the saliency map. Additionally, we incorporate a depth optimization module to improve the extraction of depth information, particularly from low-quality depth maps.}, archive = {J_TASE}, author = {Wujie Zhou and Fan Sun and Weiwei Qiu}, doi = {10.1109/TASE.2024.3410182}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4341--4353}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MSNet: Multiple strategy network with bidirectional fusion for detecting salient objects in RGB-D images}, volume = {22}, year = {2025}, }
• (2025). Deep reinforcement learning with multiple unrelated rewards for AGV mapless navigation. TASE, 22, 4323--4340. (www)

  @article{ , abstract = {Mapless navigation for Automated Guided Vehicles (AGV) via Deep Reinforcement Learning (DRL) algorithms has attracted significantly rising attention in recent years. Collision avoidance from dynamic obstacles in unstructured environments, such as pedestrians and other vehicles, is one of the key challenges for mapless navigation. Autonomous navigation requires a policy to make decisions to optimize the path distance towards the goal but also to reduce the probability of collisions with obstacles. Mostly, the reward for AGV navigation is calculated by combining multiple reward functions for different purposes, such as encouraging the robot to move towards the goal or avoiding collisions, as a state-conditioned function. The combined reward, however, may lead to biased behaviours due to the empirically chosen weights when multiple rewards are combined and dangerous situations are misjudged. Therefore, this paper proposes a learning-based method with multiple unrelated rewards, which represent the evaluation of different behaviours respectively. The policy network, named Multi-Feature Policy Gradients (MFPG), is conducted by two separate Q networks that are constructed by two individual rewards, corresponding to goal distance shortening and collision avoidance, respectively. In addition, we also propose an auto-tuning method, named Ada-MFPG, that allows the MFPG algorithm to automatically adjust the weights for the two separate policy gradients. For collision avoidance, we present a new social norm-oriented continuous biased reward for performing specific social norm so as to reduce the probabilities of AGV collisions. By adding an offset gain to one of the reward functions, vehicles conducted by the proposed algorithm exhibited the predetermined features. The work was tested in different simulation environments under multiple scenarios with a single robot or multiple robots. The proposed MFPG method is compared with standard Deep Deterministic Policy Gradient (DDPG), the modified DDPG, SAC and TD3 with a social norm mechanism. MFPG significantly increases the success rate in robot navigation tasks compared with the DDPG. Besides, among all the benchmarking algorithms, the MFPG-based algorithms have the optimal task completion duration and lower variance compared with the baselines. The work has also been tested on real robots. Experiments on the real robots demonstrate the viability of the trained model for the real world scenarios. The learned model can be used for multi-robot mapless navigation in complex environments, such as a warehouse, that need multi-robot cooperation. Our source code and supplementary material is available at https://github.com/dornenkrone/MFPG Note to Practitioners—Autonomous navigation for AGVs in complex and large-scale environments, such as factories and warehouses, is challenging. AGVs are usually centrally controlled and depend on reliable communications. However, centralized control is not always reliable due to poor signal strengths or crashes of the server, and hence unsuitable due to the requirements of accurate information of the dynamic environments and fast responses of decision making. Therefore, it is necessary for the vehicles to perform reliable decision making based on only onboard sensors and processors, for efficient and safe autonomous navigation. Existing methods, such as simultaneous localization and mapping (SLAM) and motion planning algorithms, have been widely used. However, they are neither flexible nor generalizable enough. This paper proposes a method for autonomous navigation based on reinforcement learning (RL), which allows vehicles to gain experience through cumulative rewards by continuously interacting with the environment. The RL-based controller is designed for optimising its performance in two independent aspects, namely collision avoidance and navigation, which are quantified as separate rewards. Instead of carefully hand-crafting a combined reward, our proposed approach trains the agent using the two rewards separately to obtain one optimal policy. It is clearly easier and more practical to design the individual rewards than manually combining them. Besides, the algorithm includes a mechanism for incorporating social norms to encourage the vehicles to follow the right-hand rule, such that they can avoid pedestrians or other vehicles in a socially acceptable manner. This is achieved by adding a continuous bias on the collision avoidance reward. Experiments using simulation environments and real robots suggest that the method is generalizable to multi-robot systems, while guaranteeing safety. In future research, we will focus on incorporating uncertainties of sensor readings for safe and reliable autonomous navigation.}, archive = {J_TASE}, author = {Boliang Cai and Changyun Wei and Ze Ji}, doi = {10.1109/TASE.2024.3410162}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4323--4340}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep reinforcement learning with multiple unrelated rewards for AGV mapless navigation}, volume = {22}, year = {2025}, }
• (2025). Event-triggered guaranteed performance control of nonaffine uncertain nonlinear systems via intermittent feedback. TASE, 22, 4311--4322. (www)

  @article{ , abstract = {This paper addresses the event-triggered adaptive control problem with prescribed performance guarantees for a class of nonaffine uncertain nonlinear systems. Compared with existing works, intermittent communication in the sensor-to-controller channel for high-order nonaffine nonlinear systems is considered. Since the controller design involves intermittent variables arising from the event triggers in the sensor-to-controller channel and completely unknown nonaffine nonlinearities, traditional control schemes encounter significant challenges and usually render intricate controllers. To alleviate these complexities, a system transformation is involved to convert the system into an affine-like form, followed by a performance transformation independent of initial conditions, thus enabling the proposed method to encompass a wide range of such problems. Unlike traditional backstepping-based methods, a non-recursive control approach with an observer is employed in this work, which simplifies both controller and event-triggered mechanism design. It is theoretically proven that all signals in the closed-loop system are uniformly ultimately bounded, and the event-triggered scheme significantly reduces the communication burden while precluding the Zeno phenomenon. Simulation examples are provided to demonstrate the effectiveness of the proposed method. Note to Practitioners—This work is motivated by the control desiderata, including computational and communication resources efficiency, transient performance guarantees, and adaptability to the uncertain nonlinear, which emerge in the networked control systems, such as unmanned aerial vehicles, robot manipulators, and chemical processes. Although numerous works have been reported to solve part of these problems, there is no unified solution covering these control objectives. In this paper, we propose an event-triggered guaranteed performance control via intermittent feedback signals in a unified framework, which is applicable to a wide class of applications without controller redesign. The intermittent communication in the sensor-to-controller channel is co-designed with the adaptive controller, which provides a provable sampling scheme for the digital implementation of nonlinear networked control systems. The proposed method permits flexible performance specifications, including exponential and prescribed-time types, rather than theoretical asymptotic behavior, lending it readily to engineering purposes. Furthermore, the performance functions are designed to be independent of the initial error, eliminating the need for recalibration that hinders practical applications. The non-recursive controller with intermittent feedback and adaptive updates reduces both the communication and computation burdens, which is simple yet efficient, making our approach user-friendly and less demanding in practical applications.}, archive = {J_TASE}, author = {Jun Chen and Shuzong Xie and Wenchao Meng and Qinmin Yang}, doi = {10.1109/TASE.2024.3409760}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4311--4322}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered guaranteed performance control of nonaffine uncertain nonlinear systems via intermittent feedback}, volume = {22}, year = {2025}, }
• (2025). Cross-domain bilateral transfer learning for fault diagnosis under incomplete multisource domains. TASE, 22, 4298--4310. (www)

  @article{ , abstract = {Recently, transfer learning (TL) approaches have been extensively applied in industrial cross-domain fault diagnosis, most of which depend on the consistency assumption of the source and target fault categories. In practice, it is common to utilize multiple source domains for transfer learning, but each of them may not include all fault categories in the target domain, which are referred to as incomplete multisource domains. For the challenge of fault diagnosis under incomplete multisource domains, a cross-domain bilateral transfer learning (CDBTL) method is proposed in this article. First, a cross-domain bilateral transfer strategy is developed, where the source and target domains are reconstructed from each other and their distribution differences are reduced by minimizing the reconstruction error to avoid negative transfer. Then, for the source domain with label information, CDBTL maximizes the between-class distance of different fault categories and minimizes the within-class distance of the same fault category to ensure the discriminative nature of its feature representation. Afterwards, the common projection matrix is learned through the mutual cooperation of projection matrices between different incomplete source domains and target domain to compensate for the missing fault categories in a single source domain. The key to discriminate CDBTL from many exiting TL algorithms is that it relaxes the restriction of consistent fault categories in the source and target domains, and skillfully integrates the knowledge of multiple incomplete source domains. Extensive experiments on Tennessee Eastman process demonstrate the superiority of CDBTL in solving cross-domain fault diagnosis problem, whose accuracy is averagely improved by 17.99% compared with eleven existing algorithms.Note to Practitioners—The changing industrial operating modes (domains) may result in different data distributions and fault categories between the historical mode (source domain) and current mode (target domain). Traditional machine learning methods usually fail to diagnose under the above domain and category inconsistencies. The key work in this article is to develop a cross-domain bilateral transfer learning (CDBTL) algorithm to realize cross-domain fault diagnosis under incomplete multisource domains. The proposed algorithm can avoid negative transfer while reducing inter-domain differences, and utilize the mutual cooperation of multiple source domains to fully cover the fault categories in the target domain. The constructed CDBTL model can be combined with various classifiers, and the learned classifiers can be directly applied to fault diagnosis in advanced manufacturing industry under incomplete multisource domains.}, archive = {J_TASE}, author = {Shumei Zhang and Sijia Wang and Qi Lei and Chunhui Zhao}, doi = {10.1109/TASE.2024.3409621}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4298--4310}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cross-domain bilateral transfer learning for fault diagnosis under incomplete multisource domains}, volume = {22}, year = {2025}, }
• (2025). Optimal design of smoothed raster scan trajectory for repetitive control based high-speed atomic force microscopy imaging. TASE, 22, 4283--4297. (www)

  @article{ , abstract = {Due to the uniform sampling pattern and concise trajectory generation, the raster scan has become the mainstream scan mode for atomic force microscopes (AFMs). However, the high-frequency components of the triangular trajectory and the stair trajectory for generating the raster scan tend to motivate the lightly damped resonance of the nano-positioner insides AFMs and thus severely limit the imaging speed of AFMs. To handle this issue, a novel smoothed raster scan trajectory generation method is proposed in this article via replacing the non-imaging path of both fast axis and slow axis with an optimally designed smoothed transition trajectory. A universal mathematical expression of the smoothed raster scan trajectory is developed, following by the detailed frequency spectral analysis. The analysis results reveal that the inner model of the smoothed slow-axis trajectory is the sum of two cascaded integrators and a series of sinusoidal internal models distributed at the scanning frequency along with its harmonic frequencies. As a result, an embedded repetitive control (RC) scheme is adopted to achieve the high-bandwidth and high-accuracy tracking for both fast- and slow-axis trajectories so as to contribute to a high-speed raster scanning. Comprehensive trajectory tracking experiments and imaging experiments are performed on a commercial AFM. The experimental results show that, under the scanning frequency of 100 Hz, the root-mean-square tracking error of the proposed smoothed trajectory is reduced from 47.1 nm (fast axis) and 7.5 nm (slow axis) of the conventional trajectory to 6.9 nm (fast axis) and 3.8 nm (slow axis) with the same RC schemes, and maximum tracking error of the same smoothed slow-axis trajectory is reduced from 387.7 nm with PI scheme and 138.2 nm with PI+PDOB scheme to 13.0 nm with the RC scheme, which verify the effectiveness and advancement of the proposed trajectory smoothing method and the developed control scheme. Note to Practitioners—Raster trajectory implemented via fast- and slow-axis coordinated motion has been widely used within modern precision manufacturing and measuring systems such as the atomic force microscope, coordinate measuring machine, laser or ion-beam machine tool. The high-accuracy and high-speed tracking of the raster trajectory is the basic requirement for these sophisticated systems to achieve the intended function. However, the trajectory mutation of the raster trajectory induces large amounts of high-frequency trajectory components for each servo axis and thus brings a significant challenge for its high-performance motion control. In this article, a novel smoothed raster scan trajectory generation method is developed to reduce the high-frequency components of the conventional raster trajectory and thus facilitate its trackable performance on the basis of the frequency spectral analysis of the smoothed raster scan trajectory. The spectral analysis results reveal for the first time that the frequency spectral function of the smoothed slow-axis trajectory is the linear combination of those of two cascaded integrator and a periodic trajectory. This conclusion motivates the usage of a simple yet efficient embedded repetitive control (RC) scheme by integrating a proportional-integral (PI) controller with a repetitive controller for the high-performance tracking of both fast and slow axes whilst suppressing their cross-coupling errors without extra compensators. The effectiveness of the proposed smoothed raster scan trajectory as well as the developed embedded RC scheme have been preliminarily verified via high-speed AFM imaging. This development can also be applied on other systems involved in the high-speed tracking of the raster trajectory.}, archive = {J_TASE}, author = {Wei-Wei Huang and Xiangyuan Wang and Linlin Li and Yixuan Meng and Li-Min Zhu}, doi = {10.1109/TASE.2024.3409408}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4283--4297}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimal design of smoothed raster scan trajectory for repetitive control based high-speed atomic force microscopy imaging}, volume = {22}, year = {2025}, }
• (2025). Absolute monocular depth estimation on robotic visual and kinematics data via self-supervised learning. TASE, 22, 4269--4282. (www)

  @article{ , abstract = {Accurate estimation of absolute depth from a monocular endoscope is a fundamental task for automatic navigation systems in robotic surgery. Previous works solely rely on uni-modal data (i.e., monocular images), which can only estimate depth values arbitrarily scaled with the real world. In this paper, we present a novel framework, SADER, which explores vision and robot kinematics to estimate the high-quality absolute depth for monocular surgical scenes. To jointly learn the multi-modal data, we introduce a self-distillation based two-stage training policy in the framework. In the first stage, a boosting depth module based on vision transformer is proposed to improve the relative depth estimation network that is trained in a self-supervised method. Then, we develop an algorithm to automatically compute the scale from robot kinematics. By coupling the scale and relative depth data, pseudo absolute depth labels for all images are yielded. In the second stage, we re-train the network with 3D loss supervised by pseudo labels. To make our method generalize to different endoscopes, the learning of endoscopic intrinsics is integrated into the network. In addition, we did cadaver experiments to collect new surgical depth estimation data about robotic laparoscopy for evaluation. Experimental results on public SCARED and cadaver data demonstrate that the SADER outperforms previous state-of-art even stereo-based methods with an accuracy error under 1.90 mm, proving the feasibility of our approach to recover the absolute depth with monocular inputs. Note to Practitioners—This paper aims to solve the problem of absolute monocular depth estimation in automatic surgical navigation by leveraging the multi-modal data from the robot-based endoscopic system. Accurate depth perception with real scales of the monocular scene is essential for the control of surgical robots in automatic navigation. However, current methods can only predict the relative depth of the surgical scene using monocular images. In this article, we propose a self-supervised learning-based method to achieve high-quality absolute depth estimation of monocular endoscopic images. It neither needs manual data annotation, nor other imaging modalities. The experiments extensively validate the feasibility and high performance of our framework for absolute depth estimation on monocular endoscopes. This absolute depth perception framework can be potentially encapsulated into the automatic navigation system in the near future.}, archive = {J_TASE}, author = {Ruofeng Wei and Bin Li and Fangxun Zhong and Hangjie Mo and Qi Dou and Yun-Hui Liu and Dong Sun}, doi = {10.1109/TASE.2024.3409392}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4269--4282}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Absolute monocular depth estimation on robotic visual and kinematics data via self-supervised learning}, volume = {22}, year = {2025}, }
• (2025). Autonomous hyperspectral characterisation station: Robot aided measuring of polymer degradation. TASE, 22, 4257--4268. (www)

  @article{ , abstract = {This paper addresses a gap between the capabilities and utilisation of robotics and automation in laboratory settings and builds upon the concept of Self Driving Labs (SDL). We introduce an innovative approach to the temporal characterisation of materials. The article discusses the challenges posed by manual methods involving established laboratory equipment and presents an automated hyperspectral characterisation station. This station integrates robot-aided hyperspectral imaging (HSI), complex material characterisation modelling, and automated data analysis, offering a non-destructive and comprehensive approach. This work explains how the proposed assembly can automatically measure the half-life of biodegradable polymers with higher throughput and accuracy than manual methods. The investigation explores the effect of pH, number of average molecular weight (Mn), end groups, and blends on the degradation rate of polylactic acid (PLA). The novel contributions of the paper lie in introducing an adaptable classification station for characterisation and presenting an innovative methodology for polymer degradation rate measurements. The proposed system holds promise for expediting the development of high-throughput screening and characterisation methods within advanced material and chemistry laboratories. Note to Practitioners—The characterisation and classification of materials hold significant importance within the realms of materials science, chemistry, manufacturing and the circular economy. We introduce an innovative approach by employing robotics to automate hyperspectral imaging and processing to accelerate and improve material characterisation within a laboratory environment. This methodology facilitates the time-dependent characterisation of materials. This automated system encompasses sample manipulation and handling for hyperspectral scanning, accompanied by automated image and data processing procedures to minimise manual interventions effectively. The developed system can be seamlessly integrated into lab settings, offering objective, accurate, and automated measurement of biodegradable polymer degradation rates.}, archive = {J_TASE}, author = {Shayan Azizi and Ehsan Asadi and Shaun Howard and Benjamin W. Muir and Riley O’Shea and Alireza Bab-Hadiashar}, doi = {10.1109/TASE.2024.3409186}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4257--4268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous hyperspectral characterisation station: Robot aided measuring of polymer degradation}, volume = {22}, year = {2025}, }
• (2025). Autonomous heavy object pushing using a coaxial tiltrotor. TASE, 22, 4243--4256. (www)

  @article{ , abstract = {Aerial physical interaction (APhI) with a multirotor-based platform such as pushing a heavy object demands generation of a sufficiently large interaction force while maintaining the stability. Such requirement can cause rotor saturation, because the rotor thrust enlarged for interaction force may leave a reduced margin for attitude stabilization. We first design an H-shaped coaxial tiltrotor that can generate a sufficiently large interaction force than a conventional multirotor. We then propose an overall framework composed of high-level robust controller and low-level control allocation for the coaxial tiltrotor to ensure robustness against uncertain motion of the unknown interacting object and to overcome the saturation issue. To guarantee the robustness at all time, we design a controller based on a nonlinear disturbance observer (DOB). Then, we formulate a problem of computing low-level actuator inputs avoiding rotor saturation as a tractable nonlinear optimization problem, which can be solved real-time. The proposed framework is validated in extensive real-world experiments where the 3.3 kg tiltrotor successfully pushes a cart weighing up to 60 kg. An ablation study with the tiltrotor shows effectiveness of the proposed control allocation law in avoiding rotor saturation. Furthermore, a comparative experiment with a conventional multirotor shows failure in the same setting, which validates the use of the coaxial tiltrotor. An experimental video can be found at https://youtu.be/Gdmcmoz_UjU Note to Practitioners—The motivation of this work is to enable heavy object manipulation, especially pushing operation, using a multirotor platform. To push a heavy object, which has not been treated in existing works, is challenging due to the presence of a large unknown interaction wrench and rotor saturation issue. To resolve this problem, we first build an H-shaped coaxial tiltrotor that can effectively utilize rotor thrusts to generate horizontal interaction force. Then, a robust controller is designed to address uncertainty in the interaction wrench for the tiltrotor. Finally, we propose an optimization-based control allocation to avoid rotor saturation. Real-world experiments including heavy object pushing and ablation study with respect to the control allocation law are conducted to show the effectiveness of the proposed framework. We validate the use of the coaxial tiltrotor in additional comparative experiments with a conventional multirotor which fails in pushing a 60 kg cart. The proposed framework can be utilized in various applications such as disaster relief operations, emergency rescue and aerial delivery.}, archive = {J_TASE}, author = {Sunwoo Hwang and Dongjae Lee and Changhyeon Kim and Hyoun Jin Kim}, doi = {10.1109/TASE.2024.3409058}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4243--4256}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous heavy object pushing using a coaxial tiltrotor}, volume = {22}, year = {2025}, }
• (2025). A static modeling method considering friction for the shape prediction or force sensing of concentric cable-driven manipulators. TASE, 22, 4231--4242. (www)

  @article{ , abstract = {Owing to the flexibility and redundancy, concentric cable-driven manipulators are widely used in confined space applications. However, it is challenging to establish shape prediction or force sensing for this type of manipulator. This paper proposes a static modeling method to address the shape prediction or force sensing challenges of concentric cable-driven manipulators. The Newton-Euler method is adopted to establish the static model of the concentric cable-driven manipulator with consideration of friction. On the one hand, the static model can accurately predict the shape when the external load and cable tension of the concentric cable-driven manipulator are known. On the other hand, the static model can also assist force sensing when the cable tension and shape of the concentric cable-driven manipulator are known. Finally, the accuracy of the shape prediction and force sensing of this model are both verified by experiment, separately. The results show that the accuracy of shape prediction is 95% and the accuracy of force sensing is 90%. The model is thus suitable for the shape prediction or force sensing of the concentric cable-driven manipulator. In addition, the errors associated with shape prediction or force sensing were found to be significantly reduced when the influence of friction was taken into account. Therefore, the proposed static modeling method considering friction is effective for the shape prediction or force sensing of concentric cable-driven manipulators. Note to Practitioners—Concentric cable-driven manipulators featured by small size and high flexibility are excellent candidates for operations in confined environments. This paper proposes a static modeling method to address the shape prediction or force sensing challenges of concentric cable-driven manipulators. The shape prediction can be used to facilitate the accurate positioning and control of such manipulators in various situations. The force sensing can be used to avoid unnecessary harm within the working environment of such manipulators. It is beneficial to promote the application of such manipulators in minimally invasive surgery, pharyngeal swab sampling, and other fields. Moreover, the static method applies not only to the specific model of a concentric cable-driven manipulator but also to various cable-driven manipulators. Similarly, the importance of the static model is not only that it improves the accuracy of shape prediction or force sensing but also that it can be applied to automation control of motion, flexibility, and stiffness.}, archive = {J_TASE}, author = {Zonggao Mu and Yuming Gao and Weitao Li and Guikun Lv and Shun Zhao and Ruichun Dong}, doi = {10.1109/TASE.2024.3408916}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4231--4242}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A static modeling method considering friction for the shape prediction or force sensing of concentric cable-driven manipulators}, volume = {22}, year = {2025}, }
• (2025). Strategic two-stage diagonal quadratic approximation method for economic dispatch with energy storage. TASE, 22, 4216--4230. (www)

  @article{ , abstract = {Solving large-scale stochastic dynamic economic dispatch problems involving energy storage over multiple time periods is inherently challenging. Decomposing the optimization problem across time scales to achieve parallel optimization is an effective strategy for enhancing computational efficiency. However, the introduction of energy storage complicates this task significantly due to its non-convex characteristics. To address this challenge, this paper presents a novel strategic two-stage Diagonal Quadratic Approximation Method (DQAM) that transforms the original problem into a two-stage structure amenable to parallel solving. In this structure, each sub-problem focuses on a single time period, effectively handling the mixed-integer and strong temporal coupling characteristics of energy storage. This extension broadens the applicability of the original DQAM to non-convex problems. Numerical results under various conditions demonstrate the effectiveness and performance of the proposed two-stage DQAM, achieving significantly improved solution efficiency while maintaining good solution accuracy. Leveraging computational processor and memory resources, the proposed method can achieve a maximum enhancement of 36.86 times in computational efficiency on large-scale systems with multiple time periods. Note to Practitioners—With the increasing scale of power systems and the growing penetration of renewable energy, the accuracy and real-time requirements for economic dispatch solutions have been continuously enhanced. Solving large-scale optimization problems over multiple time periods in power grids has become more challenging, especially with the additional computational burden introduced by the deployment of energy storage to mitigate renewable energy fluctuations. The configuration of energy storage generates a large number of binary decision variables, which results in the original problem becoming a large-scale mixed-integer programming model which is difficult to solve. This paper proposes a strategic two-stage DQAM to solve the stochastic dynamic economic dispatch problem with energy storage, utilizing a two-stage structure to determine an approximate energy storage optimization schedule through an optimization approximation function. Subsequently, the equivalent dynamic economic dispatch problem is transformed into a sizable form with smaller and more easily solvable sub-problems. By relaxing the temporal coupling constraint, the computational efficiency is significantly improved. Test results on systems of various scales and time periods demonstrate that the two-stage DQAM reduces computation time in all scenarios.}, archive = {J_TASE}, author = {Ziyu Zhang and Tao Ding and Chenggang Mu and Yuhan Huang and Jinbo Liu and Yishen Wang}, doi = {10.1109/TASE.2024.3408893}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4216--4230}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Strategic two-stage diagonal quadratic approximation method for economic dispatch with energy storage}, volume = {22}, year = {2025}, }
• (2025). Event-based distributed bipartite consensus of high-order nonlinear multi-agent network: Applications to hypersonic flight vehicles. TASE, 22, 4206--4215. (www)

  @article{ , abstract = {This paper endeavors to address the output-feedback bipartite consensus issue for uncertain high-order multi-agent systems (MASs) through the utilization of a dynamic event-triggered control methodology. To achieve this goal, a novel output-feedback distributed control strategy is developed for each follower agent. This strategy utilizes a bounded time-varying gain along with a predefined-time compensator. This combined approach facilitates the achievement of bipartite consensus within a specified timeframe. The key highlight of this methodology lies in the introduction of a newly devised time-varying gain, which not only enhances the convergence rate but also significantly improves transient performance. Additionally, a new dynamic event-triggered mechanism is developed for each agent, leveraging the newly introduced predefined-time internal dynamic variable. This mechanism ensures the achievement of predefined-time bipartite consensus while significantly reducing communication frequency. Comparative simulations are conducted to validate the superiority of the obtained results. Note to Practitioners—For real physical agent systems, it is impractical to gather all the state information. Hence, there is a critical need to investigate output feedback consensus algorithms that eliminate reliance on state information. However, existing output feedback consensus algorithms require the continuous transmission of control instructions to actuators, resulting in high communication costs and transmission burden. Additionally, in practical requirements, designers are eager for multiple agents to achieve the required stable performance at a predefined time rather than infinite time. For this reason, this paper presents a new output-feedback event-triggered control methodology to achieve a predefined-time bipartite consensus of uncertain nonlinear MASs. To improve the system’s convergence speed and transient performance, we introduce a novel time-varying gain. Next, we devise a dynamic event-triggering control for each agent utilizing the newly introduced predefined-time internal dynamic variable. This approach effectively reduces communication frequency while ensuring predefined-time bipartite consensus.}, archive = {J_TASE}, author = {Maolong Lv and Choon Ki Ahn and Ping Wang and Renwei Zuo and Jinde Cao}, doi = {10.1109/TASE.2024.3408650}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4206--4215}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based distributed bipartite consensus of high-order nonlinear multi-agent network: Applications to hypersonic flight vehicles}, volume = {22}, year = {2025}, }
• (2025). Adaptive prescribed-time consensus tracking control scheme of nonlinear multi-agent systems under deception attacks. TASE, 22, 4196--4205. (www)

  @article{ , abstract = {This article investigates the adaptive prescribed-time consensus tracking control problem for nonlinear multi-agent systems (MASs), where the states of systems are unmeasured and the actuators suffer from the deception attacks. Firstly, a novel coordinate transformation technology is developed by introducing a time-varying constraint function, such that the prescribed-time tracking control problem of nonlinear MASs is converted into the constraint problem of the error variables. Then, a new attack compensator is proposed to address the unknown time-varying attack gains caused by the actuator deception attacks. Further, the state observers are designed to estimate the unavailable state variables and fuzzy-logic systems (FLSs) are employed to handle the unknown functions that exist within the systems. In addition, the attack compensator-based controller ensures the boundedness of all signals, while the error variables converge to the predefined region in a specified time. The upper bound of the whole tracking errors in the mean square sense can be decreased by selecting the appropriate design parameters. At last, the simulation example illustrates the availability of the developed control method. Note to Practitioners—In the industry, consensus tracking control of nonlinear MASs exists in many different systems, such as mobile robot networks, intelligent transportation management, surveillance and monitoring. Since the above systems operate in a network environment, the security problems of the systems cannot be ignored. Hence, considering the unmeasured states, the unknown functions, and the unknown time-varying attack gains existing simultaneously in the studied systems, it is a challenging and meaningful task to achieve the desired security control objectives. On the other hand, based on a time-varying constraint function, this article presents an adaptive prescribed-time consensus tracking control scheme for the nonlinear MASs under the deception attacks. It provides a viable strategy for industrial applications.}, archive = {J_TASE}, author = {Ben Niu and Yahui Gao and Guangju Zhang and Xudong Zhao and Huanqing Wang and Ding Wang and Chao Liu}, doi = {10.1109/TASE.2024.3408453}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4196--4205}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive prescribed-time consensus tracking control scheme of nonlinear multi-agent systems under deception attacks}, volume = {22}, year = {2025}, }
• (2025). Distributed fault-tolerant control of nonlinear multiagent systems with generally uncertain semi-markovian switching topologies. TASE, 22, 4180--4195. (www)

  @article{ , abstract = {This paper centers on the distributed fault-tolerant control (DFTC) issues of time-varying delayed nonlinear multiagent systems (TVDNMASs) with switching topologies and external disturbances by considering multiple faults and event-triggered consensus strategy (ETCS). The switching topologies satisfy generally uncertain semi-Markovian switching topologies (GUSMSTs), and they contain uncertain and partially unknown semi-Markovian transition rates (TRs). In addition, the ETCS is adopted in this paper to decide the update of controllers, which alleviates the load of the correspondence network. In view of Lyapunov-Krasovskii functional (LKF), the tracking control protocols are presented to guarantee the DFTC of nonlinear multiagent systems (MASs). Moreover, the controller gain and observer gain matrices are derived through the solution of linear matrix inequalities (LMIs). Finally, a simulation example is proposed to exhibit the capability of our design technique. Note to Practitioners—Due to the complexity of engineering environment, the cooperative control of MASs has gained widespread attention. Nowadays, the MASs are generally utilized in diverse fields, such as multi-motor synchronization, drone swarm formation, and smart grids. As one of the significant research interests in cooperative control, the consensus control of MASs has become a research hotspot. However, in practical applications, due to stochastic system failures and sudden changes in the external environment, it is hard for the fixed communication topologies to cope with these unexpected situations. Therefore, the DFTC issues of delayed nonlinear MASs with GUSMSTs and external disturbances by considering multiple faults are investigated in this paper. Moreover, the mode-dependent distributed time-delay intermediate observers and active fault-tolerant consensus controllers are designed on the basis of distributed fault-tolerant control consensus protocol.}, archive = {J_TASE}, author = {Junyi Wang and Zhonglin Gui and Jiayue Sun and Xiangpeng Xie and Qinggang Meng}, doi = {10.1109/TASE.2024.3408294}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4180--4195}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed fault-tolerant control of nonlinear multiagent systems with generally uncertain semi-markovian switching topologies}, volume = {22}, year = {2025}, }
• (2025). Cooperative output regulation of MASs with a novel parameter convergence condition and prescribed-time event-triggered observers and its application to robot formation. TASE, 22, 4168--4179. (www)

  @article{ , abstract = {An enhanced control scheme for multiagent cooperative output regulation (COR) is presented in this paper. The proposed scheme incorporates prescribed-time observers and controllers with concurrent learning adaptive laws. By introducing prescribed-time event-triggered (ET) observers, the estimation of exosystem information can be achieved within a prescribed time, eliminating continuous communication between followers and avoiding reliance on global information. Additionally, Zeno behavior is excluded. The convergence of controller parameters is ensured by the established concurrent learning adaptive laws. Specifically, the challenge of designing correlation filter signals caused by the lack of a matching condition between the exosystem matrix and the follower system matrix, essential for concurrent learning, is addressed by employing the regulator equation. The desired relationship between two sets of developed filtered signals is ensured by the developed prescribed-time observers after a prescribed time. This establishes the basis for a concurrent learning strategy. Moreover, a cooperative controller with cooperative adaptive laws is developed, effectively addressing the difficulty arising from the mismatch between the matrix estimation of the exosystem and the regulator equation. Compared with existing studies, the developed controller can accurately identify controller parameters, resulting in better system performance. The successful application of the cooperative finite-time excitation (CFTE) condition relaxes the widely used persistent excitation (PE) and cooperative persistent excitation (CPE) conditions in parameter identification. Finally, a multi-robot formation simulation verifies the effectiveness of the developed method. Note to Practitioners—The COR problem explored in this paper can be applied to address the actual robot formation issue, utilizing coordinate transformation. Unlike existing studies, this paper employs easily satisfied cooperative persistent excitation conditions, making them readily applicable to real-world systems. The developed approach enables the convergence of the designed controller parameters to their true values, achieving superior control effects. Moreover, this paper overcomes some current research difficulties. It successfully integrates the most advanced event-triggered mechanism (ETM) into the prescribed-time observer, realizing the prescribed-time observation of the exosystem. Uniquely, the method presented here does not depend on global information, making it suitable for complex large-scale multiagent connections and the Internet of Things. It holds significant guidance for practical multiagent networked systems. The effectiveness of the proposed method is demonstrated through a nonholonomic dynamic robot formation problem. It is anticipated to be scalable to multiagent systems (MASs) with broader practical constraints, finding applications in more realistic engineering scenarios.}, archive = {J_TASE}, author = {Qinghua Hou and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3408161}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4168--4179}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative output regulation of MASs with a novel parameter convergence condition and prescribed-time event-triggered observers and its application to robot formation}, volume = {22}, year = {2025}, }
• (2025). Finite-time stabilization of nonlinear systems via hybrid control and application to state estimation of complex networks. TASE, 22, 4154--4167. (www)

  @article{ , abstract = {This paper investigates the problems of finite-time stabilization (FTS) and finite-time contractive stabilization (FTCS) for nonlinear systems by designing a novel class of hybrid control consisting of aperiodically intermittent control (APIC) and impulsive control (IC). Different from the single APIC, our proposed hybrid control specifically introduces IC on the intervals without control input in APIC. It shows that IC in hybrid control may not only flexibly shorten the length of the intervals with continuous control input, but also alleviate the design pressure of gain of APIC. Based on the hybrid control, some Lyapunov-based criteria are derived to guarantee the FTS and FTCS, where a potential relationship between system structure, APIC law, and impulse actions is established. Specifically, some linear matrix inequalities (LMIs) based criteria are presented to design the hybrid control gains. As an application, the theoretical results are extended to the finite-time state estimation of complex networks involving the unavailable network states. With the available measurement outputs, the hybrid state estimator involving both APIC and IC is proposed. Finally, three numerical examples are provided to illustrate the effectiveness of our results. Note to Practitioners—Unlike the Lyapunov stability results, the transient performances described by FTS and FTCS can be quantitatively guaranteed in the finite-time sense and have better applicability for the practical engineering requirements. This paper was motivated to solve FTS and FTCS problems. To guarantee the FTS and FTCS, a novel class of hybrid control is designed, which integrates the advantages of APIC and IC. Compared with the single APIC, the introduction of IC may reduce the dependence on the continuous control input in APIC, which significantly reduces the energy consumption of communication. The introduction of IC can also alleviate the design pressure of APIC to some extent. Hence, some easy-checked FTS and FTCS criteria are presented based on the proposed hybrid control. In practical applications, the theoretical results are extended to investigate the finite-time state estimation of complex networks considering unavailable network states. Utilizing the available measurement outputs, the corresponding hybrid state estimator consisting of APIC and IC is constructed. Finally, note that the impulsive transmission of ball motion is considered to further illustrate the practicability of our proposed results. In the future work, our proposed hybrid control strategy can be applied in other practical projects, such as the finite-time control of omnidirectional mobile robot.}, archive = {J_TASE}, author = {Luyao You and Shuchen Wu and Xiaodi Li}, doi = {10.1109/TASE.2024.3408105}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4154--4167}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time stabilization of nonlinear systems via hybrid control and application to state estimation of complex networks}, volume = {22}, year = {2025}, }
• (2025). Stabilization of a class of lipschitz nonlinear systems through an event-triggered impulsive controller. TASE, 22, 4145--4153. (www)

  @article{ , abstract = {In this paper, the stabilization problem is addressed for a class of globally Lipschitz nonlinear dynamical systems. An event-triggered impulsive control (ETIC) method is introduced to establish asymptotic stability and exponential stability criteria for the resulting closed-loop systems without and with time delay, respectively. Compared with the extant results, a new type of switching and event-triggered impulsive systems are considered, and some general asymptotic stability and exponential stability conditions are provided. It is proved that Zeno behaviors can be excluded and the control frequency can be appropriately regulated through the appropriate selection of event parameters. Finally, an application example of the Chua’s circuit is given to verify the efficiency of the approach and theoretical analyses. Note to Practitioners—Since impulsive control provides an easier and cheaper strategy than continuous control, it has been widely applied to many practical situations, such as control of satellite rendezvous, synchronization of permanent magnet synchronous motors, consensus of multi-agent systems and so on. Different from the time-triggered impulsive control methods, the proposed ETIC method can flexibly regulate the control frequency, which meets the increasing requirement for reducing unnecessary waste of communication resources in modern industries. Compared with the extant results, an ETIC method is established with time-varying threshold parameters and event interval parameters in this paper. Moreover, the effects of event parameters and time delay on convergence rate are analyzed, which can provide reference for improving system performance in practical applications.}, archive = {J_TASE}, author = {Zidong Ai and Guangdeng Zong and Wei Xing Zheng}, doi = {10.1109/TASE.2024.3407956}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4145--4153}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stabilization of a class of lipschitz nonlinear systems through an event-triggered impulsive controller}, volume = {22}, year = {2025}, }
• (2025). Robotic microscopic vision-free cell transportation based on pressure variation model inside bilayer micropipette. TASE, 22, 4134--4144. (www)

  @article{ , abstract = {Cell transportation is a key step in many biological applications. Currently, most cell transportation research relies on microscopic vision feedback, which limits its application at places when a microscopic view is unavailable. For the first time, this paper proposed a robotic microscopic view-free cell transportation method based on pressure variation model inside a self-made bilayer micropipette. The bilayer micropipette, made by inserting a thin micropipette into a thicker one, picks the target cell with the outsider micropipette and holds it with the inner micropipette in cell transportation. To achieve this, the appropriate fluidic forces to pick and place the cell were determined based on force analysis. Then, the appropriate inner diameter of the inner thin micropipette was determined based on the micropipette aspiration model. Further, the pressure variation inside the bilayer micropipette during holding and releasing process were modeled to conduct a robotic microscopic vision-fee cell transportation. Experimental results demonstrated that our system was capable of transporting zebrafish embryos at an average speed of 25 s/cell with a success rate of 90%, with a transportation efficiency comparable to the related methods with microscopic view feedback. Besides, ignorable mechanical harm to the development competence of the zebrafish embryos was confirmed through culturing experiments. Note to Practitioners—Cell transportation is an important step in many biomedical researches. Currently, cell transportation usually relies on microscopy for picking up, and releasing cells, limiting its application in special environments without microscopic view feedback. In this paper, a robotic microscopic vision-free cell transportation method is proposed for the first time. In this method, cells are transported within a self-made bilayer micropipette. Pressure variations inside the micropipette are utilized to determine cell picking up and release. This robotic transport method demonstrated a high success rate and low harm to cells in experiments. Our research may contribute to the future development of vision-free and fully automated robotic cell culture techniques. Our research may spike inspiration for future research on microscopic vision-free cell transportation in fully automated cell culture system.}, archive = {J_TASE}, author = {Jinyu Qiu and Ripeng Zhu and Yidi Zhang and Minghui Li and Ruimin Li and Shaojie Fu and Xin Zhao and Qili Zhao}, doi = {10.1109/TASE.2024.3407841}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4134--4144}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robotic microscopic vision-free cell transportation based on pressure variation model inside bilayer micropipette}, volume = {22}, year = {2025}, }
• (2025). Data-driven control for magnetic actuation capsule: Dynamic compensation and input constraints. TASE, 22, 4121--4133. (www)

  @article{ , abstract = {In the magnetic actuation system, MAC (Magnetically Actuated Capsule) motion is disturbed by gastrointestinal resistance, and dynamic constraints exist between MAC and EPM (External Permanent Magnet), leading to control difficulties. This paper presents a data-driven control method for magnetic actuation system. Firstly, a data-driven modeling approach is proposed for addressing the modeling challenges, relying solely on MAC visual positioning input and EPM position output data. Secondly, To effectively track rapidly changing MAC desired trajectories, determining the upper bound of system inputs through analysis of the magnetic field relationship, and integrating it with adaptive parameter reset conditions, enables the establishment of dynamic constraints for the MAC system. Finally, dynamic compensation is applied to account for non-linear resistance terms and inaccuracies in data model representation. A dual-visual positioning experimental platform simulating the gastrointestinal environment is established to validate the proposed algorithm’s effectiveness in MAC trajectory tracking under different conditions. Note to Practitioners—This article aims to design a data-driven method that incorporates dynamic constraint relationships between MAC and EPM, enabling capsules to move rapidly within the gastrointestinal system. This paper aims to increase control speed, enabling the energy to be primarily used for capturing the lesion area rather than during motion, and improving MAC tracking for desired trajectory accuracy. To validate the algorithm’s effectiveness under various resistances and driving forces, a dual-camera positioning system is designed and tested on a simulated gastrointestinal platform. In future work, we plan to replace the external camera with an internal capsule camera for MAC localization and design a medical system that actively utilizes patient lesion data to drive WCE for examinations.}, archive = {J_TASE}, author = {Peng Chen and Xiongxiong He and Jinhui Zhu and Sheng Li}, doi = {10.1109/TASE.2024.3407825}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4121--4133}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven control for magnetic actuation capsule: Dynamic compensation and input constraints}, volume = {22}, year = {2025}, }
• (2025). An online growing-and-pruning algorithm of a feedforward neural network for nonlinear systems modeling. TASE, 22, 4109--4120. (www)

  @article{ , abstract = {In recent decades, many researchers and practitioners have always been focusing on the optimal design approach of feedforward neural network (FNN). However, it is still a challenge that FNN has a compact structure, while meeting some special requirements. This paper investigates the interrelation of nodes and the diversity of samples in sliding windows based on previous work for nonlinear system modeling and develops a new adaptive growing-and-pruning algorithm of FNN (AGPA-FNN). More specifically, the basis of AGPA-FNN is to dynamically grasp the interactions between hidden nodes through the local sensitivity analysis and mutual information method, which can seamlessly prune the redundancy hidden nodes through the punishing mechanism for weight decay. Thus, AGPA-FNN can prevent breaking the network structure and forget the acquired knowledge due to the sudden change in weights. In learning algorithm, an effective learning technique is developed to accelerate the learning efficiency of the gradient descent method (AOGD) through keeping the diversity of samples and a suitable learning rate in each window. Experimental investigations show that the proposed AGPA-FNN can effectively self-adjust hidden nodes based on data changes and finally achieve a compact structure, while it can outperform other approaches in terms of generalization performance. Note to Practitioners—Nonlinear system modeling has always been a key and difficult problem in real-world. In recent decades, Neural network has become an indispensable technique of nonlinear system modeling due its universal approximation property and strong modeling ability. However, finding the trade-off between the network structure and the generalization capability to meet some prespecified requirement for neural network has always been the fundamental challenge. The existing methods are always neglecting the interrelation between the pruned neurons and other neurons in pruning process, and usually compensate for the weight of other neurons after pruning redundancy neurons, which may seriously disrupt the stability of feedforward neural network (FNN). Therefore, this paper focuses on simultaneously self-organizing and self-updating both the structure and parameters of FNN for data streams, and develops a new adaptive growing-and-pruning algorithm of FNN (AGPA-FNN) in a bid to make a trade-off between the network structure and property. AGPA-FNN can dynamically self-organize its structure and seamlessly prune the redundancy hidden nodes through the punishing mechanism for weight decay based on the interactions between nodes using the local sensitivity analysis and mutual information method. Furthermore, the online gradient descent algorithm is improved to enhance the learning efficiency of FNN by keeping the diversity of samples and the suitable learning rate in each window. Finally, the proposed AGPA-FNN is validated by several benchmark and real-word problems.}, archive = {J_TASE}, author = {Xin Guo and Wei-Sheng Wang and Jie Zhang and Li-Shuang Gong}, doi = {10.1109/TASE.2024.3407518}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4109--4120}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An online growing-and-pruning algorithm of a feedforward neural network for nonlinear systems modeling}, volume = {22}, year = {2025}, }
• (2025). A failure tree model for cascading failure in power grid with uncertain renewable energy generation. TASE, 22, 4092--4108. (www)

  @article{ , abstract = {The increasing penetration of renewable energy generation (REG) introduces high levels of uncertainty into power grid, potentially causing significant impacts on the evolution of cascading failure. In this paper, we propose a failure tree model that encompasses all possible failure paths resulting from the uncertain power injections from REG to describe the dynamic process of cascading failure in power grid. In order to obtain the failure paths of cascading failure, we propose an interval overload tripping mechanism to model relay protection based on the uncertainty set of REG and dynamic interval power flow. On the basis of the proposed model, we design a forward-backward tree search to efficiently evaluate the impact of the uncertain REG on cascading failure. Compared with the probabilistic power flow (PPF) model and scenario-based model, the simulation results of our model are more accurate because the statistical distribution of demand loss in our model is closer to Monte Carlo simulation (MCS). The efficiency of the proposed simulation method is demonstrated by comparing our model with the MCS under various sample numbers and two existing models. Finally, we analyze the influence of REG uncertainty level and penetration level on cascading failure and simulation performance. Note to Practitioners—To achieve accurate and fast cascading failure analysis in power grid with renewable energy generation (REG), this paper develops a failure tree model that considers the impact of uncertain injected power of REG on the dynamic process of cascading failure. In the model, the dynamic interval power flow and interval overload tripping mechanism are proposed to simulate the physical responses during cascading failure, including power flow redistribution, transmission branch outage and frequency regulation. Therefore, the model is more accurate in describing the actual characteristics of cascading failure in power grid with REG. This will facilitate the development and evaluation of control strategies aimed at improving the stability of power grid. Meanwhile, the model provides a good example for researchers and engineers to simulate network systems without detailed information about the probability distribution of uncertain injection variables. Based on the proposed model, we develop a forward-backward tree search, which allows the decision-maker to make a satisfactory trade-off between accuracy and time consumption. This algorithm allows for fast control strategy implementation to prevent failure propagation.}, archive = {J_TASE}, author = {Yujie Yang and Yadong Zhou and Jiang Wu and Zhanbo Xu and Sizhe He and Ting Liu and Xiaohong Guan}, doi = {10.1109/TASE.2024.3407203}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4092--4108}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A failure tree model for cascading failure in power grid with uncertain renewable energy generation}, volume = {22}, year = {2025}, }
• (2025). Toward air operation aerial manipulator control with a refined anti-disturbance architecture. TASE, 22, 4076--4091. (www)

  @article{ , abstract = {Due to the presence of strong inner dynamic coupling and changes in center of mass (CoM) during tasks execution, the precise tracking control problem of aerial manipulator systems becomes challenging. When the mounted manipulator is performing a task, the movements of the manipulator will cause the inherent dynamic coupling force/torque disturbances. Such disturbances are quickly acted on the position loop and the orientation loop of the UAV body, to the detriment of the control accuracy. On the other hand, the fluctuation of the UAV body leads to the base floating, resulting in a significantly adverse influence on the precision of the manipulator end-effector. Since different disturbances have distinct mathematical properties, i.e., norm bounded and rate bounded, currently, there is no control framework that is able to tackle the dynamic coupling, model uncertainty and base floating simultaneously. In this paper, a refined anti-disturbance control architecture is proposed for the decentralized aerial manipulator model, where various disturbances with different mathematical properties are well explored and tackled according to their positions and effects acting on the system. The stability of the proposed control framework is ensured by using Lyapunov-like analysis. Experimental results are presented to illustrate the performance of the proposed control framework. Note to Practitioners—One of the key challenges that hinder the aerial manipulator potential applications is its stability and accuracy due to the existence of various disturbances. Most of disturbance rejection control methods in the literature for aerial manipulator always deal with the various disturbances as the lumped one. Their distinct mathematical properties and different impacts on the system are not well explored, which may result in the performance degradation and limit its practical implementation. In this article, a refine anti-disturbance architecture is proposed, which is able to handle various disturbances in a more systematical way according to their mathematical properties and effects acting on the system. Physical experiments suggest that finely tackling the different disturbances enjoys a better performance in aerial manipulator trajectory tracking control problem and thus can promote the aerial manipulator to be deployed in the tasks demanding on the accuracy. In addition, such control strategy can also be extended to other robotic systems suffering from various disturbances.}, archive = {J_TASE}, author = {Shangke Lyu and Yu Zhang and Jianliang Wang and Chien Chern Cheah and Xiang Yu}, doi = {10.1109/TASE.2024.3407191}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4076--4091}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward air operation aerial manipulator control with a refined anti-disturbance architecture}, volume = {22}, year = {2025}, }
• (2025). Vision-based closed-loop control with spatiotemporal multiplexing strategy for noncontact trapping of multiple micro-particles. TASE, 22, 4063--4075. (www)

  @article{ , abstract = {Noncontact trapping of micro objects has great application potential in fields like material science and biomedical engineering due to its label-freeness and biocompatibility. In this paper, an automated acoustic micro-particle trapping system implemented with phased transducer array (PTA) is prototyped. The system is incorporated with a stereo vision to provide visual feedback benefited from localization of the invisible acoustic field through hydrophone scanning. Binocular vision calibration and stereo matching are realized using image Jacobian matrix. An efficient phase modulation algorithm is proposed for the calculation of desired PTA phase profile in real-time and a spatiotemporal multiplexing control strategy is adopted to dynamically generate multiple trappings. Experimental results well demonstrated that the stable trapping of multiple particles can be robustly realized by the system, leading to the improvements of robotic noncontact manipulation with invisible acoustic end-effector. Note to Practitioners—This paper is motivated by the problem that previous classic acoustic trapping was achieved as a physical phenomenon that particles within the trapping zone would be automatically trapped and thus required people to place the particle into the invisible trapping zone, which is neither precision nor efficient. Such problem is a crucial factor that limits acoustic tweezer to be further readily usable in bioengineering, surface manufacturing, and quantitative micromechanical characterization. In this work, automated acoustic trapping is presented in the context of robotics, as grasping task in conventional industrial robots, that can generate the acoustic trap exactly in the location where particles are detected (by microscopic vision, or micro-CT or acoustic imaging, etc.). This paper proposes a full pipeline to automatically trap multiple particles using ultrasonic transducer array and binocular microscopic vision. The experiments verified the ability of proposed method in simultaneously trapping three micro particles with opposite acoustic properties. Such trapping method is the foundational technology for further acoustic manipulation such as arraying and sorting, which will be the tasks in our future work.}, archive = {J_TASE}, author = {Mingyue Wang and Jiaqi Li and Chengxi Zhong and Teng Li and Zhenhuan Sun and You-Fu Li and Hu Su and Song Liu}, doi = {10.1109/TASE.2024.3407144}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4063--4075}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Vision-based closed-loop control with spatiotemporal multiplexing strategy for noncontact trapping of multiple micro-particles}, volume = {22}, year = {2025}, }
• (2025). Multi-scale alignment domain adaptation for ship classification in multi-resolution SAR images. TASE, 22, 4051--4062. (www)

  @article{ , abstract = {Synthetic aperture radar (SAR) images obtained from multi-sensor systems usually exhibit significant shift in data distribution, known as the domain shift. It is challenging to utilize the relevant knowledge of multi-sensor datasets for SAR image cross-domain classification with general supervised learning methods. While domain adaptation (DA) methods can alleviate the domain shift by aligning distributions, they are limited in handling the scale/resolution variations observed in multi-sensor SAR images. These methods mainly focus on feature representations at a single scale for distribution alignment, which may fail to fully align the distributions across different scales. To solve this problem, we propose a new multi-scale alignment domain adaptation network (MSADAN) for SAR ship cross-domain classification. MSADAN explicitly considers the scale factor, enabling us to overcome the weak generalization observed in existing DA methods when dealing with SAR images exhibiting significant scale/resolution variations. Specifically, we develop a scale-aware feature extractor to effectively capture the multi-scale information present in the datasets, facilitating comprehensive representation learning. Furthermore, we propose a multi-level bilinear fusion (MLBF)-based adversarial learning strategy to overcome the limitations of single-scale feature extraction for distribution alignment, aiming to enhance the generalization ability of the model across domains. In addition, a customized class contrastive loss is designed to improve the inter-class separability and intra-class compactness by penalizing cross-class confusion. Experimental results on datasets demonstrate the superiority of MSADAN in SAR ship cross-domain classification. Note to Practitioners—Ship classification using Synthetic aperture radar (SAR) images proves to be a promising strategy in modern maritime monitoring systems. The primary motivation of this paper is to develop a cross-domain classification system tailored for SAR ship target. In real-world scenarios, SAR data often presents significant resolution variations due to the different imaging modes and conditions across multiple sources. Existing DA methods fail to effectively address this unique challenge in the remote sensing field, resulting in poor cross-domain classification performance. The proposed method considers the scale or resolution variations of the targets during feature learning. Furthermore, explicitly incorporating resolution into the distribution alignment enhances the generalization ability of the model to target tasks when dealing with the significant scale or resolution variations across datasets. Experimental results confirm the practicality and robustness of our SAR ship cross-domain classification method in real scenarios. This is of significant importance in promoting the application of machine learning approaches in practical scenarios. In the future, we plan to integrate the physical scattering information of SAR images for the interpretable cross-domain classification for SAR targets.}, archive = {J_TASE}, author = {Zhunga Liu and Kun Li and Longfei Wang and Zuowei Zhang}, doi = {10.1109/TASE.2024.3407130}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4051--4062}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-scale alignment domain adaptation for ship classification in multi-resolution SAR images}, volume = {22}, year = {2025}, }
• (2025). Autonomous target docking with obstacle avoidance and final velocity control for non-holonomic mobile robots. TASE, 22, 4039--4050. (www)

  @article{ , abstract = {Robust, reliable, low-cost, and precise target docking solutions are in high demand in the industry. This paper investigates the docking problem of two non-holonomic robots with a fork attached to one of them and a docking station attached to the other. To secure latching the fork to the docking station, it must be oriented in a particular direction and distance from the docking station. Toward this and to perform docking operations, a navigation strategy is proposed which enables orientations and distances to be matched with the desired values. In addition, the mobile robot must intercept the target at a desired closing velocity for the docking mission to latch the fork and the docking station. This method also allows the final closing velocity to be set at any arbitrary value. This paper considers the docking operation of the assembling robots, and also proposes a solution to the docking problem, when the target moves at a constant velocity. As a result, the presented method allows the assembly of the robots, while moving. In addition, since safety is a critical factor in robotic systems operating in dynamic environments, a control-barrier-functions-based technique is adopted to avoid both static and dynamic obstacles. The performance of the proposed method is validated through a simulation and comparison study, then real-world implementation is performed using QBot robots.Note to Practitioners—Autonomous target docking for mobile robots has gained prominence, driven by its diverse applications like pallet picking, parking, refueling, automatic charging, and robot assembly. This paper specifically addresses the challenges of mobile robot docking in industrial settings, where shared workspaces involve static and dynamic obstacles, including equipment and humans. Two main challenges emerge: navigating a complex environment with obstacles and the time-consuming requirement for the robot to come to a complete stop before docking, particularly in scenarios with multiple docking targets or frequent operations. The paper proposes a solution allowing mobile robots to dock with a moving target at a constant velocity. This approach not only improves efficiency by eliminating the need for stopping during docking but also enhances safety through clear navigation paths and controlled closing velocity for precise docking. The presented solution shows promise in addressing the autonomous target docking problem in industrial environments, offering possibilities for further advancements and applications. Simulation and experimental results validate the effectiveness of the proposed method.}, archive = {J_TASE}, author = {Meysam Yadegar and Zahra Kashi and Hadi Ghaderi and Nader Meskin and Mohammad Noorizadeh}, doi = {10.1109/TASE.2024.3406901}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4039--4050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous target docking with obstacle avoidance and final velocity control for non-holonomic mobile robots}, volume = {22}, year = {2025}, }
• (2025). APTMRS: Autonomous prism target maintenance robotic system for FAST. TASE, 22, 4022--4038. (www)

  @article{ , abstract = {The Five Hundred Meter Spherical Radio Telescope (FAST) is the largest spherical radio telescope in the world, and there are more than 2,000 prism targets distributed on its reflector that require regular maintenance. These prism targets are screwed into the corresponding threaded holes by target bolts. At present, manual maintenance is mainly used, which is inefficient and unsafe. To address this issue, we develop an autonomous prism target maintenance robotic system, called APTMRS. This system integrates a mobile robot with an assembly robot and a collaborative robot, enabling it to move and perform the prism target replacement task on the reflector. Its technical development is threefold. First, a robotic end-effector that can be used for screwing the non-standard target bolt from the side is designed. Second, a novel pose measurement framework, which incorporates a feature point extraction and matching method, is utilized to measure the pose of the prism target. Third, the manipulation policies for the challenging steps in the workflow are presented, including picking and placing, initial thread mating, and final tightening. Independent experiments indicate that the performances of these methods are satisfactory. Moreover, tests in both simulated and real FAST scenarios confirm the effectiveness of APTMRS in achieving autonomous prism target maintenance. Note to Practitioners—Due to its excellent versatility, threaded connections have been widely adopted in a multitude of scenarios. Threads are not only present on standard bolts and nuts but also on a myriad of non-standard components. The robotic system proposed in this paper offers a solution for assembling and disassembling target bolts for FAST. Specifically, the scenario applied in this paper covers several extremely challenging problems in autonomous threaded assembly. Firstly, the object of assembly is a non-standard bolt, with obstacles above it, rendering traditional tools inapplicable. Secondly, the surface of the object is weakly textured and highly reflective. These pose significant challenges for the vision system to accurately acquire its pose. Lastly, the assembly environment is harsh. The threaded holes reside on an incline that is hard to access, with a maximum slope of 56°. To address these challenges, this paper presents a robotic system supplemented by three techniques including a novel robotic end-effector, a new pose measurement framework, and manipulation policies. The experimental results indicate that this robotic system is capable of autonomously disassembling and assembling non-standard target bolts. We explore the potential for generalizing these designs and methods to other scenarios, with the hope of contributing new insights to the development of threaded assembly systems.}, archive = {J_TASE}, author = {Rui Tao and Fengshui Jing and Jun Hou and Shiyu Xing and Yichen Fu and Junfeng Fan and Min Tan}, doi = {10.1109/TASE.2024.3406772}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4022--4038}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {APTMRS: Autonomous prism target maintenance robotic system for FAST}, volume = {22}, year = {2025}, }
• (2025). Precise end-effector control for an aerial manipulator under composite disturbances: Theory and experiments. TASE, 22, 4006--4021. (www)

  @article{ , abstract = {One of inescapable challenges in facilitating the application of aerial manipulators is to achieve the high precision control performance of the end-effector. The manipulator motions beneath the UAV platform constantly contend with composite disturbances, such as floating base, strong inner coupling effects, and model uncertainties. These factors collectively contribute to an inadequate control performance. In this paper, a composite control scheme is presented to tackle this issue. Specifically, a joint velocity planner is proposed to handle the base-floating disturbance in kinematic loop. By virtue of the generated joint reference signal, the base-floating disturbance can be effectively alleviated. The tracking error of the end-effector can be ensured within a small set. Moreover, in a complementary manner, neural network (NN) approximation and nonlinear disturbance observer (NDO) compensation are combined to track the joint references. The NN is adopted to estimate composite dynamic model including inner coupling effects and model uncertainties, while the NDO is designed to handle the remaining uncompensated part. The stability of the closed-loop system including the manipulator kinematics and dynamics is guaranteed using the Lyapunov-like method. Experimental results are reported to manifest the effectiveness of the proposed composite control scheme.Note to Practitioners—This work is driven by the precise end-effector control problem of an aerial manipulator subject to base-floating, strong dynamic coupling, and model uncertainties. Most of existing approaches implicitly address this issue by improving the flight performance of the aerial platform. However, the composite disturbances acted on the manipulator, which would deteriorate the operation accuracy of the end-effector, are not systematically addressed. In this work, a composite control scheme is constructed, which consists of the manipulator joint velocity planner and the dynamic controller. The idea is intuitive. The joint velocity is generated to counteract the fluctuation of the aerial platform. Furthermore, the dynamic controller is developed to accurately track the planned joint velocity in the presence of strong coupling and model uncertainties. The proposed scheme guarantees the stability of the close loop system, making our approach especially promising solution for aerial manipulation under composite disturbances.}, archive = {J_TASE}, author = {Meng Wang and Shangke Lyu and Qianyuan Liu and Ziqi Yang and Kexin Guo and Xiang Yu}, doi = {10.1109/TASE.2024.3406754}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {4006--4021}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Precise end-effector control for an aerial manipulator under composite disturbances: Theory and experiments}, volume = {22}, year = {2025}, }
• (2025). Contrast, imitate, adapt: Learning robotic skills from raw human videos. TASE, 22, 3993--4005. (www)

  @article{ , abstract = {Learning robotic skills from raw human videos remains a non-trivial challenge. Previous works tackled this problem by leveraging behavior cloning or learning reward functions from videos. Despite their remarkable performances, they may introduce several issues, such as the necessity for robot actions, requirements for consistent viewpoints and similar layouts between human and robot videos, as well as low sample efficiency. To this end, our key insight is to learn task priors by contrasting videos and to learn action priors through imitating trajectories from videos, and to utilize the task priors to guide trajectories to adapt to novel scenarios. We propose a three-stage skill learning framework denoted as Contrast-Imitate-Adapt (CIA). An interaction-aware alignment transformer is proposed to learn task priors by temporally aligning video pairs. Then a trajectory generation model is used to learn action priors. To adapt to novel scenarios different from human videos, the Inversion-Interaction method is designed to initialize coarse trajectories and refine them by limited interaction. In addition, CIA introduces an optimization method based on semantic directions of trajectories for interaction security and sample efficiency. The alignment distances computed by IAAformer are used as the rewards. We evaluate CIA in six real-world everyday tasks, and empirically demonstrate that CIA significantly outperforms previous state-of-the-art works in terms of task success rate and generalization to diverse novel scenarios layouts and object instances.Note to Practitioners—This work aims to study robot skill learning from raw human videos. Compared with teleoperation or kinesthetic teaching in the laboratory, such learning method can flexibly utilize large-scale human videos available on the Internet, thereby improving the robot’s ability to generalize to various complex scenarios. Previous works on learning from videos usually have some issues, including requirements for robot actions, consistent viewpoints, similar layouts and low sample efficiency. To alleviate these issues, we propose a three-stage skill learning framework CIA. Temporal alignment is utilized to learn task priors through our proposed transformer-based model and self-supervised loss functions. A trajectory generation model is trained to learn the action priors. To further adapt to diverse scenarios, we propose a two-stage policy improvement method by initialization and interaction. An optimization method is introduced to ensure safe interaction and sample efficiency, where the optimization objective is guided by the learned task priors. The experimental results show that our CIA outperforms other state-of-the-art methods in task success rate and generalization to novel scenarios.}, archive = {J_TASE}, author = {Zhifeng Qian and Mingyu You and Hongjun Zhou and Xuanhui Xu and Hao Fu and Jinzhe Xue and Bin He}, doi = {10.1109/TASE.2024.3406610}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3993--4005}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Contrast, imitate, adapt: Learning robotic skills from raw human videos}, volume = {22}, year = {2025}, }
• (2025). Online iterative optimization for angle tracking of electromagnetic scanning micro-mirror with unmatched interference and hysteresis. TASE, 22, 3980--3992. (www)

  @article{ , abstract = {In this paper, a precise angle tracking control strategy of electromagnetic driven scanning micro-mirror (EDSM) using online iterative optimization control (OIOC) scheme is addressed. Considering modeling difficulty caused by nonlinear hysteresis and complex dynamic characteristics, a simplified model (SM) is applied to describe the essential motion trait of EDSM approximately, more than that, the predictive model (PM) is developed based on the SM to improve the accuracy of the model. As for the omitted section caused by unmolded nonlinear and dynamic factor with unmatched interference (e.g., oscillation, nonlinear hysteresis and stochastic noise), that’s all being introduced into the angular deviation compensator (ADC) once again. Subsequently, a compensating structure SM+ADC is acquired to describe the dynamic performance and to design angular tracking control. Due to the nonlinear performance of multi-valued mapping, the generalized gradient is applied to optimize the control variables located in the nonlinear interval aiming to suppress the model deviation and eliminate unmatched interference. Then, a new iterative optimization control OIOC+ADC is obtained on the basis of compensation structure and generalized gradient. Besides, by designing Lyapunov function, the proposed control design incorporated by the relationship between optimized step-size and optimization gradient that needs to be satisfied guarantees the high accuracy positioning and stability. Conducted quantitative experiment results are provided to validate the effectiveness of the proposed control strategy in precise angle tracking to the fast dynamic system.}, archive = {J_TASE}, author = {Qingmei Cao and Yonghong Tan}, doi = {10.1109/TASE.2024.3406166}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3980--3992}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online iterative optimization for angle tracking of electromagnetic scanning micro-mirror with unmatched interference and hysteresis}, volume = {22}, year = {2025}, }
• (2025). Piezoelectric planar parallel microrobot with high bandwidth and precision for micromanipulation. TASE, 22, 3971--3979. (www)

  @article{ , abstract = {Parallel micro/nano robots hold great potential in micro-manufacturing/assembly, microsurgery, and precision engineering because of their high precision and stiffness. However, it is challenging to develop a millimeter-scale robot with a large workspace, high bandwidth, and high precision. In this paper, we present the design, fabrication, tests, and potential applications of a piezoelectric planar parallel microrobot. The developed microrobot consists of a parallel mechanism, three amplification mechanisms, and three independently controlled piezoelectric actuators. The microrobot is miniaturized to the millimeter scale through a monolithic integrated manufacturing process, and achieves a dimension of $36\times 36\times 34$ mm, a weight of 4.9 g, and a static workspace of 33.9 mm2. The resonant frequencies reach 55–65 Hz in x and y directions, and 95 Hz in rotation. The microrobot exhibits high positioning accuracy in the trajectory tracking experiment on three different lines, circle, and triangle trajectories at a board bandwidth. Moreover, the microrobot can repeat 50 periodic circle trajectories in one second, with a velocity of 628.3 mm/s and a precision of $10.9~\mu $ m. Furthermore, we conducted three validation experiments to demonstrate the potential applications of the microrobot in tremor compensation for micromanipulation, 3D printing electronics, and minimally invasive surgery. Note to Practitioners–This work is motivated by the need to design a microrobot with a large workspace, high bandwidth, high precision, and compact structure. Such capabilities are essential across various fields of micromanipulation, including micro-manufacturing/assembly, microsurgery, and precision engineering. This paper presents a detailed manufacturing process for this type of microrobot, followed by comprehensive experimental tests. The resonant frequency, workspace, quasi-static and dynamic trajectory tracking experiments of the microrobot are thoroughly tested. The experimental results demonstrate the excellent performance of the microrobot in executing various operational tasks. Additionally, the potential applications in the fields of micromanipulation, 3D printing electronics, and minimally invasive surgery are demonstrated. By assembling various end effectors on the robot platform, the microrobot could have greater potential for application in more fields.}, archive = {J_TASE}, author = {Jiaxu Shen and Qin Fang and Xizheng Fang and Junqiang Lou and Yue Wang and Rong Xiong and Haojian Lu}, doi = {10.1109/TASE.2024.3405347}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3971--3979}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Piezoelectric planar parallel microrobot with high bandwidth and precision for micromanipulation}, volume = {22}, year = {2025}, }
• (2025). Automated non-invasive analysis of motile sperms using sperm feature-correlated network. TASE, 22, 3960--3970. (www)

  @article{ , abstract = {An unbiased assessment of sperm morphology and motility is crucial for assessing fertility potential and guiding visual feedback for microrobotic manipulation. Automated analysis and selection of optimal sperm are essential for in vitro fertilization treatments, such as robotic intracytoplasmic sperm injection. However, conventional image processing methods face limitations in analyzing small sperm objects under microscopic imaging. While convolutional neural networks (CNNs) have brought promising advancements in microscopic image analysis, previous CNN methods have struggled to accurately differentiate tiny objects. These methods often require staining or fluorescence techniques to enhance visual contrast between sperm and culture medium, leading to clinical impracticality. To address these limitations, we introduce a novel sperm recognition network named the sperm feature-correlated network (SFCNet), for accurate and efficient segmentation and tracking of minute sperm objects. The SFCNet employs innovative modules, including collateral multi-scale convolution, cross-scale feature map guide, atrous spatial pyramid convolution with pooling, lateral attention, and multi-scale tracking proposal, to preserve essential sperm details despite their small size. Experimental results indicate that the SFCNet surpassed the state-of-the-art models designed for segmenting or tracking small objects, achieving up to a 28.39% higher Sørensen-Dice coefficient in segmentation and a 10.33% higher average precision in tracking. Additionally, the SFCNet excelled in sperm morphometric analysis, achieving errors below 15%. Moreover, the SFCNet also secured top-tier performance in sperm motility analysis, acquiring errors below 13% in seven sperm motility parameters. Note to Practitioners—This study is stimulated by the need to analyze the quality of motile sperms and select the optimal one for in vitro fertilization. Existing methods for detecting sperm fall short as they require a relatively high-magnification microscopic image or the usage of stain or fluorescence to increase sperm visualization, which limits the selection process or even makes the sperm clinically unavailable. To overcome these limitations, the present work proposes a new framework based on deep learning, which includes the design of extracting multi-scale sperm features. Experimental results suggest that the proposed method can perform better than existing methods in real-time analysis of multiple motile sperms’ morphology and motility at $20\times $ objective. In the future, there is a high potential for fertility specialists and healthcare workers to apply the presented framework in fertility treatment with higher accuracy and efficiency.}, archive = {J_TASE}, author = {Wei Dai and Zixuan Wu and Rui Liu and Tianyi Wu and Min Wang and Junxian Zhou and Zhuoran Zhang and Jun Liu}, doi = {10.1109/TASE.2024.3404488}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3960--3970}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated non-invasive analysis of motile sperms using sperm feature-correlated network}, volume = {22}, year = {2025}, }
• (2025). A novel robust imitation learning framework for complex skills with limited demonstrations. TASE, 22, 3947--3959. (www)

  @article{ , abstract = {Imitation learning allows us to directly encode manipulation skills based on human demonstrations, facilitating rapid transfer of skills without any expert knowledge. Autonomous dynamic systems (DS) offer reliable stability and time-independence though sacrificing part of accuracy, and are increasingly attractive as an encoding method. As unstructured environments become more challenging, skill trajectories become more complex, and various disturbances are encountered, existing state-of-the-art encoding methods struggle to adapt to these complex tasks. This paper introduces a novel robust DS-based framework for learning skills in complex tasks, which consists of trajectory regularization, adaptive segmentation, skill modeling, and skill organization based on new task requirements. It achieves a task-level generalization so that the operator only needs to focus on the semantic deconstruction of a task. Additionally, we propose an online modulation policy for the skill decision engine to address two types of disturbances: enhancing convergence speed for large-scale disturbances and improving fitting capability for small-scale disturbances while still keeping stability. To evaluate the effectiveness of the proposed framework, we conduct various comparison experiments in simulation and a real-world sugar-scooping task to assess the generalization performance and the ability of resistance to disturbances. Note to Practitioners—Imitation learning for complex tasks is crucial to the development of robot intelligence. However, achieving a balance between maintaining generalization accuracy and robustness remains a challenging problem that necessitates continuous exploration in the field of imitation learning. The purpose of this paper is to propose a robust imitation learning framework from human demonstration, which includes preprocessing, learning and generalization, that can be applied in industrial production or daily life. Considering appropriate trajectory segmentation and self-organization strategies can effectively improve the generalization accuracy by prior research, it is necessary to introduce them into our framework. Most importantly, we design novel disturbance-resistant online modulation strategies from both task-level and motion-level aspects. To validate the effectiveness of our approach, we conduct simulations and coffee scooping experiments. The results show that skills acquired through demonstration can reliably, accurately, and safely perform tasks even in uncertain environments. This paper is a systematic and pioneering attempt to implement.}, archive = {J_TASE}, author = {Weiyong Wang and Chao Zeng and Hong Zhan and Chenguang Yang}, doi = {10.1109/TASE.2024.3403833}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3947--3959}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel robust imitation learning framework for complex skills with limited demonstrations}, volume = {22}, year = {2025}, }
• (2025). Introducing novice operators to collaborative robots: A hands-on approach for learning and training. TASE, 22, 3933--3946. (www)

  @article{ , abstract = {Collaborative robots (cobots) have seen widespread adoption in industrial applications over the last decade. Cobots can be placed outside protective cages and are generally regarded as much more intuitive and easy to program compared to larger classical industrial robots. However, despite the cobots’ widespread adoption, their collaborative potential and opportunity to aid flexible production processes seem hindered by a lack of training and understanding from shop floor workers. Researchers have focused on technical solutions, which allow novice robot users to more easily train collaborative robots. However, most of this work has yet to leave research labs. Therefore, training methods are needed with the goal of transferring skills and knowledge to shop floor workers about how to program collaborative robots. We identify general basic knowledge and skills that a novice must master to program a collaborative robot. We present how to structure and facilitate cobot training based on cognitive apprenticeship and test the training framework on a total of 20 participants using a UR10e and UR3e robot. We considered two conditions: adaptive and self-regulated training. We found that the facilitation was effective in transferring knowledge and skills to novices, however, found no conclusive difference between the adaptive or self-regulated approach. The results demonstrate that, thanks to the proposed training method, both groups are able to significantly reduce task time, achieving a reduction of 40%, while maintaining the same level of performance in terms of position error.Note to Practitioners—This paper was motivated by the fact that the adoption of smaller, so-called collaborative robots is increasing within manufacturing but the potential for a single robot to be used flexibly in multiple places of a production seems unfulfilled. If more unskilled workers understood the collaborative robots and received structured training, they would be capable of programming the robots independently. This could change the current landscape of stationary collaborative robots towards more flexible robot use and thereby increase companies’ internal overall equipment efficiency and competencies. To this end, we identify general skills and knowledge for programming a collaborative robot, which helps increase the transparency of what novices need to know. We show how such knowledge and skills may be facilitated in a structured training framework, which effectively transfers necessary programming knowledge and skills to novices. This framework may be applied to a wider scope of knowledge and skills as the learner progresses. The skills and knowledge that we identify are general across robot platforms, however, collaborative robot interfaces differ. Therefore, a practical limitation to the approach includes the need for a knowledgeable person on the specific collaborative robot in question in order to create training material in areas specific to that model. However, with our list of identified skills, it provides an easier starting point. We show that relatively few skills and knowledge areas can enhance a novice’s programming capability.}, archive = {J_TASE}, author = {Andreas Kornmaaler Hansen and Valeria Villani and Andrea Pupa and Astrid Heidemann Lassen}, doi = {10.1109/TASE.2024.3403709}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3933--3946}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Introducing novice operators to collaborative robots: A hands-on approach for learning and training}, volume = {22}, year = {2025}, }
• (2025). Clustering weighted envelope spectrum for rolling bearing fault diagnosis. TASE, 22, 3922--3932. (www)

  @article{ , abstract = {Spectral coherence (SCoh) is a powerful tool to reveal the hidden periodicities of signals, which has been widely used for rolling bearing fault diagnosis. However, most SCoh-based methods focus on searching a single demodulation band, which results in their inability to compound fault diagnosis and discrete frequency band localization. Moreover, many studies are conducted based on prior fault characteristic frequencies (FCFs), which limits their application in limited vision cases. To solve such issues, a prior knowledge-needless method namely clustering weighted envelope spectrum (CWES) is proposed for rolling bearing fault diagnosis. Firstly, based on the algorithms of peak searching and multiple relation checking, the potential FCFs (PFCFs) of each spectral frequency slice (SFS) of SCoh are automatically identified without any prior knowledge. The PFCFs of each SFS are regarded as its fault type label and are used to design a weight to evaluate its fault information abundance. Then, the SFSs with similar labels are clustered and other SFSs are ignored. Each cluster is considered to be associated with a potential cyclostationary component, and the importance of all clusters is sorted based on their maximum weights. Finally, to further enhance the fault characteristics, CWESs are defined as the weighted average of the SFSs in each top-ranked cluster. By using this method, the discrete informative frequency bands of multiple faults can be quickly located without prior FCFs and iterative optimization. The advantages of CWES over the state-of-the-art methods are validated by the experimental data of bearing single and compound faults. The results indicate that CWES has the best completeness in fault information extraction and the highest accuracy of fault diagnosis compared with other methods. Moreover, the robustness and computational efficiency of the proposed method are also advantageous.Note to Practitioners—This paper is motivated by the problems of discrete frequency band localization and compound fault separation in the field of rolling bearing fault diagnosis. Different from other prior FCF-oriented methods, we design a prior knowledge-needless algorithm to identify the PFCFs of each SFS of the SCoh. The PFCFs of each SFS can not only indicate the fault type but also quantify the abundance of fault information. Based on the identified PFCFs, several CWESs can be generated for fault diagnosis through the clustering algorithm and the weighted mechanism. Our experimental results show the proposed method has higher diagnostic accuracy than the existing methods.}, archive = {J_TASE}, author = {Tao Chen and Liang Guo and Hongli Gao and Tingting Feng and Yaoxiang Yu}, doi = {10.1109/TASE.2024.3403665}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3922--3932}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Clustering weighted envelope spectrum for rolling bearing fault diagnosis}, volume = {22}, year = {2025}, }
• (2025). Time-varying formation-containment control via PDE backstepping technique. TASE, 22, 3911--3921. (www)

  @article{ , abstract = {This paper focuses on the formation-containment (FC) control for multi-agent systems (MASs) with long input delay under jointly connected switching topologies. The considered MASs contain one virtual leader, multiple leaders and followers. The leaders complete the formation first, then the followers move into the formation created by the leaders. Compared with existing FC control, each agent in the considered MASs can have different dynamics. To handle long input delay, based on the partial differential equation (PDE) backstepping technique, a new distributed predictor feedback control method is proposed. Utilizing this method and the certainty equivalence principle, the MASs with long input delay can be mapped into new coupled ODE-PDE target systems. Then, Lyapunov-Krasovskii functions are used to prove the stability of the system. Simulations verify the validity of the proposed results.Note to Practitioners—In actual applications, multiple agents need to perform complex tasks together. These agents can be divided into leaders and followers. The leaders need to establish a time-varying formation (TVF) around a virtual leader to adapt to the changing circumstances, then the followers can move into the multiple leaders’ spanning convex hull so they can be protected from outside obstacles. In practical, input delays frequently occur in MASs due to network transmission, which brings difficulties to the FC control. Although there are plenty of works on FC control, research on the FC control of MASs with long input delay is still lacking.}, archive = {J_TASE}, author = {Jiachun Sun and Shiqi Zheng and Choon Ki Ahn and Yifeng Zha and Peng Shi}, doi = {10.1109/TASE.2024.3403130}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3911--3921}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Time-varying formation-containment control via PDE backstepping technique}, volume = {22}, year = {2025}, }
• (2025). Twofold weighted-based statistical feature KECA for nonlinear industrial process fault diagnosis. TASE, 22, 3901--3910. (www)

  @article{ , abstract = {In order to ensure the safe operation of industrial systems, the timely diagnosis of incipient faults is gradually gaining attention. The kernel entropy component analysis (KECA) has been widely used in the fault diagnosis of nonlinear industrial processes. However, the KECA often performs unsatisfactorily in the case of incipient faults. Therefore, a novel incipient fault detection and diagnosis method based on the statistical feature KECA integrating the twofold weighted (TWSFKECA) is proposed. The residual function in the local approach is combined with the KECA to construct statistical features of the data. Then, in order to highlight the influence of incipient faults of statistical features, the statistical feature sample weighting strategy is established based on the dissimilarity analysis between the test and training samples. Furthermore, the statistical feature component weighting strategy is developed for the sensitive components, which are judged by applying the Durbin-Watson (DW) criterion to calculate the extent to which the sample-weighted statistical feature components contain significant information. Moreover, based on the statistical features of twofold weights, two statistics indexes are created for incipient fault detection. In addition, the strategy for process fault diagnosis using a variable contribution plot method is proposed to isolate faulty variables. Finally, the continuous stirred tank reactor control system and the Tennessee Eastman process illustrate the superiority of the proposed method for incipient fault detection and diagnosis. Note to Practitioners—Effective detection of incipient faults prevents the evolution of accidents and ensures the smooth operation of the production process. In nonlinear industrial processes, the statistical feature KECA integrating the twofold weighted is proposed for incipient fault detection and diagnosis. The residual function is introduced in the kernel entropy component analysis to construct the statistical features of the data, which helps extract the incipient fault information. Then, the twofold weighting strategy weights the statistical features in terms of samples and components, highlighting the influence of the main samples and sensitive components in the incipient faults, respectively. In addition, a variable contribution plot method is developed to solve the problem of not being able to find out the cause of faults through control plots. The experimental results further verify the applicability of the proposed method for monitoring the occurrence of incipient faults.}, archive = {J_TASE}, author = {Tao Li and Yongming Han and Xuan Hu and Bo Ma and Zhiqiang Geng}, doi = {10.1109/TASE.2024.3402653}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3901--3910}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Twofold weighted-based statistical feature KECA for nonlinear industrial process fault diagnosis}, volume = {22}, year = {2025}, }
• (2025). Transparent fault diagnosis for complex equipment considering expert reliability based on belief rule base and linguistic Z-number. TASE, 22, 3887--3900. (www)

  @article{ , abstract = {Fault diagnosis is crucial for mastering the operation status of complex equipment. However, due to the complexity of large-scale data-driven models, the process of fault diagnosis is difficult to comprehend and provide convincing results. In this paper, a transparent rule-based fault diagnosis method is proposed by introducing expert reliability, which covers three aspects: modeling, reasoning and optimization. Firstly, a transparent knowledge base is proposed using belief rule base and linguistic Z-number, namely LZ-BRB, in which the expert reliability is modeled. LZ-BRB characterizes the mapping relationship reflected by incompletely reliable knowledge between fault features and fault states. Secondly, a new evidential reasoning rule with the unreliable proposition (ER-UP) is developed to perform the reasoning of LZ-BRB and obtain the diagnosis result. The reliability of diagnosis results is acquired and measures the trustworthiness of the diagnosis result. Thirdly, to improve the diagnosis accuracy and maintain the transparent modeling process, expert reliability is considered in model optimization and a knowledge-data hybrid-driven gradient descent (KDGD) method is developed. Finally, a fault diagnosis case of the aerospace relay verifies the proposed method. The case results show that the diagnosis performance of the proposed method is competitive and maintains superb transparency. Note to Practitioners—Fault diagnosis holds paramount importance in the health management of industrial equipment. Nowadays, large-scale intelligent fault diagnosis methods have gained widespread application in the diagnosis of various mechanical and electronic equipment. Nevertheless, due to the complex structure and excessive parameters, it is hard for managers to make sense of the modeling, reasoning and optimization process of these models. Hence, there is an urgent need for a transparent fault diagnosis approach that can provide trustworthy diagnosis results. In this paper, based on BRB and linguistic Z-number, we present a transparent rule-based fault diagnosis method considering the knowledge reliability of human experts in engineering, including a transparent knowledge base, inference engine and model optimization. A fault diagnosis case of aerospace relay validates the effectiveness of the method and demonstrates the transparency of the fault diagnosis process. The proposed method can effectively model incompletely reliable expert knowledge and achieve transparent fault diagnosis, contributing to the health management of complex equipment.}, archive = {J_TASE}, author = {Zheng Lian and Zhijie Zhou and Changhua Hu and You Cao and Shuaiwen Tang and Yijie Sun}, doi = {10.1109/TASE.2024.3402099}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3887--3900}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transparent fault diagnosis for complex equipment considering expert reliability based on belief rule base and linguistic Z-number}, volume = {22}, year = {2025}, }
• (2025). Distributed set-membership fusion estimation for complex networks with communication constraints. TASE, 22, 3877--3886. (www)

  @article{ , abstract = {This paper concerns the distributed set-membership fusion estimation (SMFE) problem of complex networks subject to communication constraints and unknown-but-bounded (UBB) noises, where nodes communicate with their neighbours based on a given topology. There are three main contributions: 1) an event-based coding-decoding mechanism (CDM) is designed in each communication channel to code the transmitted data with a finite bit rate, where the purpose is to save communication resources and enhance transmission security; 2) a fusion estimation method, which can fusion a group of local estimation sets, is introduced to estimate the system state with a trace-maximal ellipsoid, where the local estimators make full use of the information from itself as well as its neighbours. It helps to improve the accuracy and precision of the estimation effectively; 3) a genetic algorithm (GA) is firstly adopted to tackle the co-design issue of bit rate allocation protocol and estimator gain. It aims to reduce decoding errors while ensuring good estimation performance. To demonstrate the superiority of the above contributions, a numerical simulation is provided to validate the proposed SMFE method. Note to Practitioners—Complex networks can be applied to various practical problems, such as biological networks, neural networks, and social networks. The set-membership estimation (SME) method limits the state of the system to an ellipsoidal region. This paper studies the SMFE method, which fuses an ellipsoidal set containing the intersection of all local ellipsoidal sets to form a relatively small region containing the system state, thereby improving the estimation performance. Each communication channel equals with event-based CDM, which improves the robustness and security of data transmission. The bit rate allocation mechanism optimized by GA breaks the traditional uniform allocation scheme, which can reduce decoding error and improve estimation accuracy. Therefore, the method proposed in this paper can be used for navigation and positioning devices.}, archive = {J_TASE}, author = {Changzhen Hu and Xiangpeng Xie and Sanbo Ding and Yanhui Jing}, doi = {10.1109/TASE.2024.3401740}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3877--3886}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed set-membership fusion estimation for complex networks with communication constraints}, volume = {22}, year = {2025}, }
• (2025). Hierarchical Q-learning path planning for cooperative tracking control of multi-agent systems with lumped uncertainties. TASE, 22, 3866--3876. (www)

  @article{ , abstract = {This paper presents the hierarchical Q-learning path planning (HQPP) architecture for solving the cooperative tracking control problem of multi-agent systems (MASs) with lumped uncertainties in an unknown environment. The presented architecture consists of three layers, namely, the decision layer, the distributed estimated layer, and the local control layer. Specifically, in the decision layer, we propose the dynamic parameter and trajectory fitting Q-learning (DPTF-Q-learning) algorithm to find a feasible continuous trajectory to the target in an unknown environment. In addition, two dynamic parameters are proposed and introduced into the DPTF-Q-learning algorithm to shorten the required minimum number of steps in the training process. Then, the distributed estimated layer is designed to broadcast the continuous trajectory generated from the decision layer based on the directed communication topology containing a spanning tree. In the local control layer, the cooperative tracking control (CTC) algorithm is proposed to achieve cooperative tracking for MASs in the presence of uncertain dynamics and external disturbances. The sufficient conditions for achieving cooperative tracking control are rigorously derived by employing Lyapunov argument. Finally, numerical simulations are presented to verify the effectiveness of the proposed architecture.Note to Practitioners—This paper is motivated by the need of developing an integrated path planning and control method for cooperative tracking of multi-agent systems in a no-signal environment and without the presence of users. Most related works are limited to separate fields: 1) most existing path planning techniques are only applicable to a single agent and discrete environments, and 2) most existing cooperative tracking algorithms focus on guaranteeing control stability and error convergence without decision-making capabilities. To address the above issues, this work proposes a hierarchical control architecture based on reinforcement learning for multi-agent systems to achieve path planning and cooperative tracking tasks. In addition, multi-agent systems exhibit strong robustness and fault tolerance due to their inherent characteristics, so the above mentioned research can be well applied to post-disaster rescue, intelligent logistics, future war, and so on. Numerical simulations based on Matlab and Python verify the effectiveness of the proposed architecture.}, archive = {J_TASE}, author = {Mai-Kao Lu and Ming-Feng Ge and Zhi-Wei Liu and Teng-Fei Ding}, doi = {10.1109/TASE.2024.3401456}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3866--3876}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical Q-learning path planning for cooperative tracking control of multi-agent systems with lumped uncertainties}, volume = {22}, year = {2025}, }
• (2025). An input saturation-tolerant position control method for magnetic microrobots using adaptive fuzzy sliding-mode method. TASE, 22, 3852--3865. (www)

  @article{ , abstract = {The position control of magnetic medical microrobots is influenced by several environmental uncertainties including the unknown characteristics of the medium and the imaging precision. Furthermore, measuring the physical attributes of the microrobots is a challenging issue. To provide a model-free position control approach for the magnetic medical microrobots, a saturation-tolerant Adaptive Fuzzy Sliding-Mode Control (AFSMC) is designed in this study. In the proposed approach, each control input comprises a fuzzy inference term utilized to approximate an unknown nonlinear function including uncertain forces, a robust part derived to compensate for the fuzzy approximation error and disturbances, and a compensating gain for input saturation. By utilizing the second theorem of Lyapunov and Barbalat’s lemma, it is proved that the closed-loop system is asymptotically stable. The effectuality of the presented controller is assessed by means of two experimental scenarios. The results show that the magnitudes of the tracking errors corresponding to a spiral reference path are less than 0.2 mm at the end of the motion. Moreover, in the test conducted in a 3D printed Aorta phantom, the minimum and the maximum values of the tracking errors are −1.22 mm and 0.95 mm, respectively.Note to Practitioners—This article introduces an innovative, model-free technique specifically designed to tackle the complex challenges of position control in magnetic medical microrobots. Achieving precise control over these microrobots is a challenging task, compounded by the complexity of accurately measuring their physical properties and the characteristics of their surrounding medium. This challenge is further exacerbated by the issue of input saturation, which can compromise system stability. Our pioneering control method is designed to navigate these obstacles effectively. It functions under the assumption that both the lower and upper saturation limits are unknown, and it eliminates the necessity to model the forces acting on the microrobot. Experimental results confirm the method’s effectiveness in accurately tracking various reference trajectories. These findings suggest that our method holds significant promise for various medical applications.}, archive = {J_TASE}, author = {Alireza Mousavi and Awais Ahmed and Hesam Khaksar and Hongsoo Choi and Ali Kafash Hoshiar}, doi = {10.1109/TASE.2024.3400602}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3852--3865}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An input saturation-tolerant position control method for magnetic microrobots using adaptive fuzzy sliding-mode method}, volume = {22}, year = {2025}, }
• (2025). Cloud-edge cooperative MPC for large-scale complex systems with input nonlinearity. TASE, 22, 3835--3851. (www)

  @article{ , abstract = {Nonlinear model predictive control (NMPC) is a promising approach for controlling large-scale complex systems (LSS) that exhibit nonlinearity and constraints. However, its computational and real-time limitations hinder its widespread adoption. To address this challenge, we propose a cloud-edge cooperative model predictive control (MPC) scheme that overcomes these limitations while ensuring the desired control performance. Specifically, our proposed approach involves designing a cloud-based NMPC with a high-fidelity nonlinear model for the cloud layer. Meanwhile, the edge layer is equipped with a simplified backup linear model predictive control (LMPC) that uses a linearized model based on constraint tightening to mitigate model mismatch errors. Additionally, we develop an automatic strategy that employs a sliding weighted average method to switch between the cloud and edge controllers, enhancing the system’s reliability under non-ideal networking conditions. We provide a thorough analysis of the recursive feasibility and asymptotic average performance of the control scheme with different prediction models in the cloud and edge layers. To validate our approach, we apply it to a charging system for plug-in hybrid electric vehicles (PHEVs). Furthermore, we compare the performance and computation efficiency of our proposed cloud-edge cooperative MPC scheme with four other MPC schemes. Note to Practitioners—This work aims to overcome the challenge of reducing the computational burden and increasing the applicability of nonlinear model predictive control (NMPC) in large-scale complex systems (LSS) with input nonlinearity. To address these issues, we propose a cloud-edge cooperative MPC scheme that involves deploying the cloud layer on a remote cloud server and the edge layer locally onboard the system. The proposed scheme not only provides a practical solution for LSS with input nonlinearity but is also applicable to general nonlinear systems. Further research will focus on exploring the relevant theory and practical aspects of the scheme.}, archive = {J_TASE}, author = {Yaling Ma and Li Dai and Huan Yang and Junxiao Zhao and Runze Gao and Yuanqing Xia}, doi = {10.1109/TASE.2024.3400598}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3835--3851}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cloud-edge cooperative MPC for large-scale complex systems with input nonlinearity}, volume = {22}, year = {2025}, }
• (2025). An improved maximal continuity graph solver for non-redundant manipulator non-revisiting coverage. TASE, 22, 3822--3834. (www)

  @article{ , abstract = {This paper proposes an improved solver for the maximal continuity graph painting problem. The problem is motivated by the real-world surface non-revisiting coverage path planning (NCPP) task carried out by manipulators, where the physical meaning of maximising the colouring continuity in the graph translates to minimising the number of undesirable transitions between end-effector force/torque control discontinuities. Early works have formulated the graph-based representation of the task and finitely solved the graph. However, the exponential growth of its algorithmic complexity makes the problem intractable for even relatively simple graphs. The improved solver proposed in this paper demonstrates exponential improvement compared to the state-of-the-art algorithm, setting guaranteed bounds on performance, whereby the algorithmic complexity is proven reduced by a factor of $2^{N}$ , N being the number of internal edges in the graph. Challenging simulated experiments are presented to validate the computational advantage, and an open source implementation is also provided for the benefit of the community. Note to Practitioners—To solve a non-redundant manipulator NCPP task, the first step is collecting all valid inverse kinematics configurations that lead to coverage on the surface. Continuous configurations can then be grouped and assigned the same colour. This process creates a spatial distribution of colours on the target surface, forming a topological graph as detailed in this paper. Using the proposed algorithm, each coverable point on the surface is assigned a colour, which specifies the unique inverse kinematic configuration that the manipulator should adopt to cover such a point. A suitable geometric coverage planner can then be employed to generate the path that a manipulator end-effector must follow on the surface that is guaranteed to have the minimum number of end-effector discontinuities. In this paper, an (open-sourced) solver is proposed to solve the graph optimally from a computational point of view, most notably increasing productivity from a manufacturing/automated perspective.}, archive = {J_TASE}, author = {Tong Yang and Jaime Valls Miro and Yue Wang and Rong Xiong}, doi = {10.1109/TASE.2024.3400518}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3822--3834}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An improved maximal continuity graph solver for non-redundant manipulator non-revisiting coverage}, volume = {22}, year = {2025}, }
• (2025). Distributed event-triggered output-feedback time-varying formation fault-tolerant control for nonlinear multi-agent systems. TASE, 22, 3810--3821. (www)

  @article{ , abstract = {This paper studies the event-triggered time-varying formation control problem for nonlinear multi-agent systems with actuator faults. Based on the neural network approximation technique, a neural observer is constructed to estimate the unmeasured states of systems. Then, a distributed adaptive event-triggered time-varying formation control manner is proposed utilizing the intermittent estimated states information from the agent and its neighbors. To overcome the problem that estimated states triggering leads to virtual control laws is non-differentiable, a distributed continuous control scheme under regular output-feedback is designed firstly, upon which a distributed event-triggered controller is constructed by replacing estimated states with intermittent estimated ones. It is shown that the designed event-triggered output-feedback time-varying formation fault-tolerant controller can compensate for actuator faults, and all signals in closed-loop systems are semi-globally uniformly ultimately bounded. Finally, simulation results of a practical example are given to verify the effectiveness of the proposed control manner.Note to Practitioners—Formation control has broad application prospects in modern military and civilian fields, such as combat aircraft flying formation, satellite formation, autonomous vehicle formation, etc. In formation control systems, when agents occur actuator faults, it may break the original formation and even cause collision between agents. As a result, the security of formation control systems is facing great challenges in practical engineering applications. On the other hand, communication bandwidth is limited in practical engineering systems, and how to make systems quickly form formation under the limited communication bandwidth has become a key topic. Inspired by the above discussions, a distributed state-triggered output-feedback time-varying formation fault-tolerant control scheme is designed in this paper, in which actuator faults are compensated by using adaptive technology. Meanwhile, to sufficiently save the usage of system communication resources, a dual-channel event-triggered mechanism is designed.}, archive = {J_TASE}, author = {Xiangjun Wu and Ning Zhao and Shuo Ding and Huanqing Wang and Xudong Zhao}, doi = {10.1109/TASE.2024.3400325}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3810--3821}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed event-triggered output-feedback time-varying formation fault-tolerant control for nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Resilient human-in-the-loop formation-tracking of multi-UAV systems against byzantine attacks. TASE, 22, 3797--3809. (www)

  @article{ , abstract = {This study addresses resilient human-in-the-loop (HiTL) formation-tracking of multi-UAV systems against f-local Byzantine attacks. In the HiTL settings, a human operator plays a key role in detecting any physical hazard, monitoring the whole UAV swarm, and sending secure execution signals to a non-autonomous leader UAV. Moreover, there exists a fraction of Byzantine UAVs in the multi-UAV systems, which propagate incorrect information to their neighbors (called Byzantine edge attacks (BEAs)) and adopt false input signals (called Byzantine node attacks (BNAs)) when swarming. In order to suppress the above aggressive Byzantine attacks, this paper proposes a Byzantine-resilient hierarchical control scheme, including a virtual Digital Twin Layer (DTL) apart from a Cyber-Physical Layer (CPL). First, a distributed resilient estimation scheme is proposed on the DTL, which can realize resilient estimation on the state of the non-autonomous leader UAV against BEAs on the premise that the DTL topology is strongly $(2f+1)$ -robust. Second, a series of decentralized and chattering-free controllers is formulated on the CPL, which is resilient to both BNAs and inter-layered faults. The asymptotical control performance of the above controllers is strictly proven based on Cromwell-Bellman Lemma. To demonstrate the practicality of the theoretical results, a resilient HiTL multi-UAV systems experiment has been further conducted. The experimental results verify the effectiveness and practicality of the designed two-layered controllers against f-local Byzantine attacks.Note to Practitioners—Owing to the wide application of multi-UAV systems, the resilience of the whole swarm against malicious attacks has grasped the great attention of both academia and industry. This work considers a rather aggressive kind of attacks, named Byzantine attacks, where a fraction of unidentified UAVs act as traitors. Inspired by the digital twin technology, a two-layered control architecture for multi-UAV systems is formatted, including a Digital Twin Layer (DTL) and a Cyber-Physical Layer (CPL). Here are the highlights: 1) Control Architecture: The DTL handles Byzantine edge attacks (BEAs), while the CPL addresses Byzantine node attacks (BNAs), ensuring reliable human-swarm cooperation in adversarial environments. 2) Resilient Estimation against BEAs: A novel resilient estimation scheme on the DTL is designed, using edge-based feedback, which can estimate the states of the leader UAV manipulated by human operators. 3) Adaptive Controller against BNAs and Inter-layered Faults: On the CPL, a decentralized adaptive controller with adjustable and exponential convergence is proposed, enhancing its precision and flexibility. 4) Practical Application: A UAV swarm formation-tracking experiment validates the control architecture’s effectiveness in human-in-the-loop scenarios, demonstrating its practicality in the realm of swarm robotics and human-swarm interaction.}, archive = {J_TASE}, author = {Xin Gong and Jie Gui and Yong Chen and Xuefei Yang and Wenwu Yu and Tingwen Huang}, doi = {10.1109/TASE.2024.3400155}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3797--3809}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient human-in-the-loop formation-tracking of multi-UAV systems against byzantine attacks}, volume = {22}, year = {2025}, }
• (2025). Scheduling cluster tools for concurrent processing: Deep reinforcement learning with adaptive search. TASE, 22, 3783--3796. (www)

  @article{ , abstract = {We address the scheduling problem of single-armed cluster tools that concurrently process two wafer types without assuming cyclic scheduling. These cluster tools, consisting of multiple processing modules and a transport robot, are commonly used in semiconductor manufacturing processes, such as etching, deposition, and lithography. To optimize the tool’s throughput, we propose a reinforcement learning approach for determining both the robot task sequence and the release sequence of wafer types. By incorporating an adaptive search, our method intelligently explores future states to gather crucial information, enabling the selection of the best action. Extensive experiments demonstrate that our proposed method outperforms the well-known optimal robot task sequence for cyclic scheduling in single-armed cluster tools with concurrent processing. These findings underscore the effectiveness and superiority of our approach in optimizing the throughput of single-armed cluster tools, without relying on cyclic scheduling assumptions. Note to Practitioners—Single-armed cluster tools, comprising multiple processing modules and a transport robot, are commonly employed in semiconductor manufacturing processes. In recent times, there has been a trend towards concurrent processing of multiple wafer types within the fab, aimed at improving the PMs’ utilization. The concurrent backward sequence (CBS) has emerged as an efficient approach for such scenarios, assuming cyclic scheduling. However, the CBS relies on a cycle plan that specifies the number of wafers from each wafer type to be produced and their release sequence in a cycle. While the CBS provides an optimal schedule that maximizes the throughput of the tool in some cases under the cyclic scheduling assumption, its performance has not been guaranteed for other general cases. We propose a highly efficient reinforcement learning approach for the scheduling problem including the robot task sequence and release sequence of wafer types in a single-armed cluster tool. Our method outperforms the CBS, demonstrating its superior performance. By implementing our proposed approach, we believe that practitioners can effectively maximize the throughput of single-armed cluster tools and significantly enhance the fab productivity.}, archive = {J_TASE}, author = {Hyun-Jung Kim and Jun-Ho Lee}, doi = {10.1109/TASE.2024.3399818}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3783--3796}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling cluster tools for concurrent processing: Deep reinforcement learning with adaptive search}, volume = {22}, year = {2025}, }
• (2025). Learning to detect critical nodes in sparse graphs via feature importance awareness. TASE, 22, 3772--3782. (www)

  @article{ , abstract = {Detecting critical nodes in sparse graphs is important in a variety of application domains, such as network vulnerability assessment, epidemic control, and drug design. The critical node problem (CNP) aims to find a set of critical nodes from a network whose deletion maximally degrades the pairwise connectivity of the residual network. Due to its general NP-hard nature, state-of-the-art CNP solutions are based on heuristic approaches. Domain knowledge and trial-and-error are usually required when designing such approaches, thus consuming considerable effort and time. This work proposes a feature importance-aware graph attention network for node representation and combines it with dueling double deep Q-network to create an end-to-end algorithm to solve CNP for the first time. It does not need any problem-specific knowledge or labeled datasets as required by most of existing methods. Once the model is trained, it can be generalized to cope with various types of CNPs (with different sizes and topological structures) without re-training. Computational experiments on 28 real-world networks show that the proposed method is highly comparable to state-of-the-art methods. It does not require any problem-specific knowledge and, hence, can be applicable to many applications including those impossible ones by using the existing approaches. It can be combined with some local search methods to further improve its solution quality. Extensive comparison results are given to show its effectiveness in solving CNP. Note to Practitioners—This work is motivated by the problems of identifying influential nodes from a sparse graph or network. Various practical applications can be naturally modeled as critical node problems, e.g., finding the most influential stations or airports within a transportation network, identifying a specific number of people to be vaccinated in order to reduce the overall transmissibility of a virus, and reinforcing the protection over some most important nodes to make the electric network more stable. It proposes an effective end-to-end deep learning algorithm to solve the critical node problem. The proposed approach combines feature importance-aware graph attention network with dueling double deep Q-network. Extensive numerical experiments and comparisons show that our proposed algorithm is highly comparable to state-of-the-art algorithms and can help decision-makers to discover valuable knowledge or influential nodes in a real-world network.}, archive = {J_TASE}, author = {Xuwei Tan and Yangming Zhou and MengChu Zhou and Zhang-Hua Fu}, doi = {10.1109/TASE.2024.3399322}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3772--3782}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning to detect critical nodes in sparse graphs via feature importance awareness}, volume = {22}, year = {2025}, }
• (2025). Cost-driven scheduling for workflow decision making systems in fuzzy edge-cloud environments. TASE, 22, 3756--3771. (www)

  @article{ , abstract = {Workflow decision making is critical to performing many practical applications of scientific principles and data. Scheduling in edge-cloud environments can address the high complexity of workflow applications, while decreasing the data transmission delay between the cloud and end devices. However, due to the heterogeneous resources in edge-cloud environments and the complicated data dependencies between the tasks in a workflow, significant challenges for workflow scheduling remain, including the selection of an optimal tasks-servers solution from the possible numerous combinations. Existing studies are mainly done subject to rigorous conditions without fluctuations, ignoring the fact that workflow scheduling is typically present in uncertain environments. In this study, we focus on reducing the execution cost of multiple workflow applications mainly caused by data transmission and task computation, while satisfying the required deadline constraints. Triangular fuzzy numbers are employed to represent the computing performance of servers and transmission bandwidth in fuzzy edge-cloud environments. A cost-driven scheduling strategy for multiple Poisson-arrived workflow applications using partial critical paths is proposed. It firstly merges cut edges through preprocess to reduce the workflow scale, then uniformly schedules all tasks on each partial critical path to avoid data transmission between dependent tasks and reduce the data transmission cost. The experimental results show that our strategy can obtain the optimal feasible scheduling scheme and have better robustness and real-time performance with different deadline constraints, compared with other benchmark strategies. Note to Practitioners—Vehicle identification is one of the workflow decision making systems in transportation environments, whose core technology is Deep Neural Networks (DNN). Traffic cameras with limited process capacity periodically record the images of on-road vehicles, and usually fail to complete the applications within their deadlines. Workflow decision making is one of the key issues to performance DNNs in vehicle identification applications. The uncertain environments have a great impact on the system latency for such problems, which can easily lead to the misjudgement of the optimal scheduling. In addition, it is difficult to select an optimal layers-servers solution from the numerous combinations. Therefore, we can employ the scheduling strategy (i.e., SWPCP) to make intelligent and faster workflow decisions for vehicle identification applications, which can reduce the execution cost mainly caused by layer computation and data transmission between layers within their deadlines, even in uncertain edge-cloud environments. Complex DNN layers (tasks) in vehicle identification applications can be scheduled to the cloud for execution, while simple ones are processed on the edge. The cloud and edge platforms collaborate with each other and execute the DNN layers with low system cost and latency.}, archive = {J_TASE}, author = {Bing Lin and Chaowei Lin and Xing Chen and Mingwei Lin and Gang Huang and Zeshui Xu}, doi = {10.1109/TASE.2024.3435026}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3756--3771}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cost-driven scheduling for workflow decision making systems in fuzzy edge-cloud environments}, volume = {22}, year = {2025}, }
• (2025). G²VD planner: Efficient motion planning with grid-based generalized voronoi diagrams. TASE, 22, 3743--3755. (www)

  @article{ , abstract = {In this paper, an efficient motion planning approach with grid-based generalized Voronoi diagrams (G2VD) is newly proposed for mobile robots. Different from existing approaches, the novelty of this work is twofold: 1) a new state lattice-based path searching approach is proposed, in which the search space is reduced to a novel Voronoi corridor to further improve the search efficiency; 2) an efficient quadratic programming-based path smoothing approach is presented, wherein the clearance to obstacles is considered to improve the path clearance of hard-constrained path smoothing approaches. We validate the efficiency and smoothness of our approach in various challenging simulation scenarios and outdoor environments. It is shown that the computational efficiency is improved by 17.1% in the path searching stage, and path smoothing with the proposed approach is 6.6 times faster than an advanced sparse-banded structure-based path smoothing approach and 53.3 times faster than the popular timed-elastic-band planner. A video showing outdoor navigation on our campus is available at https://youtu.be/iMXGthgvp58. Note to Practitioners—This paper is motivated by the challenges of motion planning problems of mobile robots. An efficient motion planning approach called G2VD planner is proposed by combining path searching, path smoothing, and time-optimal velocity planning. Extensive simulation and experimental results show the effectiveness of the proposed motion planning approach. However, the prediction information of dynamic obstacles is not incorporated in the proposed motion planner, thus the motion planner may be a bit sluggish in response to dynamic obstacles. Furthermore, we plan to integrate the intention/trajectory prediction of pedestrians/vehicles into the proposed framework to enhance the foreseeability of the motion planner.}, archive = {J_TASE}, author = {Jian Wen and Xuebo Zhang and Qingchen Bi and Hui Liu and Jing Yuan and Yongchun Fang}, doi = {10.1109/TASE.2024.3398996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3743--3755}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {G²VD planner: Efficient motion planning with grid-based generalized voronoi diagrams}, volume = {22}, year = {2025}, }
• (2025). Predictor-based active disturbance rejection control of wet flue gas desulfurization system with delay robustness and simplified tuning. TASE, 22, 3731--3742. (www)

  @article{ , abstract = {Flue gas desulfurization in thermal power plants is critical to the environmental and human health. Aiming for field application, this article presents a new predictor-and-observer-based active disturbance rejection control (ADRC) scheme for Wet Flue Gas Desulfurization (WFGD) process which is subject to long time delay and unknown disturbances. First, the working principles and control problems of WFGD system are illustrated, and the WFGD process model is identified based on the field experimental data. Then, a predictor-based ADRC approach is carefully devised to address its control issues. Specifically, a filtered smith predictor (FSP) is designed to obtain stable delay-free prediction output, and a reduced-order model-assisted extended state observer (MESO) is devised to estimate the “total disturbance” with less phase lag. Based on the equivalent control structure, its stability, including the perfect delay matching and delay mismatch cases, is theoretically investigated using the zero exclusion criterion. A simple “separation tuning procedure” is also provided, where the ADRC parameters are tuned directly based on the delay-free process, while the FSP filter is specified to achieve desired robustness. Finally, the validity of the proposed method is demonstrated by comparison simulations and field tests in a 1000MW thermal power unit. Note to Practitioners—This article focuses on sulfur dioxide (SO2) control and proposes a new predictor-based ADRC approach for the WFGD process subject to long time delay and unknown disturbances. Taking the WFGD system as an example, it systematically describes the whole process of modeling, control system design and simulation & experimental verification, which can provide an intuitive reference for practitioners. The zero exclusion criterion is used to analyze the closed-loop stability and determine the stable delay range for a given predictor-based ADRC control system. A detailed tuning procedure is also provided, which can simplify the time-consuming tuning process and help engineers grasp the performance and robustness of the control system. It is evident that the proposed method can be extended to a general class of industrial time-delay systems, which facilitates the practitioner’s practice in general single-input-single-output time-delay system control. The field test results demonstrate the superiority of the proposed method and show promise for practical application in other systems.}, archive = {J_TASE}, author = {Zhuo Chen and Yong-Sheng Hao and Zhi-Gang Su and Gang Zhao and Kwang Y. Lee}, doi = {10.1109/TASE.2024.3398776}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3731--3742}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predictor-based active disturbance rejection control of wet flue gas desulfurization system with delay robustness and simplified tuning}, volume = {22}, year = {2025}, }
• (2025). Secure recursive state estimation for singularly perturbed discrete sequential systems under round-robin-like multichannel access policy. TASE, 22, 3719--3730. (www)

  @article{ , abstract = {This paper investigates the secure state estimation problem for time-varying singularly perturbed discrete sequential systems (SPDSSs) under bandwidth-constrained and multichannel-enabled wireless communication environment. Firstly, to avoid the potential data collision as well as fully utilize the available wireless channels, a round-robin-like multichannel access policy (RRL-MAP) is presented for scheduling the transmission of measurement signals. The considered SPDSS is further assumed to be compromised by deception attack, and then the measurement outputs are updated based on the RRL-MAP and the attack driven by a Bernoulli process. In light of the constructed measurement model, distributed recursive state estimators are designed with the proposing of an algorithm that determines the estimators’ gains. Finally, the efficiency of the developed state estimation method for the envisioned SPDSS is verified by both theoretical analysis and simulation experiments.Note to Practitioners—State estimation for SPDSSs is critical given the widespread application of the systems, however it is challenged by limited communication resource and stochastic cyber attacks. Traditional RR protocol is viewed as a desirable method for mitigating the limitation of network bandwidth, but which becomes inefficient under practical multichannel communication scenario. In addition, deception attack is acknowledged as a kind of hardly detected cyber attacks, and will seriously threaten the data integrity. In view of this, this paper presents the RRL-MAP to guarantee efficient data transmission by appropriately scheduling the available communication channels, and uses Bernoulli process to properly describe deception attack. Then, by further considering the structural characteristics of SPDSSs, distributed recursive state estimators are designed to achieve desired system performance. The effectiveness of the estimators is also comprehensively evaluated, which validates the practical applicability of the proposed state estimation method.}, archive = {J_TASE}, author = {Yan Li and Lishuang Wei and Jinliang Liu and Xiangpeng Xie and Engang Tian}, doi = {10.1109/TASE.2024.3398731}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3719--3730}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Secure recursive state estimation for singularly perturbed discrete sequential systems under round-robin-like multichannel access policy}, volume = {22}, year = {2025}, }
• (2025). Effective multi-agent deep reinforcement learning control with relative entropy regularization. TASE, 22, 3704--3718. (www)

  @article{ , abstract = {This paper focused on developing an effective Multi-Agent Reinforcement Learning (MARL) approach that quickly explores optimal control policies of multiple agents through interactions with unknown environments. Multi-Agent Continuous Dynamic Policy Gradient (MACDPP) was proposed to tackle the issues of limited capability and sample efficiency in the current MARL approaches. It alleviates the inconsistency of multiple agents’ policy updates by introducing the relative entropy regularization to the Centralized Training with Decentralized Execution (CTDE) framework with the Actor-Critic (AC) structure. Evaluated by multi-agent cooperation and competition tasks and traditional control tasks including OpenAI benchmarks and robot arm manipulation, MACDPP demonstrates its significant superiority in learning capability and sample efficiency compared with both related multi-agent and widely implemented signal-agent baselines. It converges to 62% higher average return and uses 38% fewer samples compared with the suboptimal baseline over all tasks, indicating the potential of MARL in challenging control scenarios, especially when the number of interactions is limited. The open source code of MACDPP is available at https://github.com/AdrienLin1/MACDPP.Note to Practitioners—Learning proper cooperation strategy over multiple agents in complicated systems has been a challenge in the domain of Reinforcement Learning. Our work extends the traditional MARL approach FKDPP that has been successfully implemented in the real-world chemical plant by Yokogawa to the CTDE framework and AC structure that supports continuous actions. This extension significantly expands its range of applications from cooperative/competitive tasks to the joint control of one complex system while maintaining its effectiveness.}, archive = {J_TASE}, author = {Chenyang Miao and Yunduan Cui and Huiyun Li and Xinyu Wu}, doi = {10.1109/TASE.2024.3398712}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3704--3718}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Effective multi-agent deep reinforcement learning control with relative entropy regularization}, volume = {22}, year = {2025}, }
• (2025). Iterative planning for multi-agent systems: An application in energy-aware UAV-UGV cooperative task site assignments. TASE, 22, 3685--3703. (www)

  @article{ , abstract = {This paper presents an iterative planning framework for multi-agent systems with hybrid state spaces. The framework uses transition systems to mathematically represent planning tasks and employs multiple solvers to iteratively improve the plan until computational resources are exhausted. When integrating different solvers for iterative planning, we establish theoretical guarantees for recursive feasibility. The proposed framework enables continual improvement of solutions to reduce sub-optimality, efficiently using allocated computational resources. The proposed method is validated by applying it to an energy-aware UAV-UGV cooperative task site assignment problem. The results demonstrate continual solution improvement while preserving real-time implementation ability compared to algorithms proposed in the literature.Note to Practitioners—This paper presents an iterative planning solution for cooperative planning problems in multi-agent systems, which integrates multiple solvers to create an optimization framework. The proposed planning framework has been theoretically validated and applied in an energy-aware cooperative planning scenario for multi-vehicle task site assignments. The proposed framework can be applied to plan for any generalized task site assignment using multiple solvers iteratively.}, archive = {J_TASE}, author = {Neelanga Thelasingha and A. Agung Julius and James Humann and Jean-Paul Reddinger and James Dotterweich and Marshal Childers}, doi = {10.1109/TASE.2024.3398663}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3685--3703}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Iterative planning for multi-agent systems: An application in energy-aware UAV-UGV cooperative task site assignments}, volume = {22}, year = {2025}, }
• (2025). Adaptive learning control for DPS with continually emerging operational conditions. TASE, 22, 3671--3684. (www)

  @article{ , abstract = {During the operation of a distributed parameter system (DPS), its working conditions typically undergo dynamic changes. Although online learning methods can enable models to adapt to new working conditions to some extent, they often confront the “catastrophic forgetting” problem, where the updated model forgets historical working conditions. On the other hand, only a few new samples can be collected during online operation, and the sparse samples make it difficult to establish accurate models for new working conditions. Therefore, achieving precise control under full working conditions remains a challenging problem. To address these challenges, this paper proposes an adaptive predictive control method based on continuous learning that achieves stable control under full working conditions by continuously identifying working conditions and triggering adaptive model updates in real time. Specifically, a spatial-temporal feature-based working condition identification method is first proposed to identify changes in working conditions automatically. Then, to address the challenge of limited data for model updating, a parameter transfer method is proposed. Simultaneously, to ensure that the updated model retains the ability to characterize historical working conditions, an Elastic Weight Consolidation (EWC) constraint is incorporated into the loss function, thus overcoming the catastrophic forgetting problem, and ensuring the updated model can represent both the historical and new working conditions. Finally, by incorporating this condition identification mechanism and adaptive predictive model into the model predictive control framework, continuous precise control of DPS can be achieved. To demonstrate the superiority of the proposed method, extensive experiments are designed. Experimental results show that the proposed method can accurately identify new working conditions and learn new condition models with only a few online samples while overcoming model mismatch of historical conditions, ultimately achieving high-precision control under full working conditions. Note to Practitioners—Motivated by the fact that DPS often operates in different conditions and online learning methods confront “catastrophic forgetting” and “sparse sample training” problems, this paper proposes an adaptive learning control method. The proposed method can learn new working condition features with small online samples while retaining the ability to represent historical working conditions, and thus achieves accurate control under full working conditions.}, archive = {J_TASE}, author = {Yunfeng Zhang and Chunhua Yang and Keke Huang and Dehao Wu and Gui Gui and Weihua Gui}, doi = {10.1109/TASE.2024.3398655}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3671--3684}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive learning control for DPS with continually emerging operational conditions}, volume = {22}, year = {2025}, }
• (2025). Predefined-time control for multi-agent systems with input saturation: An improved dynamic surface control scheme. TASE, 22, 3661--3670. (www)

  @article{ , abstract = {This paper researches the predefined-time fuzzy adaptive dynamics surface consensus control problem for nonlinear multi-agent systems with input saturation. With regard to nonlinear functions existing in the control systems, fuzzy logic system is employed to estimated them. A novel improved predefined-time dynamics surface filter is designed that can avoid the explosion of complexity problem, and the proposed filter satisfies predefined-time stable simultaneously. With the support of piecewise function, the singularity problem that may exist in virtual controller can be dodged greatly. Furthermore, a ameliorative predefined-time auxiliary dynamic system is presented to cope with input saturation. Combining adaptive backstepping control and predefined-time theory, a predefined-time adaptive fuzzy dynamics surface controller is presented that can assure the systems are predefined-time bounded, though the followers exists in the control saturation. Note to Practitioners—Numerous actual physical systems and devices can be modeled as uncertain nonlinear MAS. Furthermore, MAS also can be widely applied to disaster relief, spacecraft and multitudinous fields. On the one hand, the practical systems may exist in input saturation phenomenon due to human factors and in most of the relevant literatures, which can affect the system performance or bring about instability. On the other hand, the initial values of actual systems usually cannot be chosen freely because of the influence of environment and other factors. In addition, it is currently ordinarily supposed that realize the stabilization of controlled systems when time approaches infinity. Consequently, a predefined-time adaptive fuzzy controller is presented for MAS with control saturation, which can achieve the predefined-time stabilization of MAS.}, archive = {J_TASE}, author = {Jiawei Ma and Huaguang Zhang and Juan Zhang and Xiyue Guo}, doi = {10.1109/TASE.2024.3398629}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3661--3670}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Predefined-time control for multi-agent systems with input saturation: An improved dynamic surface control scheme}, volume = {22}, year = {2025}, }
• (2025). GTSCalib: Generalized target segmentation for target-based extrinsic calibration of non-repetitive scanning LiDAR and camera. TASE, 22, 3648--3660. (www)

  @article{ , abstract = {Existing target segmentation methods are typically considered easy to implement and yield satisfactory results. However, they generally cannot adapt to a new environment without tiring parameter tuning, which leads to poor performance, including issues such as over-segmentation, under-segmentation, missing segmentation, and false positives. To avoid the wrong segmentation in a parameter-tuning-free and user-friendly fashion, we propose a generalized target segmentation (GTS) method based on the image-view representation of point clouds. Specifically, the method avoids devising a Euclidean space-based algorithm that is sensitive to surrounding objects and to the varied point cloud density and intensity in a new environment. The target segment produced by GTS can be used with any target-based extrinsic calibration architecture, based on which this paper further proposes a generalized target-based (in this case, chessboard) extrinsic calibration framework called GTSCalib for a non-repetitive scanning LiDAR and a camera. GTSCalib additionally introduces a novel intensity threshold method based on kernel density estimation (KDE) for 3D corner detection and the SQPnP solver for optimization to achieve more generalized and robust performance. Extensive simulations and experiments demonstrate that GTSCalib has high generalization ability, robustness, and accuracy. The code is released at https://github.com/Natsu-Akatsuki/GTSCalib. Note to Practitioners—Calibration is necessary for many non-repetitive scanning LiDAR-camera systems to enable sensor fusion in the fields of mapping, localization, and perception. Unfortunately, existing target-based (in this case, chessboard) calibration methods are weakly adaptable to the surrounding environment with variable density or intensity of point clouds, resulting in unstable performance, particularly for the target segmentation submodule. To solve this problem, we introduce a new target segmentation approach, GTS, and a more generalized and robust extrinsic calibration framework, GTSCalib. The proposed GTSCalib is very suitable for practitioners looking for a robust and accurate target-based calibration without limits on the target’s pose or its surrounding environment and without the need for time-consuming parameter tuning.}, archive = {J_TASE}, author = {Hongqian Huang and Meng Zhang and Lin Li and Jianchen Hu and Hesheng Wang}, doi = {10.1109/TASE.2024.3398623}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3648--3660}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GTSCalib: Generalized target segmentation for target-based extrinsic calibration of non-repetitive scanning LiDAR and camera}, volume = {22}, year = {2025}, }
• (2025). A redundant parallel continuum manipulator with stiffness-varying and force-sensing capability. TASE, 22, 3635--3647. (www)

  @article{ , abstract = {This paper presents the design, analysis, and validation of a novel redundant planar parallel continuum manipulator (PCM) consisting of four flexible links coupled at the rigid mid- and end-platform. To address complex geometry/static hybrid constraints at the rigid platforms, a general framework is developed for the kinetostatics modeling and analysis. Benefiting from the redundant actuation design, the Cartesian stiffness of the studied PCM can be further adjusted as a secondary task to positioning. Utilizing the proposed deflection-based force sensing method, the external load exerted on the end-effector can also be identified by measuring the pose of rigid platforms. To validate the proposed design, a prototype is built, on which a series of experiments have been conducted. The results show that, with the proposed kinetostatics models, the prototype exhibits a mean position and orientation error of 0.52 mm and 0.41°. Finally, the capability of stiffness-varying and force-sensing is demonstrated to verify the feasibility and potential applications of the designed redundant PCM.Note to Practitioners—This paper is motivated by the need for more adaptable robotic systems in industrial and interactive settings. In this work, a novel redundant parallel continuum manipulator with variable passive compliance and force-sensing capability is presented. This planar manipulator comprises four flexible limbs, each driven by independent linear modules, thus the redundancy in the actuation can be leveraged for varying Cartesian stiffness. By means of measuring the pose of the end-effector, the interaction force exerted on the end-effector can be identified according to the force sensing model. The structural compliance of the manipulator enhances safety in human-robot collaboration, while variable stiffness accommodates the specific requirements of different interaction tasks. Furthermore, the force-sensing capability allows for precise feedback control, significantly improving safety during interactions with humans. Therefore, the presented PCM has the potential for interactive manipulation applications, including delicate and fragile object handling, as well as human-robot collaboration in assembly tasks.}, archive = {J_TASE}, author = {Shujie Tang and Zhenkun Liang and Zhuang Zhang and Xuyang Duan and Hao Wang and Genliang Chen}, doi = {10.1109/TASE.2024.3398214}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3635--3647}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A redundant parallel continuum manipulator with stiffness-varying and force-sensing capability}, volume = {22}, year = {2025}, }
• (2025). A user based HVAC system management through blockchain technology and model predictive control. TASE, 22, 3621--3634. (www)

  @article{ , abstract = {This paper introduces an innovative approach to designing a user-based Heating, Ventilation, and Air-Conditioning (HVAC) system management connected with the District Energy Management System. By classifying the users into dynamic energy consumption classes to reward energy efficiency and penalize excessive use, users can modify their behavior to pass to a less expensive and more virtuous consumption class. To this aim, a blockchain platform determines the rewards and penalties and, by a K-means clustering algorithm, categorizes users into respective groups. Then, a Class Follower Problem is formulated and solved by a Model Predictive Control (MPC) strategy integrated with a Long Short-Term Memory network as a predictive model. If the users follow the suggestions proposed by the controller, i.e., the thermostat set-points and the time intervals in which the HVAC system must be switched off or on, the users can be located in a more virtuous consumption class. A case study conducted within an energy district in Bari (Italy) shows how the proposed architectural framework tuned thermal regulation in intelligent buildings while concurrently achieving energy optimization. Note to Practitioners —This paper addresses the challenge of efficiently managing HVAC systems in smart districts through a novel blockchain-based framework and an optimization strategy solved by an MPC approach. The objective is to incentivize users to optimize their energy consumption by introducing dynamic Consumption Classes that reward energy efficiency and penalize inefficient utilization. For practitioners, this strategy translates to a granular level of energy management that not only adapts to individual behaviors but also aligns with broader sustainability goals. Integrating the blockchain platform ensures a transparent and secure method for managing and recording energy usage. At the same time, adopting MPC with Long Short-Term Memory Networks offers accurate forecasts and adjustments to enhance system responsiveness. Although the study focuses on HVAC systems, the principles may be extended to other energy-intensive applications, providing a comprehensive tool for energy management and user engagement in smart cities. Future research could integrate renewable energy sources and explore the implications of user-driven adjustments on the overall energy distribution and efficiency.}, archive = {J_TASE}, author = {Giuseppe Olivieri and Gaetano Volpe and Agostino Marcello Mangini and Maria Pia Fanti}, doi = {10.1109/TASE.2024.3397561}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3621--3634}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A user based HVAC system management through blockchain technology and model predictive control}, volume = {22}, year = {2025}, }
• (2025). FAS-based anti-disturbance stabilization control of nonholonomic systems: Theory and experiment. TASE, 22, 3608--3620. (www)

  @article{ , abstract = {This article considers the control issue for chained nonholonomic systems in the presence of disturbances and model uncertainties. The original system is a cascaded form with two subsystems. In the proposed algorithm, the first subsystem is designed to be exponentially stable, while the second is reformulated into a fully actuated system (FAS) with some states remaining unavailable. Consequently, state feedback control of the original system is transformed into output feedback control of FAS. An effective observer based on Fourier transform technique is presented to simultaneously estimate states and disturbances. Then, output feedback control and relay-switching-based control are proposed to achieve exponential or asymptotical convergence of the states. Finally, simulation and experiment results are presented to illustrate the efficacy of the proposed strategy. Note to Practitioners—Nonholonomic systems, exemplified by wheeled robots, are ubiquitous in engineering applications. This paper is motivated by the practical requirements of the stabilization control tailored to such systems. Nonholonomic systems often face complex scenarios with disturbances. For example, when a wheeled robot is parking in the snow fields, skidding and slipping may occur. Handling uncertainty and disturbance increases the challenge to the control task. Furthermore, onboard computers typically have limited computing power, thereby requiring straightforward and concise control algorithms. To meet these requirements, this paper introduces the framework of FAS, which eliminates the need for conventional recursive design procedures. Consequently, two controllers are presented: the output-feedback-based controller and the state-feedback-based controller. Both two controllers offer robustness, low complexity, and strong ability in disturbance rejection, making them suitable for practical applications. On the other hand, the controllers can not only be deployed independently to meet specific requirements, but can also be combined into a relay-switching-based strategy. This strategy is a two-stage control scheme. In the first stage, accurate parameter identification is performed while the output feedback controller maintains primary control performance. Once precise parameters are obtained, the output feedback controller seamlessly transitions to the state feedback controller, achieving higher control accuracy. Overall, this approach balances the control effectiveness with the computational efficiency, addressing the practical needs of stabilization control in nonholonomic systems.}, archive = {J_TASE}, author = {Yang Gao and Zhongcai Zhang and Peng Huang and Yuqiang Wu}, doi = {10.1109/TASE.2024.3397316}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3608--3620}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {FAS-based anti-disturbance stabilization control of nonholonomic systems: Theory and experiment}, volume = {22}, year = {2025}, }
• (2025). Transient-state control for linear switched aero-engine models with input saturation utilizing input-to-state stability theory. TASE, 22, 3598--3607. (www)

  @article{ , abstract = {This paper studies the transient-state control of linear switched aero-engines with input saturation. Firstly, by formulating the problem as an input-to-state stability (ISS) property investigation of switched linear aero-engine models with saturated control input, a saturated controller coupled with an event-triggered mechanism (ETM) avoiding Zeno and asynchronous phenomena is developed. It shows that the ISS property of switched linear aero-engine models can be guaranteed utilizing the designed strategy. Then, in order to estimate the disturbance bound and the initial value domain, an optimization problem is proposed. Finally, an application study based on the GE-90K aero-engine is carried out to reveal the effectiveness of the proposed strategy. Simulation results show that the aero-engine can be transitioned to a desired steady-state point under the designed controller. In addition, comparisons and discussions demonstrate the advantages and potential significance of the work in practice. Note to Practitioners—Transient-state control is one of the most important problems in aero-engine control and affects the maneuverability of aircrafts. In addition, the saturation of controllers may lead to bad control performance. Thus, this paper is devoted to proposing a new control method for linear switched aero-engine models subject to input saturation. By designing a new controller coupled with an event-triggered mechanism, better transient-state performance compared with some existing methods, can be guaranteed.}, archive = {J_TASE}, author = {Peng Li and Ning Xu and Shuoshuo Liu and Xudong Zhao}, doi = {10.1109/TASE.2024.3396131}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3598--3607}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Transient-state control for linear switched aero-engine models with input saturation utilizing input-to-state stability theory}, volume = {22}, year = {2025}, }
• (2025). MF-ID: A benchmark and approach for multi-category fine-grained intrusion detection. TASE, 22, 3582--3597. (www)

  @article{ , abstract = {With the development of computer vision, the task of vision-based intrusion detection has been widely applied to various important fields such as intelligent monitoring, autonomous driving, and security. Previous vision-based intrusion detection tasks aim at detecting whether pedestrians invade a restricted Area-of-Interest (AoI) from a static or dynamic view. However, for real application scenarios, we need to detect whether various types of intrusion objects, not just pedestrians, intrude into the AoI of dynamic view, and the fine-grained categories of intrusion objects need to be accurately given, such a dynamic-view multi-category fine-grained intrusion detection (namely MF-ID) task is important but not yet explored. In this paper, we propose a new benchmark and approach to address this task. Firstly, due to the current lack of relevant benchmark, we develop a new publicly available dataset Cityintrusion-Multicategory, conduct statistical analysis on this dataset, and design three evaluation metrics. Secondly, we propose an end-to-end framework MF-YOLOV5, with five improvements: (1) We modify YOLOV5 to make it more suitable for our task, with a lower branch for object detection and an upper branch for segmenting AoI. (2) A new multi-category fine-grained loss (MFLoss) is designed to improve the fine-grained classification capability. (3) We improve the YOLOV5 C3 modules by enhancing the ability of cross-channel interaction. (4) A detection layer for the tiny objects is integrated into the network to improve its ability of detecting tiny objects. (5) A lightweight module with bottleneck transformer is introduced to reduce the network parameters. Finally, comprehensive experiments and comparisons demonstrate the validity of the proposed approach, and MF-YOLOV5 can reach the level of current SOTA, with 97.46% Miou, 55.29% Map@.5 and 42.97% MF-ID Acc. The relevant datasets and codes are available at https://ieee-dataport.org/documents/mf-id-1. Note to Practitioners—The motivation of this paper is to address the pitfall of the dynamic-view intrusion detection system. The existing methods mainly focus on pedestrian intrusion detection in dynamic view, ignoring the more practical and valuable task of multi-category fine-grained intrusion detection (MF-ID). Based on this, we propose a new benchmark and an advanced approach to meet the requirement of MF-ID task in dynamic view. Extensive experiments show that the proposed approach can not only reach promising performance but also maintain a high real-time intrusion detection speed. The proposed approach can be deployed in realistic scenes, e.g., autonomous driving, intelligent monitoring, security, and intelligent transportation management. In future research, we will explore a more comprehensive benchmark and more efficient approach for intrusion detection.}, archive = {J_TASE}, author = {Fujun Han and Peng Ye and Ke Li and Shukai Duan and Lidan Wang}, doi = {10.1109/TASE.2024.3396085}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3582--3597}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {MF-ID: A benchmark and approach for multi-category fine-grained intrusion detection}, volume = {22}, year = {2025}, }
• (2025). Trustworthy traffic sensors networks via fault-tolerant sensor reconciliation architectures. TASE, 22, 3567--3581. (www)

  @article{ , abstract = {In this paper, the sensor selection problem (i.e. the problem of selecting a subset of $p\ll n$ possibly redundant sensors from a larger set of n potential sensors) for large-scale Wireless Sensor Networks (WSNs) is considered along with the concomitant design problem of determining a suitable fault-tolerant full state observer for effective state reconstruction in the presence of sensor faults/failures. Given a set of redundant physical sensors, a monitoring architecture for estimating traffic flows via a novel fault-tolerant sensor reconciliation design procedure is proposed that is able to hide measurements that may be corrupted by sensor errors and provide reliable state estimates.The proposed approach utilizes a Linear Parameter Varying (LPV) Luenberger observer, which is responsible for robust and reliable system state reconstruction. The results show that the proposed estimation strategy has superior performance in both minimizing state reconstruction errors and estimating bias errors and sensor effectiveness when compared to conventional strategies and with another state-of-the-art approach exploiting Takagi-Sugeno (TS) fuzzy observers. Finally, the effectiveness of the reconciliation architecture has been evaluated in realistic road infrastructure and traffic data provided by the software Aimsun Next.Note to Practitioners—This paper is based on the results of a recently completed R&D project, SCALS (Smart Cities Adaptive Lighting System), which aimed at developing all the hardware/software components of an adaptive urban smart lighting architecture, capable of autonomously adjusting the brightness of the street lamps based on the traffic flow in specific areas of interest in order to reduce the energy consumption. In order to have enough flexibility, the road networks is split is small segments over which the lamps are operated identically by a single traffic sensor. In this context, because the number of road segments can be huge in practical situations, the selection of an appropriate (small) number of physical sensors (e.g. cameras) for traffic monitoring is a critical issue to ensure reliable estimates and unnecessary installation and maintenance costs. Moreover, unavoidable failures may affect the deployed sensors and degrading the system performance. With reference to fault-tolerance approaches in WSNs, most research and solutions mainly focus on the duplication of physical redundancy of sensors. On the contrary, the focus in this work is on the possibility of designing traffic flow estimation architectures that exploits a limited number of sensors and is resilient despite of the presence of faulty measurements. The proposed solution can be considered as a simple and practical approach to minimize the technological cost by reducing the number of sensors deployed and ensure reliable state estimation even in the presence of sensor failures.}, archive = {J_TASE}, author = {Gianfranco Gagliardi and Vincenzo D’Angelo and Franco Angelo Torchiaro and Alessandro Casavola}, doi = {10.1109/TASE.2024.3395980}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3567--3581}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Trustworthy traffic sensors networks via fault-tolerant sensor reconciliation architectures}, volume = {22}, year = {2025}, }
• (2025). A hybrid framework based on bio-signal and built-in force sensor for human-robot active co-carrying. TASE, 22, 3553--3566. (www)

  @article{ , abstract = {Human-robot collaboration represents a promising avenue for applications in future factory scenarios. Existing work predominantly concentrates on passive assistance based on real-time sensing sensor like force sensors, while only a few studies have ventured into the exploration and realization of active assistance provided by robots with the assistance of predictive sensors. This paper proposes an innovative hybrid framework that combines human bio-signal information with built-in robot force sensors to implement human-robot active collaboration. First, a Hill-type muscle-skeleton model is adopted and calibrated through partial swarm optimization (PSO). With this model, surface electromyography(sEMG) is used to estimate human limb stiffness. Then, an Radial Basis Function neural network (RBFNN) compensator is developed to account for the uncertainty in human-object-robot dynamics. Subsequently, we propose an adaptive variable impedance controller, incorporating a global bias into the neural network architecture. This innovative modification serves to augment the system’s robustness, streamline the network configuration by curtailing the number of hidden neurons, and consequently, facilitate more consistent and efficient human-robot interaction behavior. Finally, we substantiate the effectiveness of the proposed methodology through a two-link robotic simulation experiment and a real-world co-carrying task employing with the Baxter robot and human partner. These rigorous evaluations unveil a significant alleviation of task-related human workload attributed to our proposed framework. Note to Practitioners—This framework aims to address the existing research gap in human-robot collaboration, particularly involving bio-signal utilization, to facilitate perceptive active assistance within a typical industrial assembly scenario. In such representative tasks, many studies primarily employ real-time sensors such as force, position. These sensors, while essential, are limited by their detection principles and require collaborative operation to ascertain stiffness and realize passive assistance. Conversely, bio-signals intrinsically contain stiffness information and exhibit prospective characteristics that can be leveraged for stiffness prediction. In this typical task, we design a human-robot co-transport system with two crucial characteristics: first, the robot is capable of detecting human stiffness tendencies and comprehending human intent, leading to self-adjusting robotic behavior that provides enhanced protection for the transported object. Secondly, the newly proposed controller can manage sudden disturbances and execute self-repairs, thus increasing the task success rate and ensuring worker safety.}, archive = {J_TASE}, author = {Leyun Hu and Di-Hua Zhai and Dongdong Yu and Yuanqing Xia}, doi = {10.1109/TASE.2024.3395921}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3553--3566}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hybrid framework based on bio-signal and built-in force sensor for human-robot active co-carrying}, volume = {22}, year = {2025}, }
• (2025). Performance of DVR using diffusion norm penalized LMS fourth adaption algorithm with optimized FOPID gains. TASE, 22, 3543--3552. (www)

  @article{ , abstract = {This paper deals with the integrated control scheme of adapt then combine (ATC) diffusion penalized least mean square fourth (DPLMSF) algorithm and optimized fractional order proportional integral derivative (FOPID). The estimation of the fundamental component (FC) from the non-ideal input voltage is extracted by the proposed ATC-DPLMSF algorithm, which employs a penalized Lp norm in the cost function that minimizes the error utilizing the mean-fourth error of a diffusion strategy. This enhances the control scheme performance by reducing the error rate with increased convergence rate. The key contribution of the proposed approaches is adaptiveness of mixed step-size selection based on hybrid LMSF algorithm and reduced overshoot and undershoot oscillation in a dc link and ac link with optimized FOPID. These features enable accurate and fast estimation of fundamental component at abnormal grid condition. To address the error minimization, integral time square error (ITSE) is employed as a cost function by utilizing the features of the self-adaptive population (SAP)-Rao algorithm. The SAP-Rao has an inbuilt feature of self-adjusting the control parameters, which avoids local entrapment. The recommended control algorithm enhances performance of three single phases VSC-based DVR using the d-SPACE board.Note to Practitioners—In contrast to the traditional adaptive least mean square (LMS) algorithm, that employs exclusively mean square error (MSE) as a function of cost based on second order power error, this paper examines the efficacy of combined Least Mean Square/Fourth (LMS/LMF). This restricts the ability to separate the fundamental data from the noise and causes performance reduction in some dynamic situations. In order to solve this issue, a mixed norm-based step size adaptive algorithm is developed and used to estimate the basic components from the voltage of the contaminated grid. By utilising mixed LMS and LMF to enhance the error convergence quality of a filter, the proposed method’s convergence speed rate can be enhanced. Second, optimised fractional order proportional integral derivative (FOPID) acquires control voltage. Secondly, voltage regulation is performed by optimized fractional order proportional integral derivative (FOPID) gains, which can rapidly identify voltage anomalies and stabilize the voltage error with two additional “tuning gains” which provides more tuning freedom. This self-tuned FOPID controller can obtain a precise voltage regulation and with respect to industrial use the FOPID controllers has the immense benefit of improving the transitory response time over conventional PID/PI controllers in terms of fast settle time (Ts), lesser peak time (Tp), and less maximum overshoot (Mp%). The performance can be assessed by using ITSE as an objective function. Experimental and simulations results, confirms the superior and dominance performance of the pproposed control scheme weighted diffusion LMS/F and fractional order PID controller which increases its implementation and interests of the industry.}, archive = {J_TASE}, author = {Prashant Kumar and Sabha Raj Arya and Khyati D. Mistry and Papia Ray}, doi = {10.1109/TASE.2024.3395763}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3543--3552}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Performance of DVR using diffusion norm penalized LMS fourth adaption algorithm with optimized FOPID gains}, volume = {22}, year = {2025}, }
• (2025). Prototype, modeling, and control of aerial robots with physical interaction: A review. TASE, 22, 3528--3542. (www)

  @article{ , abstract = {This article aims to investigate the research achievements related to aerial robots with physical interaction. Various morphologies of aerial physical interaction (APhI) robot prototypes with fixed wing, flapping wing, single main rotor, conventional underactuated multirotor, fully actuated multirotor, even deformed multirotor, and multiple platforms are reviewed for different APhI tasks associated with momentary, loose, and strong interaction coupling. This review also covers APhI robot rigid dynamics and robot-environment coupled interaction dynamics modeling methods, interaction wrench measurement/estimation, decoupled and coupled control, active aerial interaction control, and task-constrained planning approaches. Finally, future development directions and prospects are initially anticipated for aerial robots with physical interaction.Note to Practitioners—Aerial physical interaction (APhI) has been a hot topic in the field of aerial robots in recent years, which is a reflection of the advanced capabilities of aerial robots. However, APhI robots face challenges such as difficulty in flight stability and weak adaptability to dynamic environments while exerting active influence on environments. Under this background, this review aims to offer a reference for researchers and practitioners engaged in the related field from the aspects of system design, modeling, control, and task-constrained planning, which hopes to help them apply APhI robots to polar scientific expeditions, complex environment sampling, infrastructure inspection and maintenance, and other application areas. Further, this review also highlights the design idea of rigid-soft integrated APhI robots from the perspective of design-mechanism-performance to enhance interaction stability and safety.}, archive = {J_TASE}, author = {Hang Zhong and Jiacheng Liang and Yanjie Chen and Hui Zhang and Jianxu Mao and Yaonan Wang}, doi = {10.1109/TASE.2024.3395658}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3528--3542}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prototype, modeling, and control of aerial robots with physical interaction: A review}, volume = {22}, year = {2025}, }
• (2025). Hierarchical uncertainty characterization of monthly net load in renewable power systems. TASE, 22, 3515--3527. (www)

  @article{ , abstract = {The accurate characterization of net load uncertainty can ensure the economy and stability of renewable power systems operation. In existing studies, Gaussian mixture model (GMM) and Dirichlet process mixture model (DPMM) are powerful tools for characterizing the uncertainty of monthly net load. However, the correlation among the Gaussian components and the common distribution characteristics among the monthly net load are not considered in these studies, which may lead to lower characterization accuracy. To solve these issues, we propose a hierarchical uncertainty characterization method for monthly net load considering Gaussian component reduction and correlation. Firstly, the method obtains the Gaussian mixture components and the label information of the annual net load by the Expectation-Maximum (EM) algorithm. Then, the Gaussian components of the monthly net load are corrected by the label information, which helps extract the common distribution characteristics among the monthly net load. On this basis, the corrected Gaussian components are considered as the base distribution. A component-preserving Expectation-Maximum (CPEM) algorithm is developed for component reduction. This realizes the uncertainty characterization of the monthly net load with high accuracy and lower time consumption. Importantly, the temporal correlation of net load is converted into the correlation among the Gaussian components, which are explicitly characterized by the Spearman coefficient. Finally, the superiority of the proposed method is verified with actual data collected in Australia. Note to Practitioners—In this article, we address the issue of monthly net load uncertainty characterization in renewable power systems. A hierarchical uncertainty characterization method is proposed to enable the distribution characteristic extraction of monthly net load. Most existing works on uncertainty characterization use GMM and DPMM, where the internal information in the net load data (e.g., distribution and component correlation information) is poorly utilized. This poses a considerable challenge for accurately analyzing the net load uncertainty. To this end, the article proposes a hierarchical uncertainty characterization method for monthly net load considering Gaussian component reduction and correlation. In this method, the common and typical distribution characteristics of monthly net load are considered by the CPEM algorithm. Moreover, the temporal information of the net load is transformed as the correlation among the Gaussian components, which are integrated into the E-step in the proposed CPEM algorithm. Since more internal information is utilized, the proposed method is helpful in characterizing the net load uncertainty by means of high accuracy and lower time consumption. The results of the uncertainty characterization can be readily implemented in the planning operation and real-time dispatch of renewable power systems.}, archive = {J_TASE}, author = {Xiao Yang and Yuanzheng Li and Yong Zhao and Guokai Hao and Takayuki Ishizaki and Zhigang Zeng}, doi = {10.1109/TASE.2024.3395634}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3515--3527}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical uncertainty characterization of monthly net load in renewable power systems}, volume = {22}, year = {2025}, }
• (2025). Linearized data center workload and cooling management. TASE, 22, 3502--3514. (www)

  @article{ , abstract = {With the current high levels of energy consumption of data centers, reducing power consumption by even a small percentage is beneficial. We propose a framework for thermal-aware workload distribution in a data center to reduce cooling power consumption. The framework includes linearization of the general optimization problem and proposing a heuristic to approximate the solution for the resulting Mixed Integer Linear Programming (MILP) problems. We first define a general nonlinear power optimization problem including several cooling parameters, heat recirculation effects, and constraints on server temperatures. We propose to study a linearized version of the problem, which is easier to analyze. As an energy saving scenario and as a proof of concept for our approach, we also consider the possibility that the red-line temperature for idle servers is higher than that for busy servers. For the resulting MILP problem, we propose a heuristic for intelligent rounding of the fractional solution. Through numerical simulations, we compare our heuristics with several existing algorithms. In addition, we evaluate the performance of the solution of the linearized system on the original system. Finally, the results show that the proposed approach can reduce the cooling power consumption by more than 10 percent compared to the case of continuous utilizations and a single red-line temperature. Note to Practitioners—We present a holistic approach for thermal-aware workload distribution for power consumption reduction in data centers. We suggest that when thermal and power consumption models can be linearized, a model-independent approach can be used for optimization purposes. The standard linear problem that results presents some technical challenges to solve, for which we present intuitive and effective solution heuristics. The heuristics are simple enough that they could be used for real-time calculations. The result is that customized models and problems can be avoided (a linear model could be directly constructed from operational data, if available), allowing for the simplification of operational control problems. Our approach is evaluated for a high-fidelity model of a real data center, where both the linearization and optimization components are validated. Finally, we show how this approach can be used to effectively solve the operational problem of workload distribution in the presence of utilization-dependent server red-line temperatures.}, archive = {J_TASE}, author = {Somayye Rostami and Douglas G. Down and George Karakostas}, doi = {10.1109/TASE.2024.3395471}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3502--3514}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Linearized data center workload and cooling management}, volume = {22}, year = {2025}, }
• (2025). A search and detection autonomous drone system: From design to implementation. TASE, 22, 3485--3501. (www)

  @article{ , abstract = {Utilizing autonomous drones or unmanned aerial vehicles (UAVs) has shown great advantages over preceding methods in support of urgent scenarios such as search and rescue (SAR) and wildfire detection. In these operations, search efficiency in terms of the amount of time spent to find the target is crucial since with time the survivability of the missing person decreases or wildfire management becomes more difficult with disastrous consequences. In this work, we consider the scenario where a drone is intended to search and detect a missing person (e.g., a hiker or a mountaineer) or a potential fire spot in a given area. To obtain the shortest path to the target, a general framework is provided to model the problem of target detection when the target’s location is probabilistically known. To this end, two algorithms are proposed: Path planning and target detection. The path planning algorithm is based on Bayesian inference and the target detection is accomplished by using a residual neural network (ResNet) trained on the image dataset captured by the drone as well as existing pictures and datasets on the web. Through simulation and experiment, the proposed path planning algorithm is compared with two benchmark algorithms. It is shown that the proposed algorithm significantly decreases the average time of the mission.Note to Practitioners—This article is motivated by the need for an efficient path-planning algorithm for drones during specific SAR operations. In particular, situations where someone is lost in a snow-covered hike and a fire spot that is in its initial levels are of interest. In fact, since the target location is not known, it is required that the UAV be able to efficiently search the entire area until it finds the target in the shortest possible time. The proposed Bayesian framework along with the ResNet learning algorithm shows an efficient performance in terms of average time duration and accuracy, respectively. The framework developed in this paper can be extended to a multi-UAV scenario where UAVs coordinate to optimize the overall performance.}, archive = {J_TASE}, author = {Mohammadjavad Khosravi and Rushiv Arora and Saeede Enayati and Hossein Pishro-Nik}, doi = {10.1109/TASE.2024.3395409}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3485--3501}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A search and detection autonomous drone system: From design to implementation}, volume = {22}, year = {2025}, }
• (2025). Multi-stage network attack detection algorithm based on gaussian mixture hidden markov model and transfer learning. TASE, 22, 3470--3484. (www)

  @article{ , abstract = {Multi-stage network attack (MSA) is a serious threat to data security. The high-dimensionality of the alert data along with the diverse features, leads to poor detection performance for MSA. Consequently, this paper proposes a multi-stage network attack detection algorithm based on Gaussian mixture hidden Markov model and transfer learning. Firstly, a sequence modeling framework of Gaussian mixture hidden Markov models is proposed. It uses a Gaussian mixture model to cluster high-dimensional alert data and a hidden Markov model to fully consider the temporal structure of MSA, the alert features of each stage, and transitions between stages. Secondly, optimized Baum-Welch and Viterbi algorithms are proposed, combined with the forward-backward algorithm to train the parameter of the Gaussian mixture hidden Markov model and detect the attack sequence of MSA. Finally, an improved transfer learning method is proposed, which addresses the sparsity of labeled data in MSA scenarios, a Kullback-Leibler (KL) divergence value is added as a penalty term to narrow the distribution differences between the source and target domains and solves the bias problem in the transfer learning process. The proposed algorithm is validated on the datasets DARPA 2000 and CSE-CIC-IDS2018, and the effectiveness and superiority is verified on multiple evaluation indicators. Note to Practitioners—Network attacks gradually show the large-scale, coordinated and multi-stage characteristics. Complex multi-step attacks with strong concealment and persistence have become the development trend of network attacks, which seriously threaten and infringe the secure storage and transmission of information. Most existing studies use hidden Markov model (HMM) to model multi-stage network attacks. HMM is usually more suitable for multi-step attacks occurring in a specific sequence within a continuous time interval. However, in actual multi-stage network attacks, attackers do not need to follow the exact sequence of multi-step attacks, and the intervals between successive stages of an attack can be hours, days, or even months. Attackers may also perform interleaved attacks to hide attacks. Therefore, this paper proposes a multi-stage network attack detection algorithm based on Gaussian hybrid hidden Markov and transfer learning. The optimized Gaussian hybrid hidden Markov model is used to model the alert data of multi-stage network attacks, and the improved transfer learning method is adopted to apply the knowledge learned from the source domain to the multi-stage network attack detection model of the target domain. The experimental results show that the proposed algorithm can effectively process the alert data of different attack stages under complex multi-stage network attacks, distinguish the real threat alert, false alert and irrelevant alert, and improve the performance of detecting multi-stage network attacks. The method presented in this paper can provide a valuable solution for complex multi-stage network attack detection such as advanced persistent threat (APT). Future work will further combine adversarial generation network methods to avoid the interference of adversarial attack samples, and explore more ways to improve the performance of multi-step attack detection.}, archive = {J_TASE}, author = {Qian Wang and Weilong Wang and Yan Wang and Jiadong Ren and Bing Zhang}, doi = {10.1109/TASE.2024.3395355}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3470--3484}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-stage network attack detection algorithm based on gaussian mixture hidden markov model and transfer learning}, volume = {22}, year = {2025}, }
• (2025). Fixed-time-synchronized bipartite time-varying formation tracking control of networked euler-lagrange systems. TASE, 22, 3458--3469. (www)

  @article{ , abstract = {This article solves the fixed-time-synchronized (FTS) bipartite time-varying formation (BTVF) tracking control problem of networked Euler-Lagrange systems (NELSs) with disturbances, where all state elements of the NELSs can achieve BTVF at the same time. An improved hierarchical control strategy is proposed, comprising a distributed estimator layer and a local control layer. The former is responsible for estimating the desired position and velocity, while the latter employs a novel interval observer-based algebraic reconstruction mechanism to compensate for the unknown disturbance. Subsequently, a singularity-free FTS BTVF control protocol is established, leveraging a norm-normalized sign function-based switching sliding mode control technique. This protocol guarantees that, in the presence of a signed antagonistic interaction graph, all system states converge synchronously to equilibrium within a fixed time. Finally, comparisons and simulations are conducted to validate the feasibility, significance, and merits of the developed control scheme. Note to Practitioners—In practical applications, the performance of networked systems depends not only on when the system state converges, but also significantly on when and how each subsystem state element converges. Despite the superior convergence property of fixed-time control, its standalone application proves inadequate in scenarios where all state elements of the network systems are required to reach the desired values synchronously, as observed in coordinated transportation and formation, particularly in the domain of space attack and defense. In this article, the BTVF tracking control issue with FTS convergence of networked systems is studied for the first time. The design of non-singular and FTS coordination schemes is challenging in NELSs. In addition, the developed distributed strategy based on the hierarchical control technique is suitable for NELSs with unknown disturbances and antagonistic interaction graphs, and the use of the norm-normalized sign function-based switching sliding mode control method excludes the singular problem.}, archive = {J_TASE}, author = {Li Ma and Fanglai Zhu}, doi = {10.1109/TASE.2024.3395325}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3458--3469}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time-synchronized bipartite time-varying formation tracking control of networked euler-lagrange systems}, volume = {22}, year = {2025}, }
• (2025). A self-learning approach to heterogeneous multi-robot coalition formation under uncertainty. TASE, 22, 3445--3457. (www)

  @article{ , abstract = {Coalition structure is an effective cooperation architecture for task implementation in the field of multi-robot systems. Nevertheless, the presence of uncertainty inherently complicates the decision-making process for robots and may potentially result in suboptimal coordination. To tackle this challenge, this paper considers an uncertain multi-robot coalition formation scenario in which task information (the types of all tasks) is incompletely known to robots. Given the local beliefs of robots, the problem of multi-robot coalition formation under uncertainty is formulated as a coalition formation game. In this game, each robot is a rational and self-interested player and tends to join a coalition according to its preference. A polynomial-time coalition formation algorithm is proposed to identify the social agreement, i.e., Nash stable partition, among the robots. The convergence of the proposed algorithm is strictly guaranteed as long as the communication topology of the considered system is strongly connected. The coalition formation game is then extended to a dynamic game, and we propose a belief updating algorithm that enables robots to update their beliefs as the game is played repeatedly. Simulation results demonstrate the effectiveness of our proposed algorithms, and the robots will eventually learn the true type of each task.Note to Practitioners—The work reported in this article will be beneficial for deploying multi-robot systems to support cooperative surveillance applications. In these scenarios, robots can form stable coalitions to perform surveillance tasks, even when the task information is incompletely known due to sensor noise or limited sensor range. The problem of multi-robot coalition formation is known to be NP-hard, which becomes even more challenging when uncertainty is taken into account. This paper proposes game-based algorithms to solve the problem of multi-robot coalition formation under uncertainty. Some practical schemes are introduced as benchmarks to further illustrate the performance improvement brought by our proposed algorithms. The proposed algorithms are further evaluated through a real-world experiment, demonstrating their effectiveness for practical application.}, archive = {J_TASE}, author = {Xin Huo and Hao Zhang and Zhuping Wang and Chao Huang and Huaicheng Yan}, doi = {10.1109/TASE.2024.3395283}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3445--3457}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A self-learning approach to heterogeneous multi-robot coalition formation under uncertainty}, volume = {22}, year = {2025}, }
• (2025). Game-based optimal aggregation of energy prosumer community with mixed-pricing scheme in two-settlement electricity market. TASE, 22, 3433--3444. (www)

  @article{ , abstract = {In this paper, we consider an electricity market with a batch of energy prosumers (EPs), who can trade energy with the day-ahead market (DAM) and energy balancing market (EBM). To coordinate the behavior of EPs, an energy aggregator (EA) is established, who is responsible for making aggregation strategies for the EPs to improve the benefit of the whole EP community. Specifically, two pricing schemes are provided by the EA: pay-as-you-go (PAYG) scheme and lump sum (LS) scheme, which can be flexibly selected by EPs based on their own preferences. By considering the dual-pricing principle of the EBM, a stochastic Stackelberg game between the EPs and EA is formulated and a two-level optimization algorithm is introduced based on the proposed feasible region partitioning (FRP) method. The performance of the algorithm is demonstrated with a two-settlement market model in the simulation.Note to Practitioners—This work aims to optimize the social cost of an EP community in a two-settlement electricity market. In contrast to the existing works, the main innovations and the resulting challenges lie in both the market modelling and the theoretical approach. In particular, different from the existing uniform-pricing schemes, this work explores the realization of mixed-pricing scheme, which can introduce more customized factors into the decision-making process of EPs. In addition, to adapt to a wider range of practical EBMs, the dual-pricing principle of the balancing energy is considered, which leads to a stochastic Stackelberg game where the social cost function is nondeterministic with respect to the strategy of EA. To address the aforementioned challenges, a two-level optimization algorithm is introduced based on the proposed FRP method, which enriches the existing methods for Stackelberg games. The proposed market model can be more capable of addressing some practical issues in the sense that the mixed-pricing scheme can adapt to the different preferences of market participants and the self-centric instinct of EPs is considered based on Nash games. In addition, the dual-pricing principle of the EBM is motivated by many real electricity markets and can be more general than the single-pricing counterpart.}, archive = {J_TASE}, author = {Jianzheng Wang and Guoqiang Hu}, doi = {10.1109/TASE.2024.3394892}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3433--3444}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Game-based optimal aggregation of energy prosumer community with mixed-pricing scheme in two-settlement electricity market}, volume = {22}, year = {2025}, }
• (2025). G3Reg: Pyramid graph-based global registration using gaussian ellipsoid model. TASE, 22, 3416--3432. (www)

  @article{ , abstract = {This study introduces a novel framework, G3Reg, for fast and robust global registration of LiDAR point clouds. In contrast to conventional complex keypoints and descriptors, we extract fundamental geometric primitives, including planes, clusters, and lines (PCL) from the raw point cloud to obtain low-level semantic segments. Each segment is represented as a unified Gaussian Ellipsoid Model (GEM), using a probability ellipsoid to ensure the ground truth centers are encompassed with a certain degree of probability. Utilizing these GEMs, we present a distrust-and-verify scheme based on a Pyramid Compatibility Graph for Global Registration (PAGOR). Specifically, we establish an upper bound, which can be traversed based on the confidence level for compatibility testing to construct the pyramid graph. Then, we solve multiple maximum cliques (MAC) for each level of the pyramid graph, thus generating the corresponding transformation candidates. In the verification phase, we adopt a precise and efficient metric for point cloud alignment quality, founded on geometric primitives, to identify the optimal candidate. The algorithm’s performance is validated on three publicly available datasets and a self-collected multi-session dataset. Parameter settings remained unchanged during the experiment evaluations. The results exhibit superior robustness and real-time performance of the G3Reg framework compared to state-of-the-art methods. Furthermore, we demonstrate the potential for integrating individual GEM and PAGOR components into other registration frameworks to enhance their efficacy.Note to Practitioners—Our proposed method aims to perform global registration for outdoor LiDAR point clouds. Our methodology, which extracts point cloud segments and utilizes their centers for registration, differs from conventional approaches that rely on keypoints and descriptors. We further propose GEM to model the uncertainty of the centers and embed it into our distrust-and-verify framework. In theory, our method can be applied to any registration task that involves primitives representable as sets of Gaussians or points. Additionally, practitioners should consider the following to enhance applicability. First, practitioners can fine-tune the parameters of the segmentation algorithm to generate more repeatable segmentation results. Second, although our default setting uses four compatibility test thresholds, fewer may suffice, especially when translations between point clouds are minor. Finally, for geometrically uninformative segments such as vegetation, consider extracting descriptors within these segments to increase correspondences.}, archive = {J_TASE}, author = {Zhijian Qiao and Zehuan Yu and Binqian Jiang and Huan Yin and Shaojie Shen}, doi = {10.1109/TASE.2024.3394519}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3416--3432}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {G3Reg: Pyramid graph-based global registration using gaussian ellipsoid model}, volume = {22}, year = {2025}, }
• (2025). A data-driven framework for verified detection of replay attacks on industrial control systems. TASE, 22, 3400--3415. (www)

  @article{ , abstract = {This paper addresses data-driven replay attack detection on industrial control systems. The primary challenge in detection lies in distinguishing replayed sensor measurements from normal measurements using only time series data. This is tackled through a novel two-stage detection and verification framework. The first stage consists of continuous real-time monitoring of sensor measurement patterns using matrix profile based change-point detection, used to indicate a possibility of a replay attack. The second stage verifies the presence of a replay attack by introducing spatial features to newly defined time series data. This is implemented by generating spectrograms of the time series measurements using short-time Fourier transform. Then, the spectrograms are split into image frames to form temporal sequences, creating spatio-temporal features that distinguish replay attacks. To capture both the spatial and temporal features, we utilise a Convolutional Long Short-Term Memory (ConvLSTM) neural network and implement it in an autoencoder architecture, in order to analyse data patterns in an unsupervised manner, where the replay attack is detected based on the reconstruction error. We demonstrate the effectiveness of our framework in the detection of different replay attack scenarios using the Tennessee Eastman process benchmark simulation system/process. Note to Practitioners—This paper is motivated by the importance of cyberattack detection in industrial control systems that are essential for the stable operation of many practical applications, such as in chemical processing and manufacturing plants, and power and water distribution networks. Specifically, replay attack detection using data-driven methods is explored, eliminating the need for an accurate process model which may be tedious to obtain. However, the attack’s implementation using actual/valid operational data to replicate normal behaviour, makes it difficult to detect using basic data-driven methods, resulting in an increased likelihood of false alarms or missed detection. To address this challenge, a two-stage detection and verification framework is proposed. The first stage performs real-time monitoring of sensor measurements using change-point detection on time series data patterns. The second stage verifies the occurrence of a replay attack by introducing spatial features to newly defined time series data. This framework therefore eliminates false/missed detection, and offers practitioners a robust method to enhance security measures in industrial control systems, minimising the risks posed by malicious replay attacks.}, archive = {J_TASE}, author = {Sara Gargoum and Negar Yassaie and Ahmad W. Al-Dabbagh and Chen Feng}, doi = {10.1109/TASE.2024.3394315}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3400--3415}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A data-driven framework for verified detection of replay attacks on industrial control systems}, volume = {22}, year = {2025}, }
• (2025). Autonomous trajectory tracking of bioinspired soft microrollers through a novel direction-speed decoupled control strategy. TASE, 22, 3388--3399. (www)

  @article{ , abstract = {Microrollers hold great potential for biomedical applications for its mobility and environmental adaptivity. Autonomous trajectory tracking is crucial for microrollers to conduct medical tasks precisely and safely in complex unstructured physiological environments but remains challenging. In this article, we propose a novel direction-speed decoupled control strategy for autonomous trajectory tracking of microrollers. A modified kinematic model based on extended high-order states is proposed to capture the dynamic characteristics of microrollers, enhancing adaptability to unstructured environments. Decoupled direction and speed controllers adjusting the lateral and longitudinal motions of microrollers are designed based on separate error states to offer improved autonomy and performance in trajectory tracking. A real-time online parameter identification approach is formulated that skips tedious preliminary trials and could generally be used for other variations of microrollers and environments. The trajectory tracking performance of the proposed control strategy is evaluated experimentally on bioinspired magnetic microrollers (BMMs) in biological-like environments. The BMM successfully navigates through cell clusters and reaches the target destination, showing the potential of the proposed control strategy in real physiological environments. Note to Practitioners—This article is motivated by the growing interest in enhancing the autonomy of microrollers in biomedical applications. Controlling the microrollers along an intended trajectory is crucial for applications such as targeted drug delivery and minimally invasive surgery. Traditional control strategies focus solely on spatial path following, disregarding the motion speed of the microrollers. Moreover, kinematic model generally neglects the high-order states for simplification, which cannot facilitate microllers to track rapid changing trajectories in unstructured environments. The identification in real-time and amidst unknown disturbances is also crucial for microrollers to operate efficiently in complex physiological environments. However, existing control strategies rely heavily on preliminary experiments to obtain kinematic parameters for each form of microrollers and environments, which is impractical for real-world applications. To address these challenges, we propose a novel control strategy encompassing an extended state kinematic model, decoupled direction and speed controllers, and an online approach to updating kinematic parameters. Precise trajectory tracking containing direction and speed of BMM is achieved in a biological-like environment consisting of real cell clusters, showing improved performances compared to conventional control methods, and can be adapted to other forms of microrollers without preliminary trials.}, archive = {J_TASE}, author = {Wei Zhang and Bin He and Zhen Yin and Chang Xu and Yanmin Zhou and Yu Cheng}, doi = {10.1109/TASE.2024.3394238}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3388--3399}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous trajectory tracking of bioinspired soft microrollers through a novel direction-speed decoupled control strategy}, volume = {22}, year = {2025}, }
• (2025). Adaptive nonlinear power regulation control of floating wind turbines with platform motion reduction. TASE, 22, 3377--3387. (www)

  @article{ , abstract = {Blade pitch actuators for traditional onshore wind turbines primarily serve the purposes of regulating generated power to the rated value in high wind speed regions. Yet, when considering the floating offshore wind turbines (FOWTs) subject to wind and wave disturbances, it can be observed that improving power regulation often comes at the expense of exacerbating motion in the floating platform, leading to elevated platform loads. To address this issue, this paper proposes a novel robust nonlinear pitch controller specifically designed to achieve power production regulation while simultaneously mitigating platform pitch motion for FOWTs. Moreover, to tackle the challenge posed by unknown dynamics under wind-wave joint loads, the presented controller integrates a two-layer neural network (NN) for real-time learning of these unknown system dynamics. Meanwhile, a robust continuous term is introduced to alleviate the effects of residual reconstruction errors from the NN and external disturbances. Finally, the viability and efficacy of the proposed scheme are clearly demonstrated through comprehensive comparative studies with traditional pitch controllers conducted on the National Renewable Energy Laboratory (NREL) FAST platform. Note to Practitioners—This paper is motivated by the challenge of achieving the competing objectives between power regulation and load mitigation for floating offshore wind turbines (FOWTs) in high speed region. FOWTs have garnered significant interest in the field of renewable energy due to their advantages, including the ability to install high-powered wind turbines and reduced costs in deeper waters. However, there exist some limitations that existing FOWT controllers face including inefficiency in directly applying onshore WT control schemes, typically depend on precise turbine models and lack adaptability to potential uncertainties and errors encountered in practical situations. To tackle these limitations, this paper proposes a novel robust nonlinear controller for FOWTs considering the platform motion, which has the ability to simultaneously achieve power production regulation while platform stabilization. A two-layer neural network for real-time learning is developed to capture potential changes in system dynamics, while a robust continuous term is introduced to alleviate potential errors in the control scheme. This naturally provides adaptability and robustness against uncertainties while enhancing overall control performance. The effectiveness of the proposed control scheme is validated through simulation results.}, archive = {J_TASE}, author = {Yingjie Gong and Qinmin Yang and Wenchao Meng}, doi = {10.1109/TASE.2024.3393918}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3377--3387}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive nonlinear power regulation control of floating wind turbines with platform motion reduction}, volume = {22}, year = {2025}, }
• (2025). Beetle swarm with constrained lévy flight for image matching. TASE, 22, 3366--3376. (www)

  @article{ , abstract = {Image matching is an essential part of many processes in practical industrial applications. Using optimization algorithms to optimize the actual problems in industrial production can lead to more efficient use of resources. This paper presents a new algorithm called Beetle Swarm with Constrained Lévy Flight (BSL) algorithm for solving problems in industrial production where image matching cannot be done quickly and accurately. This algorithm is based on the beetle antennae search algorithm. It combines the swarm intelligence algorithm with the constrained Lévy flight and quickly finds the optimal solution through the long-horned beetle’s judgment of the left and right odour concentration, which reduces the blindness of Lévy flight and dramatically improves the convergence speed of the algorithm. In the performance test, compared with other commonly used meta-heuristic algorithms, BSL shows stable performance with low time cost of convergence and better fitness results, which means BSL avoids the effect of random direction caused by a single beetle. In addition, this algorithm significantly improves the speed and accuracy of image matching in the application of Printed Circuit Board (PCB) defect detection. In practical application tests, the BSL algorithm was faster than the BAS algorithm, with a reduction in the fitness value. The subsequent robustness experiments further prove that the BSL algorithm has better noise immunity and is more suitable for application in actual production than Cuckoo Search (CS) and beetle antennae search (BAS). Note to Practitioners—This paper was motivated by the problem of matching images in PCB defect detection. Existing meta-heuristic approaches cannot achieve quick and accurate image matching due to the low convergence speed. This paper suggests a method that combines the beetle antennae search algorithm with the constrained Lévy flight. In this paper, we improve the Lévy flight mechanism to update the next position of each beetle to eliminate invalid solutions. Then, new directions are generated in combination with the odour concentration around the aspen whiskers, which reduces the blindness of Lévy flight to a certain extent and, indeed, leads to the optimal solution. We also incorporate the Normalized cross-correlation (NCC) algorithm to improve the matching speed and noise immunity. The application in automated optical inspection (AOI) shows that the algorithm significantly improved the speed and accuracy when applied to production.}, archive = {J_TASE}, author = {Yubing Wu and Danfeng Sun and Mingjia Zhang and Jianyong Zhao and Huifeng Wu}, doi = {10.1109/TASE.2024.3393897}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3366--3376}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Beetle swarm with constrained lévy flight for image matching}, volume = {22}, year = {2025}, }
• (2025). A matheuristic approach for delivery planning and dynamic vehicle routing in logistics 4.0. TASE, 22, 3345--3365. (www)

  @article{ , abstract = {In distribution logistics, the planning of vehicles’ routes and vehicles’ loads are traditionally managed separately, despite these activities are correlated. This often leads to various re-designs to make the routes and load plans compatible and applicable in practice. Moreover, the planned routes, which are static by definition, cannot always cope with unexpected events. Traffic congestion, vehicle failures, adverse meteorological conditions, and further undesired events can make the planned routes inapplicable and require vehicles’ re-routing. This results in lower service levels, undesired delays, and higher costs for logistics companies. With the aim of overcoming the above limitations, this work proposes a novel approach based on a matheuristic algorithm that jointly solves the problem of delivery planning and dynamic vehicle routing to automate the delivery process in a logistics 4.0 perspective. The presented algorithm includes two different phases: the static phase, which is executed offline and in advance with respect to the delivery day, and the dynamic phase, which is executed in real-time to cope with unexpected events during the delivery. For the first phase, a matheuristic approach is defined to efficiently solve the combined vehicle routing and loading problems. Differently, for the second phase, a genetic algorithm is proposed to re-route vehicles in real-time, considering both the redefinition in real-time of the nominal trip and/or of the sequence of the customers to be visited. The algorithm is tested both on a literature benchmark and on a real dataset provided by an Italian logistics company. The obtained results show that, on the one hand, the proposed algorithm can automatically provide feasible solutions that minimise travel costs, total travelled distance, and empty space on the vehicles; on the other hand, it can ensure in real-time effective re-routing solutions in case of unexpected events occurring during delivery. Note to Practitioners—This work is motivated by the need for facilitating the operations of planning and routing deliveries in the external logistics sector. We propose an algorithm that automatically generates feasible routing and loading plans for a set of Transport Units (TUs) (i.e., the static phase), and then updates in real-time the nominal route in case of unexpected events (i.e., the dynamic phase). More specifically, the first phase of the algorithm takes as input the set of different clients, the list of products packed into bins (i.e., standard packing units) to be delivered to each client, and the set of transport units available for the deliveries, and provides as output the number and type of TUs to be used, the composition of the bins in each transport unit, and the corresponding route, while optimising the space occupation in each TU and the travel costs. The second phase, instead, takes as input the nominal routes computed in the first phase and, in case of unexpected events (e.g., accidents, slowdowns, etc.) affecting one or more routes, it re-routes the involved trucks guaranteeing the maximum efficiency in regards to travel cost, travel time, and quality of service. The adoption of this algorithm by logistic companies supports the automation of the delivery process and drastically improves the efficiency of logistic operations, with particular regard to the number of used TUs, costs, safety of goods, and customers’ satisfaction.}, archive = {J_TASE}, author = {Giulia Tresca and Hadrien Salem and Graziana Cavone and Hayfa Zgaya-Biau and Sarah Ben-Othman and Slim Hammadi and Mariagrazia Dotoli}, doi = {10.1109/TASE.2024.3393507}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3345--3365}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A matheuristic approach for delivery planning and dynamic vehicle routing in logistics 4.0}, volume = {22}, year = {2025}, }
• (2025). Optimized formation control of nonlinear systems with full-state constraints using adaptive fixed-time techniques. TASE, 22, 3331--3344. (www)

  @article{ , abstract = {This paper proposes an approach for fixed-time (FxT) adaptive optimized formation control of nonlinear multi-agent systems (MASs) with unknown nonlinear dynamics and full-state constraints. To address system uncertainty and state constraints while achieving optimality in FxT settings, the paper presents a novel adaptive estimation and analysis. The proposed approach first introduces a tan-type nonlinear mapping to handle state constraints, eliminating the feasibility condition of the conventional barrier Lyapunov function method. Next, the actual optimal controller is iteratively designed using the identifier-actor-critic structure and optimized backstepping method, with neural approximators used to learn system uncertainty. Finally, a monotonically decreasing function is constructed to prove that the designed actor-critic update laws have an upper bound, which is essential for stability analysis. The proposed scheme can ensure that the formation is realized at a fixed time while optimizing a given performance index and meeting the constraint requirements. The simulation results verify the effectiveness of the proposed approach. Note to Practitioners—There are inevitable system constraints and model uncertainties in actual physical systems, which may degrade the operational performance of the system. In addition, for practical requirements, designers are eager for multiple agents to achieve the required formation performance at a fast convergence speed. While meeting the requirements of system constraints, it is of practical and theoretical significance to improve the convergence speed and robustness of MASs formation and ensure optimal performance. For this reason, this paper focuses on proposing an effective adaptive FxT-optimized formation control scheme to enhance system performance and convergence speed, which combines nonlinear mapping to address full-state constraints. To achieve optimality under FxT settings, an optimal controller with learning laws is first designed using an identifier-actor-critic structure, in which the identifier is used to learn uncertainty. Then, by constructing a quadratic function, it is proved that the estimation error of the learning law is bounded, thereby forming a new adaptive estimation and analysis scheme.}, archive = {J_TASE}, author = {Ping Wang and Chengpu Yu and Maolong Lv}, doi = {10.1109/TASE.2024.3392936}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3331--3344}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimized formation control of nonlinear systems with full-state constraints using adaptive fixed-time techniques}, volume = {22}, year = {2025}, }
• (2025). Adaptive stabilization control for a class of non-strict feedback underactuated nonlinear systems by backstepping. TASE, 22, 3320--3330. (www)

  @article{ , abstract = {Inspired by backstepping method, this paper proposes a novel adaptive stabilizing control method for a class of uncertain underactuated nonlinear systems, which is named as underactuated adaptive backstepping. Firstly, based on a virtual controller and a partial differential equation (PDE), a residual system is constructed. Then, a novel coordinate transformation is introduced to analyse the adaptive stabilizing control problem of the residual system, which breaks through the restriction of strict feedback form and achieves the estimation of unknown parameters. The proposed method first applies backstepping to form a residual system, and then uses backstepping again to design an adaptive controller for the residual system. The stability of the proposed method is proven based on Lyapunov’s theorems. Finally, two actual underactuated examples with physical damping are provided to demonstrate the effectiveness of the proposed underactuated adaptive backstepping. Note to Practitioners—This paper aims to develop a novel underactuated adaptive scheme called underactuated adaptive backstepping for underactuated mechanical systems. This scheme can effectively address the parameter uncertainty problem and achieve good control effect, even if there is uncertain physical damping which may change the equilibrium point in the system. Compared with existing methods, this scheme provides a novel design approach without linearization and approximation, which fully promotes the development of backstepping in the underactuated field. Moreover, the simulation experiments are carried out on the inertia wheel pendulum and ball and beam system with physical damping, thus effectively proving the feasibility of this scheme. In the future, the scheme will be applied to practical mechanical systems.}, archive = {J_TASE}, author = {Ning Li and Xiaoping Liu and Cungen Liu and Weikai He and Huanqing Wang}, doi = {10.1109/TASE.2024.3392877}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3320--3330}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive stabilization control for a class of non-strict feedback underactuated nonlinear systems by backstepping}, volume = {22}, year = {2025}, }
• (2025). Cooperative tracking of quadrotor UAVs using parallel optimal learning control. TASE, 22, 3308--3319. (www)

  @article{ , abstract = {The cooperative tracking control issue of quadrotor unmanned aerial vehicles (UAVs) is investigated, where a cluster of UAVs is required to maintain a preassigned pattern while tracking a reference trajectory. Since only a part of UAVs can access the reference trajectory, an adaptive distributed estimator is developed such that each UAV could obtain the accurate estimate exponentially. Based on the hierarchical development, a parallel optimal learning control strategy is proposed by introducing virtual artificial systems that generate the practical control commands. In particular, a saturated force command is exploited for the position loop tracking to the estimated trajectory and a torque command is utilized to ensure the command attitude tracking, respectively. Moreover, a data-based learning law is designed for the critic weight under the finite excitation (FE), which is made available for the inadmissible initial control. It is shown that the overall closed-loop system is uniformly ultimately bounded and the tracking errors eventually converge to small sets around zero. Simulation and experiment results further verify the proposed control strategy. Note to Practitioners—This paper is inspired by the cooperative formation control issue of quadrotor UAVs. In practical missions like surveillance and reconnaissance, a UAV comes into or leaves out of the cluster may lead to re-tune the control parameters for all the involved UAVs. To save the onboard resource, the communication topology among the UAVs cannot be tedious. The proposed parallel learning strategy is implemented as follows. First, we design an adaptive distributed estimator for each UAV that could obtain the reference information accurately. Then, we propose a data-based parallel optimal learning strategy for each UAV that tracks the estimated reference trajectory. Since the proposed strategy does not require any global information regarding the communication topology, the parameters can be remained when the UAV cluster needs to extend or reduce, which makes it more practical. Finally, the proposed strategy is validated by both simulation and experiment results.}, archive = {J_TASE}, author = {Kewei Xia and Xinyi Li and Kaidan Li and Yao Zou and Zongyu Zuo}, doi = {10.1109/TASE.2024.3391942}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3308--3319}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Cooperative tracking of quadrotor UAVs using parallel optimal learning control}, volume = {22}, year = {2025}, }
• (2025). Sliding mode quantized control of interval type-2 fuzzy systems under event-triggered scheme: Stability analysis and performance optimization. TASE, 22, 3295--3307. (www)

  @article{ , abstract = {Focusing on the sliding mode control of interval type-2 fuzzy systems with restricted communication capability, this study designs an event-triggered mechanism and a dynamic quantizer. Firstly, a membership-function-based dynamic event-triggered scheme is elaborately designed to reduce the transmission of the system state. As a result, only the triggered state is available for the fuzzy sliding mode controller and its membership functions (MFs). Since the event-triggered fuzzy controller contains different MFs from the system and the controller’s MFs are only updated at triggering instants, the above two factors result in the mismatching phenomenon between the system’s and the controller’s MFs. Under the event-triggered scheme, the MFs’ local information of both controller and system is extracted to derive sufficient conditions for ensuring the control performance of the closed-loop dynamic system and the sliding mode dynamics. Furthermore, a dynamic quantizer’s parameter is adjusted online to improve the quantization procedure of the control signal, and an optimization problem is introduced to balance the system’s dynamic performance with communication resource utilization by reducing the convergence domain and increasing the triggering threshold. Finally, the practicability of the proposed control scheme is illustrated via a numerical example and the single-link manipulator. Note to Practitioners—In many practical engineering applications, the equipment units in the control system are generally implemented on the network platform, and measurement and control signals are transmitted through wired/wireless network channels. The usage of communication resources is important to prevent packet loss and communication congestion. The event-triggered scheme and quantization are two effective ways to reduce the communication burden. In this work, an event-triggered control scheme and a quantization transmission mechanism are proposed for a networked control system under IT2 fuzzy modeling. In order to improve the limited quantizer’s levels due to the hardware cost, a dynamic variable is introduced, and the adjustment scheme is proposed to ensure that the dynamic quantizer can be implemented on the practical digital platform. The co-design scheme of event-triggered, quantization, and controller is proposed for ensuring the dynamic performance of networked control systems.}, archive = {J_TASE}, author = {Yekai Yang and Jiarui Li and Yugang Niu}, doi = {10.1109/TASE.2024.3391910}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3295--3307}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Sliding mode quantized control of interval type-2 fuzzy systems under event-triggered scheme: Stability analysis and performance optimization}, volume = {22}, year = {2025}, }
• (2025). Robustness optimization of air transportation network with total route cost constraint. TASE, 22, 3280--3294. (www)

  @article{ , abstract = {Designing a robust air transportation network is critical for aviation activities to maintain highly efficient operations. Given certain total route cost, designing a strategy to improve the robustness of the network is a challenging issue. In this paper, first, a comparison experiment based on a small example shows the total effective resistance is a superior measure satisfying two critical criteria that other measures do not satisfy. Consequently, we explicitly formulate the network optimization problem as the minimization of total effective resistance under several constraints. The problem is often too time-consuming to solve to the optimal by exact methods. To achieve better efficiency, we propose a convex relaxation method for medium-scale networks making use of the problem properties. For large-scale networks, a clustering based convex relaxation method, reducing the dimensions of the network via selecting critical airports, is further proposed. To demonstrate the efficiency and efficacy of the proposed methods, simulations are performed on nine typical scale-free network and two real air transportation networks including Jetstar Asia Airway and domestic American Airlines. Note to Practitioners—The crucial need for designing robust air transportation networks in the aviation industry serves as the impetus for this work. The application scenarios range from the design of airport and flight operation networks for airlines, to air traffic networks with airways and waypoints, extending to drone cargo networks. An essential characteristic of a well-designed network is its ability to preserve connectivity, as much as possible, when it experiences disruptions such as flight cancellation or airway navigation equipment failure. The paper presents an argument for using total effective resistance as the key measure of network robustness, showing it to outperform other measures. We’ve developed specialized algorithms that leverage the unique properties of this optimization problem to minimize the total effective resistance of air transportation networks subject to certain total route cost to enhance the network structure. Although this study focuses on air transportation networks, the applicability of the algorithms extends beyond this field. Practitioners can apply them in other scenarios to design robust networks in various aviation operations.}, archive = {J_TASE}, author = {Changpeng Yang and Jianfeng Mao and Xiongwen Qian and Edmond Q. Wu}, doi = {10.1109/TASE.2024.3391763}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3280--3294}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robustness optimization of air transportation network with total route cost constraint}, volume = {22}, year = {2025}, }
• (2025). Temporal finite-time adaptation in controlling quantized nonlinear systems amidst time-varying output constraints. TASE, 22, 3265--3279. (www)

  @article{ , abstract = {Using the backstepping technique, this paper formulates innovative adaptive finite-time stabilizing controllers for uncertain nonlinear systems featuring nonuniform input quantization and asymmetric, time-varying output constraints. These novel controllers leverage the consistent characteristics of both hysteresis quantizers and logarithmic quantizers. Quantization errors, when consistent, become unbounded and contingent on control input, rendering them incompatible with the growth conditions of nonlinear systems. Consequently, the developed adaptive controllers eliminate the reliance on growth conditions, effectively addressing the impact of unbounded quantization errors on finite-time stability. This adaptability allows the controllers to function effectively with systems employing either hysteresis quantizers or logarithmic quantizers. The paper establishes the convergence of these controllers through the finite-time Lyapunov stability theorem. It also provides a comprehensive guideline for tuning settling time, enabling fine-grained control over finite-time convergence and adjustable tracking error performance. Additionally, the controllers rigorously maintain system output within predefined limits. Their effectiveness and low computational burden are demonstrated through three comparative numerical simulations and a practical simulation in collision-free trajectory tracking control of an autonomous vehicle platoon using the vehicle motion software CarSim. These simulations confirm the advanced performance of the adaptive controllers. Note to Practitioners—This paper introduces an innovative approach to control uncertain nonlinear systems encountering intricate input quantization and output constraints. Employing the sophisticated backstepping technique, the authors present adaptive finite-time-stabilizing controllers engineered to address nonuniform input quantization and asymmetric, time-varying output restrictions. What distinguishes these controllers is their reliance on the consistent behavior exhibited by hysteresis and logarithmic quantizers. This unique feature equips them to effectively counteract unbounded quantization errors influenced by control input. Most notably, these controllers eliminate the conventional growth conditions typically demanded by nonlinear systems. As a result, they extend their applicability to a broad spectrum of systems employing either hysteresis or logarithmic quantizers. The research also provides practitioners with a valuable guideline for precisely adjusting settling time. This enables the attainment of desired convergence rates while permitting adaptable tracking error performance. Additionally, these controllers guarantee that the system’s output adheres to predefined limits. The practical significance of this study is highlighted through three comparative numerical simulations and a real-world application simulation. This real-world simulation involves collision-free trajectory tracking control of an autonomous vehicle platoon, executed using the vehicle motion software CarSim. These simulations unequivocally demonstrate the effectiveness and low computational burden of the developed controllers, thereby establishing them as a valuable resource for practitioners facing complex control challenges in various domains.}, archive = {J_TASE}, author = {Shaohua Cui and Yongjie Xue and Maolong Lv and Kun Gao and Bin Yu and Jinde Cao}, doi = {10.1109/TASE.2024.3391629}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3265--3279}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Temporal finite-time adaptation in controlling quantized nonlinear systems amidst time-varying output constraints}, volume = {22}, year = {2025}, }
• (2025). Opacity enforcement in discrete event systems using modification functions. TASE, 22, 3252--3264. (www)

  @article{ , abstract = {Opacity of a discrete event system is a confidentiality property that characterizes instances when the secret behavior of the system cannot be revealed. This paper considers current-state opacity with respect to a set of secret states, and addresses the opacity enforcement problem via modification functions that are capable of replacing in real-time the output of the underlying system with another output, aiming to prevent the intruder from inferring the secret. We study modification functions that are constrained, i.e., they may not be able to modify certain outputs or they may have restrictions in how they modify a particular output. A system is said to be emc-enforceable if opacity can be enforced via modification functions under the given event modification constraints (emc). To verify whether a system is emc-enforceable, we construct a verifier, based on which a necessary and sufficient condition is obtained. If a system is emc-enforceable, an algorithm is developed to compute a modification strategy based on the verifier to enforce opacity. For m-enforceability, a special case of emc-enforceability without constraints, two necessary and sufficient conditions to verify and enforce opacity with reduced complexity are presented, one based on a system estimator and the other based on a detector. Note to Practitioners—Discrete event systems provide appropriate mathematical models for analysing information-flow properties of cyber-physical systems, such as intelligent transportation systems, command-control systems, and networked communication systems, in dynamically changing environments. Opacity is a general property capable of capturing safe information-flow in a cyber-physical system. This article targets the enforcement of opacity using modification functions that mask the output of the system into another output in a manner that appears consistent to an external observer as the system evolves. The designed modification function can be realized via triggers equipped with digital logic and storage functionality. As an output is generated by a system, a trigger displays a fake output based on the stored real and fake states.}, archive = {J_TASE}, author = {Xiaoyan Li and Christoforos N. Hadjicostis and Zhiwu Li}, doi = {10.1109/TASE.2024.3391617}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3252--3264}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Opacity enforcement in discrete event systems using modification functions}, volume = {22}, year = {2025}, }
• (2025). Human-simulated intelligent walking control for biped robots. TASE, 22, 3237--3251. (www)

  @article{ , abstract = {Biped robots have received increasing attention due to their human-like mechanical structure and good environmental adaptability. In this paper, a new Human-Simulated Intelligent Walking Control (HIWC) scheme is proposed to solve the stability problem of the most popular proportional differential (PD) control under model inaccuracy and disturbance, and further improve its control performance. Specifically, based on Human-Simulated Intelligent Control (HSIC), HIWC is a hierarchical control structure composed of a foot placement compensation (FPC) strategy at the high-level planning layer, and a multi-mode compensation controller (MCC) at the low-level (execution) layer. In FPC, a foot placement compensation algorithm is proposed to plan and correct the swing foots trajectory in real time. MCC consists of a PD and two adaptive compensation algorithms. MCC under bounded uncertainty is proven to be stable in this paper using the Lyapunov theorem. HIWC was tested and compared with PD and model predictive control (MPC) in three experiments on a physical robot platform for planar walking, push-pull, and uneven-ground walking. Experimental results show that the proposed HIWC is more flexible and accurate in controlling the robot’s movement. Note to Practitioners—This paper builds on the fact that PD controllers cannot be easily proven to be stable and do not provide accurate control for the biped robot walking problem. To address these issues, this paper proposes a novel control scheme namely Human-Simulated Intelligent Walking Control (HIWC) and belonging to the family of Human-Simulated Intelligent Control (HSIC) schemes. The proposed HIWC system has been compared against the model predictive control (MPC) and a proportional differential (PD) controller. The proposed HIWC, unlike PD controllers, is rigorously proven to be stable in the presence of model inaccuracy and disturbance. Furthermore, the experiments carried out on a real-world biped robot demonstrate the superiority of HIWC over PD and MPC in terms of control performance.}, archive = {J_TASE}, author = {Xingyang Liu and Haina Rong and Ferrante Neri and Kuize Zhang and Qiang Yang and Zhangguo Yu and Gexiang Zhang}, doi = {10.1109/TASE.2024.3391358}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3237--3251}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human-simulated intelligent walking control for biped robots}, volume = {22}, year = {2025}, }
• (2025). Adaptive preassigned finite-time tracking control for PDE-ODE systems with dead-zone input and actuator failures. TASE, 22, 3226--3236. (www)

  @article{ , abstract = {This paper presents a finite-time tracking control for PDE-ODE coupled systems with dead-zone input and actuator failures. To begin with, a novel finite-time performance function and error transformation are designed for a reaction-diffusion equation to improve the transient and steady-state performance for PDE-ODE systems. In addition, to overcome the difficulties generated by the uncertain actuator failures and dead-zone input for coupled systems, a finite-time tracking controller and the adaptive laws are designed to compensate for the drastic effect of unknown nonlinearities. By applying neural networks approximation technology, the computational complexity and difficulty of controller design are significantly reduced. Unlike the existing stability results in the PDE-ODE systems, the developed method can guarantee the given transient performance by adaptive parameter tuning and the tracking error converges to the specified region in finite time. Moreover, the boundedness and convergence of all the signals in the closed-loop coupled systems are proved. Finally, the simulation example exhibits the effectiveness of the proposed control scheme. Note to Practitioners—The PDE-ODE coupled systems in modern mechanical engineering scenarios require better transient and steady-state performance as well as more agility of adjustment method. Because of the special infinite-dimensional characteristics of coupled systems compared to single PDE or ODE systems, it is more complex to design the adaptive finite-time controller for PDE-ODE systems. The finite-time performance function and error transformation designed in this paper aim to improve the stability of the coupled systems and adjust the preassigned performance more flexible. The settling time is independent and free of the parameters and initial values of the coupled systems which can be chosen by users according to actual requirements. Meanwhile, the constraints of actuator failures and dead-zone nonlinearity are compensated simultaneously for PDE-ODE systems via the presented adaptive neural networks finite-time control scheme.}, archive = {J_TASE}, author = {Lei Wan and Jiayue Sun and Yang Liu and Xiangpeng Xie}, doi = {10.1109/TASE.2024.3391312}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3226--3236}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive preassigned finite-time tracking control for PDE-ODE systems with dead-zone input and actuator failures}, volume = {22}, year = {2025}, }
• (2025). Efficient learning for selecting top-m context-dependent designs. TASE, 22, 3210--3225. (www)

  @article{ , abstract = {We consider a simulation optimization problem for context-dependent decision-making, which aims to determine the top-m designs for all contexts. Under a Bayesian framework, we formulate the optimal dynamic sampling decision as a stochastic dynamic programming problem and develop a sequential sampling policy to efficiently learn the performance of each design under each context. The asymptotically optimal sampling ratios are derived to attain the optimal large deviations rate of the worst-case probability of false selection. The proposed sampling policy is proved to be consistent, and its asymptotic sampling ratios are shown to be asymptotically optimal. Numerical experiments demonstrate that the proposed method improves the efficiency for selecting top-m context-dependent designs. Note to Practitioners—The performance of a given design may vary across different contexts, and better decision-making is possible by refining the available contextual information. We consider a context-dependent ranking and selection problem, which allows optimal selection to depend on contextual information obtained prior to decision-making. We develop a dynamic sampling scheme to efficiently learn and select the top-m designs in all contexts. Numerical experiments, including a honeypot deception game problem and a medical resource allocation problem, demonstrate that the proposed sampling scheme significantly improves the efficiency of context-dependent ranking and selection for the top-m designs.}, archive = {J_TASE}, author = {Gongbo Zhang and Sihua Chen and Kuihua Huang and Yijie Peng}, doi = {10.1109/TASE.2024.3391020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3210--3225}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Efficient learning for selecting top-m context-dependent designs}, volume = {22}, year = {2025}, }
• (2025). Five-axis contour error estimation based on multi-information dynamic time warping. TASE, 22, 3196--3209. (www)

  @article{ , abstract = {Contour accuracy is crucial for machining precision in five-axis computer numerical control (CNC) machining. This paper addresses the challenge of improving contour accuracy by proposing a novel contour error estimation and compensation method based on dynamic time warping (DTW). By incorporating time information and geometric characteristics of the machining path, the proposed method introduces a multiple-information fusion algorithm to define distance characteristics between the planned and actual trajectory sequences. This allows the calculation of a distortion path and the establishment of a mapping model between the two sequences. To mitigate the effect of DTW singularity on contour error estimation, a mapping model is established between line segments to determine reference points. The position contour error and the direction contour error of the five-axis tool are accurately estimated using segmented Hermite interpolation, and a spatial iterative learning framework is employed to compensate for them. Experimental results demonstrate the effectiveness of the proposed method in dealing with estimation errors in complex trajectories and its good performance in improving contour accuracy. Note to Practitioners—This paper proposes an effective strategy for the estimation of contour errors in five-axis machining. Currently, most methods for contour error estimation in five-axis machining are based on the nearest point principle. However, this approach fails to accurately estimate contour errors for complex trajectories with significant curvature variations, leading to ineffective contour error compensation in subsequent stages. Therefore, we introduce a contour error estimation algorithm based on DTW. This algorithm takes into account the time information and geometric features of the machining path. Experimental results validate the feasibility and advantages of this approach.}, archive = {J_TASE}, author = {Bo Peng and Zhiyu Hu and Jiangang Li and Yanan Li}, doi = {10.1109/TASE.2024.3390835}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3196--3209}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Five-axis contour error estimation based on multi-information dynamic time warping}, volume = {22}, year = {2025}, }
• (2025). Multimodal unified control method using the lie derivative and lyapunov theory for enhancing the dynamic stability of power systems with 100% renewable energy. TASE, 22, 3182--3195. (www)

  @article{ , abstract = {The multimodal dynamic stability (MDS) is a critical issue for constructing power systems with 100% renewable energy (PSRE). This paper proposes a multimodal unified control (MUC) method for enhancing the MDS in PSRE. First, an improved Heffron-Phillips model is established to demonstrate the mechanism of multimodal dynamic instability (MDI). Next, a third-order external subsystem model of the grid-forming renewable energy source (GFM-RES) is derived by the Lie derivative. Then, the MUC architecture is designed by utilizing the Lyapunov theory in the third-order external system. Finally, the application of the proposed MUC method is verified by analyzing the pertinent results for the modified IEEE 11-bus system with a 100% renewable energy generation. Note to Practitioners—The construction of PSRE has attracted widespread attention from the academic and engineering communities. Dynamic stability is one of the key issues faced in building PSRE. This work presents a MUC of GFM-RESs for improving the MDS of the PSRE. Practitioners should be able to apply the MUC to GFM-RESs like battery energy storage system, wind turbines and photovoltaic units. The MUC is designed using the Lie derivative and Lyapunov theory. From a practical point of view, the MUC achieves the MDS from a control perspective, without additional investment required.}, archive = {J_TASE}, author = {Guoteng Wang and Ying Huang and Chongyu Wang and Mohammad Shahidehpour}, doi = {10.1109/TASE.2024.3390754}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3182--3195}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multimodal unified control method using the lie derivative and lyapunov theory for enhancing the dynamic stability of power systems with 100% renewable energy}, volume = {22}, year = {2025}, }
• (2025). Finite-time control for high-order random nonlinear systems with unknown control coefficients. TASE, 22, 3172--3181. (www)

  @article{ , abstract = {Random differential equations (RDEs) involving colored noise hold more practical significance than stochastic differential equations (SDEs) driven by white noise. The research on random nonlinear systems is becoming more and more popular. This paper considers the problem of finite-time control for high-order random nonlinear systems with general growth conditions and unknown control coefficients. Based on the backstepping and domination idea, the technique of adding a power integrator is adopted to recursively design an adaptive controller that renders the closed-loop system finite-time stable in probability. This paper serves as a pioneering attempt at finite-time control design of random nonlinear systems. The theoretical results are corroborated by a numerical example.Note to Practitioners—This article was motivated to realize the finite-time stability of high-order random nonlinear systems, even though there are unknown control coefficients in the system. The practical significance of this paper is mainly twofold. Firstly, since random noise is ubiquitous in all kinds of engineering environments, the research on random nonlinear systems driven by colored noise is of great importance. Secondly, the asymptotic stability is infeasible in the practical implementation, so the achievement of finite-time control makes the method proposed in this paper applicable.}, archive = {J_TASE}, author = {Ruipeng Xi and Huaguang Zhang and Yunfei Mu and Yingchun Wang}, doi = {10.1109/TASE.2024.3390155}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3172--3181}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-time control for high-order random nonlinear systems with unknown control coefficients}, volume = {22}, year = {2025}, }
• (2025). Design and development of a new biped robotic system for exoskeleton-structure window cleaning. TASE, 22, 3160--3171. (www)

  @article{ , abstract = {Cleaning high-altitude building windows is traditionally a high-risk task for human workers due to the danger of falls. To enhance safety in facade cleaning and similar tasks, the development of window-cleaning robots has become increasingly essential. However, current robots face challenges in navigating exoskeleton structures due to obstructive beams, limiting their application. This article proposes a novel biped window-cleaning robotic system (BWCRS) designed for exoskeleton structures. The system comprises two key components: a multifunctional biped climbing robot (BCR) and a specialized cleaning robot. The BCR is adept at overcoming obstacles inherent in exoskeleton structures and transitioning across various inclined window surfaces. The BCR adopts a symmetrical construction, and the ends of the feet are equipped with vacuum suction cups and electromagnets, which can switch between different actuators at the geometric base and end of the system. The system is further enhanced by air pressure sensors that monitor the suction cup adhesion in real time, ensuring stable operation. A crucial functionality of the BCR is its manipulation mode, guided by a visual camera. This mode enables precise location and manipulation of the cleaning robot, ensuring effective coverage of all window areas. The mechanical design and mathematical modeling of the BWCRS are thoroughly examined, demonstrating its robust window-cleaning capabilities. The experimental results reveal that the proposed BCR has a favorable window-climbing and manipulation ability and that the BWCRS has a favorable cleaning effect on the exoskeleton-structure windows. Note to Practitioners—This paper is motivated by the problem of window-cleaning for exoskeleton structures, but it also applies to solving window maintenance and inspection. Previous window-cleaning robots were mainly aimed at the facades or surfaces with low obstacles that cannot be applied directly to exoskeleton-structured windows. This paper proposes an exoskeleton-structure window-cleaning method based on a biped window-climbing robot and a cleaning robot. The biped robot can climb over obstacles on the exoskeleton structures and manipulate the cleaning robot to different cleaning areas. Here, we introduce the biped window-climbing robot from the perspective of mechanical design and mathematical kinematics. We then demonstrate multiple obstacle-crossing motion modes of the biped window-climbing robot. Combined with visual-based manipulation, the window-cleaning process for the exoskeleton structures by a biped window-climbing robot integrated with a cleaning robot is demonstrated. The results verify that the proposed design can complete window-cleaning tasks on exoskeleton structures.}, archive = {J_TASE}, author = {Weijian Zhang and Zikang Li and Xianli Wang and Yonghao Huang and Junan Li and Lap Mou Tam and Qingsong Xu}, doi = {10.1109/TASE.2024.3390030}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3160--3171}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and development of a new biped robotic system for exoskeleton-structure window cleaning}, volume = {22}, year = {2025}, }
• (2025). Fixed-time neuroadaptive backstepping tracking control for uncertain nonlinear systems with predictor based learning. TASE, 22, 3147--3159. (www)

  @article{ , abstract = {This work focuses on the issue of fixed-time tracking control for a class of nonlinear systems affected by unknown uncertainties and external disturbances. First, a fixed-time neuroadaptive approximator is proposed to estimate the lumped uncertainties in nonlinear systems. Unlike existing neural network based methods, the estimation solution presented here introduces an adaptive predictor based learning mechanism, which would improve the estimation performance by removing the effect of tracking errors on the estimation process. Then, based on the reconstructed information, a fixed-time command filtered backstepping controller is developed with a fixed-time compensation system. In the compensation system, a novel bounded function is skillfully utilized such that the order and complexity of the compensation system are effectively reduced. Moreover, it is demonstrated that the designed control scheme can drive the tracking error to a small set near zero in a fixed time. Finally, the validity of the proposed control scheme is illustrated by numerical simulations. Note to Practitioners—This paper is motivated by the tracking control problem for nonlinear systems such as robotic, spacecraft, and unmanned aerial vehicle system. Existing tracking control schemes often suffer from issues such as insufficient tracking speed, redundant design process and the explosion of complexity. A fixed time neuroadaptive backstepping control scheme is proposed in this paper to ensure convergence of the error within a fixed time. An adaptive predictor-based fixed-time neuroadaptive estimator is presented to enhance the speed and accuracy of uncertainty estimation. Furthermore, a novel fixed-time compensation system is presented, which effectively addresses the issue of complexity explosion while reducing coupling of the compensation system, making the controller more concise. The effectiveness of the proposed method is validated through a numerical simulation in an uncertain spacecraft pitch motion system. Future work will involve validation of the proposed method on a hardware-in-the-loop simulation system or experiment platform.}, archive = {J_TASE}, author = {Han Gao and Jiale Wang and Yuanqing Xia and Jinhui Zhang and Bing Cui}, doi = {10.1109/TASE.2024.3390007}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3147--3159}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time neuroadaptive backstepping tracking control for uncertain nonlinear systems with predictor based learning}, volume = {22}, year = {2025}, }
• (2025). Contract-based demand response mechanism for commercial and industrial customers. TASE, 22, 3135--3146. (www)

  @article{ , abstract = {Designing an optimal Demand Response (DR) program necessitates having information on the utility of users. However, that is not easily accessible since the users may not be willing to report their personal information. In this paper, we propose an incentive-based DR mechanism for Commercial and Industrial ( $\text{C}\&\text{I}$ ) customers, leveraging contract theory to address the challenges of incomplete information and ensure the participation of customers in the DR program. Designing a DR mechanism includes specifying an incentive reward, penalty, and the corresponding demand reduction as functions of the customer’s utility information. These functions are obtained through an optimization problem that maximizes the grid operator’s profit, ensures truthful reporting of personal information and participation of the customers in the DR program. After formulating the optimal mechanism, the main technical challenge is reformulating the obtained non-convex and computationally inefficient optimization problem to a tractable and convex one. The extensive numerical evaluations demonstrate the effectiveness of our proposed framework in comparison to existing benchmark schemes. Note to Practitioners—Imbalances between supply and demand, particularly during the hot summer months, can result in issues such as power outages. This problem is detrimental for both grid operators and customers alike. DR programs have been implemented in numerous countries to address these challenges. $\text{C}\&\text{I}$ customers, due to the schedulable nature of their loads, make for ideal targets for these programs. However, the main obstacle to implementing such programs is customer participation. Incentive rewards offered in these programs for load reduction are fixed for all customers, discouraging those with low rewards from participating and imposing high costs on grid operators for those with high rewards. On the other side, designing unique rewards and penalties for each customer necessitates information about their preferences, which grid operators cannot access. In this paper, we propose an incentive mechanism based on the contract between the grid operator and $\text{C}\&\text{I}$ customers that ensures customers truthfully report their preferences while simultaneously maximizing grid operator profit and guaranteeing the participation of all customers. The proposed DR program can be implemented for large non-residential customers in order to reduce the peak load of the grid.}, archive = {J_TASE}, author = {Sajad Parvizi and Mina Montazeri and Hamed Kebriaei}, doi = {10.1109/TASE.2024.3389999}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3135--3146}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Contract-based demand response mechanism for commercial and industrial customers}, volume = {22}, year = {2025}, }
• (2025). Normalized P-D and intermittent hybrid h∞ control for delayed descriptor systems via impulsive-inputs-dependent conditions. TASE, 22, 3125--3134. (www)

  @article{ , abstract = {The problem of normalized proportional-derivative (P-D) and intermittent hybrid $H_{\infty }$ control for delayed descriptor impulsive systems (DDISs) is studied in this work. The P-D and intermittent hybrid state feedback controller is designed, such that normalisation and $H_{\infty }$ stabilisation of DDISs can be realised. A novel impulsive-inputs-dependent Lyapunov-Krasovskii functional based on a couple of auxiliary functions is constructed, which is continuous along the system state trajectories without imposing additional conditions on impulsive dynamics. Improved $H_{\infty }$ stability criterion is proposed in the form of linear matrix inequalities by utilizing free weighting matrix technique, and the $H_{\infty }$ stability criterion is described as a single condition based on a combination of discrete and continuous dynamics. Finally, the usefulness of the method set forth in this work is demonstrated by a two-loop circuit network. Note to Practitioners—This work is motivated by realizing normalized P-D and intermittent hybrid $H_{\infty }$ control for delayed descriptor impulsive systems. In practical engineering, many dynamic systems have to be formulated as descriptor systems when modeling, such as circuit systems, vehicle suspension systems, and so on. It is worth pointing out that infinite dynamic modes will occur due to the algebraic constraints in descriptor systems, and can lead to internal impulsive behavior. At the same time, these systems are often accompanied by external impulsive perturbations, such as the sudden failure of a component in circuit systems, etc. The internal and external impulses can cause serious damage to the system performances and stability. Most of the existing control schemes are not well suited for such practical systems. In addition, due to the presence of time-varying delays, the delayed dynamic systems can easily collapse and cause significant damages. In this work, the P-D and intermittent hybrid $H_{\infty }$ controller is designed so as to deal with the above challenges.}, archive = {J_TASE}, author = {Guangming Zhuang and Yiqun Liu and Jianwei Xia and Xiangpeng Xie}, doi = {10.1109/TASE.2024.3389983}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3125--3134}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Normalized P-D and intermittent hybrid h∞ control for delayed descriptor systems via impulsive-inputs-dependent conditions}, volume = {22}, year = {2025}, }
• (2025). Parameters estimation and optimal scheduling of a retrofit-free refrigerator. TASE, 22, 3113--3124. (www)

  @article{ , abstract = {Refrigerators, as a type of thermostatically controlled load (TCL), have good capacities to participate in demand response (DR) programs. However, existing methods have paid little attention to refrigerator parameter estimation and have rarely focused on retrofit-free control strategies. These control strategies often take the refrigerator’s temperature set-point or the compressor ON/OFF status as decision variables, requiring retrofitting the thermostatic controller. To address this issue, this paper proposes a 2R1C model for the refrigerator, which facilitates parameters estimation and scheduling strategy design. Based on the 2R1C model, a least-square-estimation (LSE)-based parameters estimation method is proposed to improve the accuracy and efficiency. In addition, a price-based scheduling method is proposed for an actual retrofit-free refrigerator that uses the main-power ON/OFF status as the decision variable, which can be easily implemented through an external power socket. By establishing a temperature-ramp-oriented thermostat control model considering the main-power ON/OFF status, the scheduling model is developed and linearized into a mixed integer linear program (MILP). The optimization, simulation, and practical control of an actual refrigerator verify the effectiveness of the proposed method. Note to Practitioners—This paper addresses the optimal scheduling problem of the refrigerator participating in DR, applicable to the majority of practical refrigerators. Existing optimal scheduling strategies for refrigerators suffer from two main issues. Firstly, most of these strategies assume that the thermal parameters of refrigerator loads are given, leading to inaccuracies due to the neglect of parameter estimation. Secondly, the majority of control strategies define the decision variable as the refrigerator’s power consumption rather than the temperature set-point, which deviates from the actual mode in which refrigerators indirectly control power through temperature set-points. This paper proposes a modeling approach based on 2R1C for the mode in which practical refrigerator control power indirectly through temperature set-points. On this basis, a parameter estimation strategy is designed, and an optimization strategy based on MILP is developed with the main-power ON/OFF status of the refrigerator as the decision variable. Compared to traditional methods, the proposed approach can be implemented through an external power socket, making it suitable for the vast majority of refrigerators without the need for retrofitting. In future research, we aim to extend this single refrigerator scheduling strategy to the control of a large number of refrigerators and even clusters of other TCLs, achieving more complex application scenarios.}, archive = {J_TASE}, author = {Yu-Qing Bao and Qing-He Sun and Wenhao Jia and Tao Ding}, doi = {10.1109/TASE.2024.3389711}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3113--3124}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Parameters estimation and optimal scheduling of a retrofit-free refrigerator}, volume = {22}, year = {2025}, }
• (2025). Data-driven identification model of vulnerable set for cascading failure in power grid. TASE, 22, 3097--3112. (www)

  @article{ , abstract = {The frequent blackouts around the world in the past 20 years have brought the security of the power grid to a head. Among various hazard situations, cascading failure is the one with critical threats due to its widespread propagation and long duration. An effective way to prevent cascading failure is to identify the vulnerable set, which is defined as the composition of transmission line combinations that can initialize the sequence of failures. In this paper, we first elaborate cascading failure model and its adaptation under the scenario of interest. Then a novel framework of data-driven identification model is developed to replace the traditional flow-based detection process, which is computationally heavy. Specifically, a method that seamlessly achieves the globally topological features embedding by designing a tailored messaging mechanism adjusted for the power grid is proposed, overcoming the otherwise problem of the constrained neighborhood in existing graph convolution networks (GCNs). Besides, with the proposed pruning optimization method, the sparsity of the vulnerable set can be naturally enforced and combinatorial explosion is readily alleviated. Numerical experiments are conducted on 30-bus, 200-bus, and 500-bus systems, including static and dynamic load scenarios. All of them verify the excellent performance of the identification model for both effectiveness and efficiency. Note to Practitioners—This paper is motivated by a practical need for mitigate the security threat of cascading failures to power grid through the vulnerable set identification. Existing methods for identifying the vulnerable set have limitations, including a limited number of vulnerabilities identified and challenges related to the high time complexity of cascading failure simulators and combinatorial explosion. To address this issue, we develop a data-driven identification model that integrates both the physical and topological features of the power grid. This model reduces the reliance on cascading failure simulators and enhances the efficiency of vulnerable set identification. Additionally, the pruning optimization method is proposed to further mitigate the high time complexity caused by combinatorial explosion. Simulative studies conduct in both dynamic and static load scenarios validate the performance of the developed data-driven model, demonstrating its ability to achieve rapid identification of the vulnerable set and thereby improve the overall robustness of the power grid against cascading failures.}, archive = {J_TASE}, author = {Sizhe He and Yuxun Zhou and Yadong Zhou and Jiang Wu and Ting Liu and Xiaohong Guan}, doi = {10.1109/TASE.2024.3389455}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3097--3112}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven identification model of vulnerable set for cascading failure in power grid}, volume = {22}, year = {2025}, }
• (2025). ADP-based prescribed-time control for nonlinear time-varying delay systems with uncertain parameters. TASE, 22, 3086--3096. (www)

  @article{ , abstract = {In this paper, we investigate the problem of prescribed-time optimal control using reinforcement learning technology. Unlike finite/fixed-time control methods that only achieve stability within specified time bounds, we propose a prescribed-time adaptive dynamic programming (ADP) control approach that ensures both optimality and prescribed-time stability. To address the challenge of solving the nonlinear Hamilton-Jacobi-Bellman (HJB) equation in finite horizon, we construct an actor-critic neural network (NN) with a time-varying activation function. The novel weight update laws are derived from the system’s terminal error and the approximate error of the HJB equation. This derivation eliminates the need for knowledge of dynamic conditions while ensuring compliance with terminal constraints. Based on the proposed prescribed time stability criterion, the control scheme is proven to satisfy prescribed time stability while also ensuring optimal system performance index and bounded weights. We apply the designed control scheme in a time-varying delay system and simulation examples validate the efficacy of the strategy. Note to Practitioners—Many industrial processes are nonlinear time-varying delay systems, which brings great challenges to solving the HJB equation of the prescribed-time control problem. Therefore, it is of great value to solve the prescribed-time control problem by using the advantages of finite time ADP in solving the time-varying HJB equation. The ADP-based prescribed-time optimal control (ADPPTC) scheme designed in this paper aims to optimize the steady-state performance of nonlinear time-varying delay systems, and because the user can prescribe the stability time, the system can accurately complete a task within a specific time range. At the same time, the prior demand for the system state can be eliminated by the proposed actor-critic neural network.}, archive = {J_TASE}, author = {Zhixuan Zhang and Kun Zhang and Xiangpeng Xie and Vladimir Stojanovic}, doi = {10.1109/TASE.2024.3389020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3086--3096}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ADP-based prescribed-time control for nonlinear time-varying delay systems with uncertain parameters}, volume = {22}, year = {2025}, }
• (2025). Optimization of task allocation for resource-constrained swarm robots. TASE, 22, 3068--3085. (www)

  @article{ , abstract = {While task allocation of swarm robots has been extensively researched, resource constraints of robots are rarely considered. In this work, we propose two novel task allocation methods robust to robot failures while considering the resource constraint, limited communication range, and deadline constraint of tasks. The first method, STA (static task allocation) method, finds an optimal task allocation solution at compile-time in terms of the minimum expected finish time, using answer set programming. On the other hand, the DTA (dynamic task allocation) method determines the task candidates for each robot at compile-time considering the resource constraint. It lets each robot select a task autonomously at run-time iteratively by exchanging the task allocation information with its neighbor robots. We assess the efficacy of our methods across three distinct environments: a numerical simulation, a swarm robotics simulation, and real robots. Experimental results show that the proposed methods can effectively tolerate robot failures, and the DTA method is superior to the STA method as the probability of robot failure increases. However, the STA method also exhibits consistent performance and superiority when faced with limitations in inter-robot communication. Additionally, we validate the feasibility of our method in a real-world context by conducting experiments with actual robots. Note to Practitioners—The motivation of this work is to explore how to allocate tasks efficiently to swarm robots to ensure timely completion despite occasional robot failures. In search-and-rescue scenarios, such as in the aftermath of a disaster, the effective use of swarm robots is vital, and the time taken to search is crucial to rescuing individuals within a critical time. Various approaches have been proposed to tackle this problem, taking into account time constraints. However, few studies have considered the impact of hardware constraints on robots. To address this issue, this paper proposes two new strategies to find an optimal allocation: the Static Task Allocation (STA) method and the Dynamic Task Allocation (DTA) method. Our methods are evaluated both on real robots and in simulation environments, demonstrating their suitability for practical application.}, archive = {J_TASE}, author = {Woosuk Kang and Eunjin Jeong and Sungjun Shim and Soonhoi Ha}, doi = {10.1109/TASE.2024.3389013}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3068--3085}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimization of task allocation for resource-constrained swarm robots}, volume = {22}, year = {2025}, }
• (2025). Learning an enhanced TCN-LSTM network for temperature process modeling in rotary kilns. TASE, 22, 3056--3067. (www)

  @article{ , abstract = {Accurate prediction of reaction temperature in rotary kiln is essential to realize its advanced process control and operational optimizations. However, the complexity of the physical and chemical reactions in the rotary kiln makes it difficult for the traditional mechanism model to characterize the dynamic kiln process. In this study, a deep learning-based temperature prediction model is proposed to accurately track the temperature changes during the production process of rotary kiln. The proposed model integrates a temporal convolutional network (TCN) with a long short-term memory (LSTM) network. The former enables the proposed model to be aware of local context and can effectively compute local characteristics, whereas the latter with its time-memory capability can better capture the long-term dependencies of data to extract deep-level features. To enhance the prediction accuracy, the proposed model is further improved by introducing a novel feature fusion and hybrid pooling layer to merge the original input with TCN output features, which efficiently preserve the detailed information and significantly reduce model complexity. Attention mechanism is also incorporated after the LSTM network to concentrate on the key moment information and improve the model performance. Monitoring data from a zinc rotary kiln at a field site is used for model training and testing. Results demonstrate that the proposed model can achieve the best mean square error, 0.165, exhibiting a promising prediction accuracy of the rotary kiln’s temperature. It outperforms the state-of-the-art machine learning-based prediction models such as LSTM, Gated Recurrent Unit (GRU) and TCN. Note to Practitioners—This work considers high-accuracy prediction of the temperature changes during the production process of rotary kiln. It is critical to well predict kiln tail temperature by capturing long-term dependencies and extracting global and local features from time series data. Yet, this is an extremely challenging task because rotary kiln operation is a complex thermal process with multivariate, pure hysteresis, strong interference, nonlinearity and strong coupling characteristics. As a result, these issues inhibit the establishment of a model that can accurately characterize the production process of the rotary kiln and impedes the realization of automation in volatile kiln production. Current prediction methods usually adopt mechanism-based and data-driven, there are suffer from long computation time, unknown kinetic parameters and poor accuracy. Thus, they fail to effectively learn features and accurately establish model. This work proposes a temperature prediction dynamic model named enhanced TCN-LSTM, which combines a temporal convolutional network (TCN) with a long short-term memory (LSTM) network, incorporating a unique feature fusion technique along with a hybrid pooling layer and attention layer. This integration enables the model to extract both global and local features, thereby preserving detailed information efficiently while significantly reducing model complexity and enhance the stability and robustness. Experimental results based on the zinc rotary kiln dataset demonstrate that the proposed model exhibits a promising prediction accuracy of the rotary kiln’s temperature. In the future, it can be readily implemented and applied in many industrial areas such as systems modeling, time series forecasting, and predictive maintenance.}, archive = {J_TASE}, author = {Xiaoming Wang and Yiqi Liu and Xiaojun Liang and Chaobo Zhang and Chunhua Yang and Weihua Gui}, doi = {10.1109/TASE.2024.3388709}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3056--3067}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning an enhanced TCN-LSTM network for temperature process modeling in rotary kilns}, volume = {22}, year = {2025}, }
• (2025). Mixed-gain adaption-based fault-tolerant funnel control of robotic manipulators with unknown dynamics and sensor faults. TASE, 22, 3044--3055. (www)

  @article{ , abstract = {This paper is dealt with the problem of reference tracking for the robotic manipulators with unknown dynamics and subject to multiplicative and additive sensor faults. The resulting dynamics of the closed-loop system is both unknown and structurally variable, for which the widely adopted approximation tools are not effective straightforward. Moreover, the multiplicative sensor faults result in the loss of strong controllability of the closed-loop system, which renders the conventional fault-tolerant control methods infeasible. To overcome these obstacles, an innovative fault-tolerant funnel control strategy based on a mixed-gain adaption technique is put forward in this paper. The controller involves information of neither the plant model nor the sensor faults, but does not invoke the tools for parameter identification, function approximation, data driving, iterative learning, disturbance estimation, fault estimation or fault diagnosis. It achieves reference tracking of the robotic arms with the predefined settling time and accuracy against the sensor faults. A comparative experiment on a 2-DoF serial flexible link robot is conducted to illustrate the effectiveness and advantages of the proposed approach. Note to Practitioners—This paper was motivated by the problem of fault-tolerant control (FTC) of robotic manipulators but also applies to other MIMO nonlinear plants with inputs coupling, e.g., permanent magnet synchronous motors and dynamic positioning operated vessels. The existing FTC methods work for the case where sensor faults do not alter the strong controllability of the closed-loop system. Instead, the non-strongly controllable faulty plants are considered in this paper to develop the FTC approaches to further enhance the fault tolerance of robotic manipulators. On the other hand, the involvement of sensor faults results in irregular dynamics of the closed-loop system, which cannot be addressed by the conventional approximation techniques, e.g., neural networks or fuzzy logic systems. Therefore, FTC of robotic manipulators against sensor faults that cause loss of strong controllability of control systems is challenging but significant in both academia and industry. To this end, this paper suggests a novel fault-tolerant funnel control approach based on a mixed-gain adaption technique. It is independent of the model information or the fault information, thus with a high universal property. Even so, no effort is paid for model identification or fault diagnosis, and thus the controller is simple and easy for implementation. It achieves fast accurate tracking control with prescribed settling time and accuracy, which is significant for improving efficiency and quality. The feasibility of the proposed approach is validated by a 2-DoF serial flexible link robot.}, archive = {J_TASE}, author = {Jin-Xi Zhang and Jun-Guo Song and Qingda Chen and Feng Yang}, doi = {10.1109/TASE.2024.3388592}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3044--3055}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mixed-gain adaption-based fault-tolerant funnel control of robotic manipulators with unknown dynamics and sensor faults}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time optimal control for a class of switched nonlinear systems. TASE, 22, 3033--3043. (www)

  @article{ , abstract = {This paper proposes a comprehensive framework for prescribed-time optimal switching and control (PTOSC) in switched systems. First, a new performance index function is defined, which considers the system state, specified time and accuracy, and control costs. This effectively incorporates the prescribed time control into the optimal control framework. Following this, a switched Hamilton-Jacobi-Bellman equation is derived. An event-triggered (ET) PTOSC algorithm, via reinforcement learning, is subsequently presented to solve this equation, and then the optimal control policies are derived. At each event-triggering instant, the switched controller determines which subsystem to activate, and the input controller updates the system inputs. The proposed PTOSC algorithm guarantees the stability of the switched systems and ensures the system states converge to a specified range within a specified time, all while minimizing energy consumption. Furthermore, the devised ET mechanism significantly reduces the communication burden and effectively avoids Zeno behavior. Finally, a simulation is performed to validate the proposed PTOSC algorithms’ effectiveness. Note to Practitioners—This paper addresses the critical problem in the field of switched systems of achieving prescribed-time optimal switching and control. Current practices might not adequately balance convergence accuracy, settling time, and cost-saving, which is the motivation for this paper. The presented control method is of great significance for many practical processes, such as power systems, robot control, and water-air amphibious vehicles. The feasibility of this new approach is confirmed through simulation. Future research can further refine this approach and explore how to achieve the prescribed-time optimal control for multi-agent systems or multi-player non-zero-sum games.}, archive = {J_TASE}, author = {Yan Zhang and Zhengrong Xiang}, doi = {10.1109/TASE.2024.3388456}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3033--3043}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time optimal control for a class of switched nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Learning robust model predictive control for voltage control of islanded microgrid. TASE, 22, 3021--3032. (www)

  @article{ , abstract = {This paper proposes a novel control design for voltage tracking of an islanded AC microgrid in the presence of nonlinear loads and parametric uncertainties at the primary level of control. The proposed method is based on the Tube-Based Robust Model Predictive Control (RMPC), an online optimization-based method which can handle the constraints and uncertainties as well. The challenge with this method is the conservativeness imposed by designing the tube based on the worst-case scenario of the uncertainties. This weakness is amended in this paper by employing a combination of a learning-based Gaussian Process (GP) regression and Tube-Based RMPC. The advantage of using GP is that both the mean and variance of the loads are predicted at each iteration based on the real data, and the resulted values of mean and the bound of confidence are utilized to design the tube in Tube-Based RMPC. The theoretical results are also provided to prove the recursive feasibility and stability of the proposed learning based Tube-Based RMPC. Finally, the simulation results are carried out on both single and multiple DG (Distributed Generation) units. Note to Practitioners—In this paper, we present a new way to control the voltage in an islanded microgrid to improve Power Quality (PQ). The method we propose is based on an online optimization technique called Tube-Based Robust Model Predictive Control. It can handle uncertainties and disturbances that occur when the microgrid operates independently, ensuring the voltage remains stable. However, there’s a challenge with this method. It tends to be too cautious because it assumes the worst-case scenario for uncertainties. To make the control more efficient, we improve it by combining a learning-based technique called Gaussian Process regression with Tube-Based RMPC. The advantage of using GP is that it predicts the uncertainty of the electrical devices based on real data. We use these predictions to design the control in Tube-Based RMPC more accurately. We also provide theoretical results to show that our new learning-based control is reliable and stable. We tested our approach through computer simulations on different scenarios with one or multiple power sources in the microgrid. The results show the effectiveness of our control design in regulating the voltage even with uncertain and nonlinear loads. Overall, this paper suggests a practical and reliable way to control the voltage in an independent microgrid using a combination of online optimization and learning techniques.}, archive = {J_TASE}, author = {Sahand Kiani and Ali Salmanpour and Mohsen Hamzeh and Hamed Kebriaei}, doi = {10.1109/TASE.2024.3388018}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3021--3032}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning robust model predictive control for voltage control of islanded microgrid}, volume = {22}, year = {2025}, }
• (2025). Regularized periodic gaussian process for nonparametric sparse feature extraction from noisy periodic signals. TASE, 22, 3011--3020. (www)

  @article{ , abstract = {This study proposes a nonparametric sparse feature extraction approach based on a periodic Gaussian process (PGP) for highly nonlinear sparse periodic signals, which may not be effectively modeled by conventional linear models based on user-specified dictionaries. The PGP model is reformulated as a mixed-effects model. Hence a regularization term is allowed to be imposed on the random effect of the PGP model, called regularized PGP (RPGP) in this study, for sparse feature extraction. Unlike conventional sparse models, the proposed RPGP can simultaneously model fixed and random effects (global trend and local sparsity) of the signals. A computationally scalable algorithm based on the alternating direction method of multipliers (ADMM) is tailored for RPGP to iteratively optimize the fixed and random effects. The efficient computation of RPGP is achieved by a customized circulant-based acceleration technique that utilizes fast Fourier transform on circulant matrices. The performance of RPGP is evaluated through a simulation study on synthetic signals and a case study on real vibration signals. Note to Practitioners—This work is motivated by the problem of sparse feature extraction from highly nonlinear periodic signals with a complex global trend. The key issues involved in this problem include: 1) how to model the highly nonlinear periodic signals; 2) how to separate the periodic signals and the global trend from noisy signals; 3) how to extract the sparse periodic feature. Existing approaches for sparse feature extraction assume that the signals have a constant trend and can be effectively modeled by certain dictionaries. This work proposes a novel approach that leverages a regularized PGP to simultaneously extract the sparse periodic feature and the global trend. Moreover, the nonparametric nature of the PGP model enables automatic feature extraction, eliminating the need to manually select a dictionary for sparse periodic feature extraction. The proposed approach involves four main steps: 1) collecting periodic signals from embedded sensors; 2) modeling the signals with PGP models following the fixed effects framework; 3) training the model parameters using the proposed optimization procedure; 4) extracting the global trend and sparse periodic feature based on the model parameters. This paper demonstrated the application of rolling bearing vibration signals, while the effectiveness of the proposed model is not limited to this topic. Other periodic signals with similar characteristics can also be analyzed by applying the proposed model. In the future, this approach will be extended to handle sparse feature extraction from pseudo-periodic signals in a more complex noise environment.}, archive = {J_TASE}, author = {Yongxiang Li and Yunji Zhang and Jianguo Wu and Min Xie}, doi = {10.1109/TASE.2024.3387833}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {3011--3020}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Regularized periodic gaussian process for nonparametric sparse feature extraction from noisy periodic signals}, volume = {22}, year = {2025}, }
• (2025). Bumpless transfer hybrid non-fragile finite-time control for markovian jump systems and its application. TASE, 22, 2998--3010. (www)

  @article{ , abstract = {In this article, the issue of bumpless transfer hybrid non-fragile finite-time $H_{\infty} $ control for Markovian jump systems (MJSs) is investigated, where the transition rates are partially available. The non-fragile strategy with tolerating both additive and multiplicative perturbations is designed, which greatly relaxes the application scope of the traditional controller. The bumpless transfer control idea is introduced to depict the transient behavior caused by a jumping controller. First, a bumpless transfer constraint condition is provided to restrict the amplitude of the hybrid non-fragile jumping controller, for which the additive and multiplicative perturbations are considered. Then, a bumpless transfer hybrid non-fragile controller is developed to guarantee the solvability of the finite-time $H_{\infty} $ control issue for MJSs with partially available transition rates. Finally, an electronic circuit system example is applied to illustrate the usefulness of the proposed bumpless transfer hybrid non-fragile control approach. Note to Practitioners—This article is motivated by the finite-time hybrid non-fragile $H_{\infty} $ control issue of MJSs with the bump limitation constraint. The bumpless transfer is often encountered in the Markovian jumping control field because it is usually impossible to implement unlimited control signals or bumpy control signals and is a main source of instability, and degradation of performance of MJSs. The traditional bumpless transfer technique because of the constant control gain cannot handle effectively the non-fragile bumpless transfer problem for MJSs. In this paper, the restriction of the constant control gain is relaxed. We develop a bumpless transfer control strategy considering a strong non-fragility of the control gain. In contrast to the existing results, a non-fragile bumpless transfer can tolerate the co-existing additive and multiplicative perturbations with a wider application. This study presents the method for practitioners interested in bumpless transfer controller design.}, archive = {J_TASE}, author = {Dong Yang and Qingchuan Feng and Jing Xie and Tao Liu}, doi = {10.1109/TASE.2024.3387659}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2998--3010}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bumpless transfer hybrid non-fragile finite-time control for markovian jump systems and its application}, volume = {22}, year = {2025}, }
• (2025). Real-time melt pool homogenization through geometry-informed control in laser powder bed fusion using reinforcement learning. TASE, 22, 2986--2997. (www)

  @article{ , abstract = {This paper presents a real-time geometry-informed control strategy to homogenize melt pool measurements in laser powder bed fusion (L-PBF) using reinforcement learning. The learning control strategy incorporates geometric information of the scan path as well as in-situ melt pool measurements to compute the laser power signal for reducing in-process melt pool inhomogeneities. First, we design and validate a data-driven model to train the reinforcement learning agent in simulation, with the goal of reducing the amount of experimental data needed for training. Using this simulation-based training approach has the added benefit of avoiding unsafe or infeasible experiments, an issue that is often encountered in training the reinforcement learning agent. After training, the learned control strategy attenuates the 1-norm error by 37% and standard deviation by 39% in simulation. We then deploy this learned control strategy in an experimental test bed for a new scan geometry. In this test scenario, the policy achieves a 30% reduction in error, and a 36% reduction in melt pool signal variation, thereby illustrating the potential of reinforcement learning in real-time geometry-agnostic control for L-PBF. Finally, we demonstrate that the reinforcement learning agent delivers the same level of performance as a model-based feedforward controller with PID feedback, with $20\times $ less computational time for a single geometry. Note to Practitioners—This work was motivated by the need to develop a practical control algorithm for L-PBF systems. Because L-PBF systems manufacture customized on-demand geometries, it is critical that the control strategy is extendable to and easily optimized for each geometry. Specifically, this effort develops an efficient and robust reinforcement learning control algorithm that can be used across novel part geometries, once trained. The control strategy is designed using a simulation-to-real approach, which is key for avoiding extensive training effort and avoids unsafe training experiments.}, archive = {J_TASE}, author = {Bumsoo Park and Alvin Chen and Sandipan Mishra}, doi = {10.1109/TASE.2024.3386882}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2986--2997}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time melt pool homogenization through geometry-informed control in laser powder bed fusion using reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Real-time trajectory planning and obstacle avoidance for Human–Robot co-transporting. TASE, 22, 2969--2985. (www)

  @article{ , abstract = {In this paper, a real-time obstacle avoidance approach and a trajectory planning method are proposed to avoid collisions to move an object jointly by a human and an omnidirectional mobile robot. Different from many existing approaches of local trajectory planning, the proposed method called Multiscale Local Perception Region Approach (MLPRA) is specially designed for obstacle avoidance of omnidirectional wheeled robots with direction constraints of human guidance, which can respond fast to dynamic obstacles and guarantee the safety of the robot. To solve the problem that the position of the human hand is difficult to obtain by visual sensing due to visual occlusion, a simple mechanism that can indirectly measure the change of the position of the human hand is designed, and further a method to follow the human’s intention based on this mechanism is proposed. Finally, the simulation environment on the Gazebo simulation platform is built to verify the feasibility and effectiveness of our proposed methods. Experimental results show that after embedding proposed methods into the omnidirectional mobile robot, obstacles can be effectively avoided in co-transporting processes.Note to Practitioners—The motivation of this paper is focusing on obstacle avoidance of omnidirectional mobile robots in human-robot co-transporting, application scenarios concentrated in factories and logistics warehouses, such as the mobile robot collaborates with a human (a robot teleoperated by human) transporting a table. The existing obstacle avoidance algorithm can be regarded as a local trajectory planning problem, which cannot directly adapt to co-transporting tasks in real-time. Considering the specific task if the robot cannot cooperate with human in real-time, the carried object will fall. In this paper, a local trajectory planning method is proposed, regarding the obstacle generating virtual repulsive force, regarding the human action on the robot as the virtual gravitational force, and the mobile robot under two virtual forces not only follows human action in co-transporting, but also avoids obstacles. The method shows quick calculation and good real-time performance. The method proposed is evaluated by co-transporting in Gazebo and experiments based on omnidirectional mobile robots, and ensured that the object being carried does not fall and robot can avoid collisions in the unknown environments without positioning systems.}, archive = {J_TASE}, author = {Xinbo Yu and Xiong Guo and Wei He and Muhammad Arif Mughal and Dawei Zhang}, doi = {10.1109/TASE.2024.3386814}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2969--2985}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time trajectory planning and obstacle avoidance for Human–Robot co-transporting}, volume = {22}, year = {2025}, }
• (2025). Dual-channel event-triggered quantitative LFC for micro-grid system via delay-square-dependent lyapunov functional approach. TASE, 22, 2956--2968. (www)

  @article{ , abstract = {This paper investigates the dual-channel event-triggered quantitative load frequency control for discrete-time micro-grid system via a delay-square-dependent Lyapunov-Krasovskii functional approach. Firstly, a novel dual-channel dynamic event-triggered quantified control strategy has been proposed to enhance time-delay tolerance and the utilization of communication resources. Secondly, a discrete-time Lyapunov-Krasovskii functional incorporates the square of the delay is introduced to fully deploy the desired delay information about the system. Thirdly, improved stability criteria with strict dissipative performance indexes are established by employing the proposed delay-square-dependent Lyapunov-Krasovskii functional. Finally, two numerical examples are presented to illustrate the effectiveness and superiority of the proposed control strategy. Note to Practitioners—Micro-grid typically consist of distributed power resources, loads and energy storage devices, those are deployed spatially and remotely over some wireless and digital network mediums. The constrained resource issue, posed by finite bandwidth and processing capabilities of battery-powered system components. This paper aims at designing a reasonable control strategy to deal with issues such as band-limited channels and communication delays that undermine system stability in the process of network transmission. Specifically, we propose a dual-channel asynchronous dynamic event-triggered control strategy to save the limited channel bandwidth. To mitigate the impact of time delays, we comprehensively consider the state-dependent delays and control-related delays, further establishing a novel Lyapunov functional incorporating delay squared terms. The proposed control approach is validated to be capable of better reduce the consumption of bandwidth and achieve the desired stability through theoretical investigation and simulation cases.}, archive = {J_TASE}, author = {Qishui Zhong and Hanmei Zhou and Yinsheng Chen and Jin Yang and Kaibo Shi and Lei Yan and Shouming Zhong}, doi = {10.1109/TASE.2024.3386785}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2956--2968}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual-channel event-triggered quantitative LFC for micro-grid system via delay-square-dependent lyapunov functional approach}, volume = {22}, year = {2025}, }
• (2025). Automated control of microparticle swarm in a rotating gradient-based magnetic field. TASE, 22, 2946--2955. (www)

  @article{ , abstract = {The magnetic micromanipulation of swarm microparticles has attracted considerable attention because of its advantages of non-invasiveness, high drug-carrying capacity, and easy observation in the targeted delivery in in-vivo environments. This paper presents an automated control scheme for the magnetic micromanipulation of microswarms in a rotating gradient-based field. Different from the rotating uniform magnetic field generated by Helmholtz coils, the rotating gradient-based field is a type of convergent field established by sequentially powering each coil of the electromagnetic coil system. By changing the coil currents, the field can rotate while driving the microswarm to a pre-determined position, facilitating the swarm localization and tracking. According to the preliminary motion characterization of the swarm in the rotating gradient-based field, an intuitive trapping dynamic model which can simplify the analysis of swarm dynamics is established to facilitate controller design. Based on this model, a super-twisting sliding mode estimator is first designed to estimate the position of the microswarm as well as the disturbances caused by parameter variations and unmodeled dynamics. A robust controller is then developed based on the estimator. In this way, closed-loop manipulation of the microswarm to follow a desired trajectory in the rotating gradient-based field is realized, and the system’s behavior has been significantly improved due to the capability to estimate disturbances. The proposed control scheme for the rotating gradient-based field has the potential to avoid volume loss and unexpected drug diffusion of the swarm when facing complex in-vivo environments. The stability of the control scheme is proved by the Lyapunov approach. Experiments are finally performed to demonstrate the effectiveness of the proposed control approach in a collision-free environment and in a simulated channel.Note to Practitioners—The motivation of this study is to realize automated feedback control of the microparticle swarm in a rotating gradient-based magnetic field. The rotating gradient-based magnetic field is a type of convergent field that can rotate while driving the microswarm to a pre-determined position. This magnetic actuation method facilitates microswarm tracking and enables microagents to overcome static friction with the bottom substrate by rotating them, thereby inducing movement. However, existing research on the rotating gradient-based magnetic field concentrates on moving the microswarm to the targeted position without real-time visual guidance in an open-loop manner, which provides less reliability and convenience compared to real-time closed-loop control. Moreover, it will result in unavoidable volume loss when collisions happen in the complicated environments. In addition, the existing swarm dynamics in such a field need to be simplified to apply in controller design. To solve the above problems, an automated point-to-point navigation control scheme is proposed in this study. An intuitive trapping model is first established to simplify the swarm dynamics. Then, a robust controller with a super-twisting algorithm-based estimator is designed based on the model to manipulate the microswarm. Experimental results have shown that the proposed method can successfully control the microparticle swarm in collision-free environments and a simulated channel with considerable accuracy in the rotating gradient-based magnetic field. The proposed control scheme has the potential to avoid volume loss and unexpected drug diffusion of the swarm when facing complex in-vivo vascular environments, particularly those involve fluids with completely different directions in a bifurcation, such as the vascular network of lymphatic vessels and blood vessels.}, archive = {J_TASE}, author = {Hui Cao and Liuxi Xing and Jingrong Hu and Hangjie Mo and Dong Sun}, doi = {10.1109/TASE.2024.3386745}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2946--2955}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated control of microparticle swarm in a rotating gradient-based magnetic field}, volume = {22}, year = {2025}, }
• (2025). Resilient-sampling-based bipartite synchronization of cooperative-antagonistic neural networks with hybrid attacks: Designing interval-dependent functions. TASE, 22, 2935--2945. (www)

  @article{ , abstract = {This paper investigates the bipartite synchronization problem of cooperative-antagonistic neural networks (CANNs) that suffer from hybrid attacks, i.e., denial-of-service (DoS) attacks and replay attacks. A resilient sampled-data control scheme is proposed to deal with the hybrid attacks. In addition, the Laplacian matrix with zero-row-sum is obtained based on the coordinate transformation method. On this basis, an easy-to-handle error system is constructed by integrating the sampling scheme, cooperative-antagonistic interactions and hybrid attacks. Subsequently, an interval-dependent function is designed by considering the characteristics of both replay and DoS attacks. Based on this, using Lyapunov function methods, inequality techniques, and other mathematical skills, some sufficient criteria are obtained for the bipartite synchronization of CANNs. Finally, three algorithms are proposed to minimize the coupling strength, or maximize the replay attack rate or the DoS attack rate, respectively. The effectiveness of the proposed control schemes and the advantages of the interval-dependent function are verified through numerical examples. Note to Practitioners—This work addresses the bipartite synchronization control problem of CANNs, which can be applied to some practical systems such as drone bidirectional formation and team competition. For example, the dynamics of drones within the same formation, including their velocity and position, are consistent with each other but opposite to those of other teams. Additionally, network systems across various industries often encounter DoS attacks or replay attacks during network transmission. A resilient sampled-data control scheme is proposed, and a Lyapunov function dependent on intervals is designed,taking into account the characteristics of the attack. Then, some algorithms are presented to minimize the allowable coupling strength or maximizing the allowable replay attack rate and DoS attack rate. Preliminary simulation experiments indicate that this sampling scheme is feasible. The types of network attacks are diverse. Therefore, obtaining system synchronization control schemes for more complex network attacks is a challenge.}, archive = {J_TASE}, author = {Xindong Si and Zhen Wang and Xia Huang and Yingjie Fan and Hao Shen}, doi = {10.1109/TASE.2024.3386699}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2935--2945}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient-sampling-based bipartite synchronization of cooperative-antagonistic neural networks with hybrid attacks: Designing interval-dependent functions}, volume = {22}, year = {2025}, }
• (2025). Fixed-time fuzzy resilient consensus tracking of nonlinear MASs under DoS attacks and actuator faults. TASE, 22, 2923--2934. (www)

  @article{ , abstract = {This paper addresses the output feedback fixed-time consensus tracking problem for nonlinear multiagent systems (MASs) subject to aperiodic DoS attacks and actuator faults. Different from existing resilient control methods for MASs where the attackers only destroy the connectivity between agents, the sensor-to-controller communication channels of each agent are simultaneously blocked in this paper. To overcome these difficulties, we first propose a novel fixed-time stability criterion that incorporates the characteristics of DoS attacks. Then, a fuzzy state estimator is constructed to estimate the immeasurable states as well as compensate for the adverse impact of attacks. In addition, novel adaptive laws are developed to handle an infinite number of actuator failures by estimating the bounds of failure parameters. Based on the Lyapunov stability theory, a fixed-time output feedback resilient control strategy is designed to guarantee that consensus tracking can be achieved, while the boundedness of all closed-loop signals in the MAS is ensured. At last, simulation studies are performed to validate the performance of the designed control law. Note to Practitioners—Many real-world systems, including robotic systems, manipulator systems and intelligent vehicle systems, can be represented as strict-feedback MASs. Note that these systems are notably susceptible to network attacks and operational faults. Consequently, ensuring the continued functionality of controlled MASs with DoS attacks and actuator faults has emerged as a paramount research focus. Furthermore, in the context of systems necessitating high transient performance standards, the concept of fixed-time convergence holds practical significance. In light of the challenges posed by DoS attacks and actuator faults, a novel fixed-time fuzzy fault-tolerant resilient consensus tracking approach is designed for nonlinear MASs. The proposed control laws cater to practical MASs which are characterized in strict-feedback form subject to unknown nonlinearities. Preliminary simulation results indicate the feasibility and efficacy of our developed approach.}, archive = {J_TASE}, author = {Bo Xu and Yuan-Xin Li and Zhongsheng Hou}, doi = {10.1109/TASE.2024.3386603}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2923--2934}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fixed-time fuzzy resilient consensus tracking of nonlinear MASs under DoS attacks and actuator faults}, volume = {22}, year = {2025}, }
• (2025). Deep learning-based approximation of model predictive control laws using mixture networks. TASE, 22, 2909--2922. (www)

  @article{ , abstract = {In recent years, researchers have proposed the approximation of model predictive control (MPC) using deep neural networks (DNNs). However, a limitation arises as DNNs inherently offer one-to-one mappings, posing a challenge when multiple optimal control inputs correspond to each system state, thereby leading to one-to-many mappings. Therefore, we propose an alternative scheme using mixture networks (MNs) with components of probability (density) distributions in the output layer. This method uses conditional probabilities offered by combining several estimated probability distributions, enabling generating multiple control inputs with the highest probabilities. Notably, this approach is applicable to several problems by choosing a suitable probability distribution, such as using a Gaussian distribution for nonlinear problems and a Bernoulli distribution for mixed-integer linear programming (MILP) problems. We investigate two case studies illustrating that the mixture network-based approximation outperforms the DNN-based approximation. Note to Practitioners—This article proposes using mixture (density) networks, a type of machine learning technique, to enable the online implementation of model predictive control. The prohibitive computation cost associated with model predictive control poses a significant challenge when implementing it in complex nonlinear and MILP problems. While approximation methods of control laws using deep neural networks have been studied to address this issue, it is unsuitable for problems where each state has multiple optimal control inputs. In contrast, our proposed approach can accurately approximate the control laws for these problems. This notable feature can facilitate the online implementation of model predictive control in a wide range of nonlinear and MILP problems.}, archive = {J_TASE}, author = {Morimasa Okamoto and Jiayang Ren and Qiangqiang Mao and Jianfeng Liu and Yankai Cao}, doi = {10.1109/TASE.2024.3386596}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2909--2922}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Deep learning-based approximation of model predictive control laws using mixture networks}, volume = {22}, year = {2025}, }
• (2025). A novel mixed integer programming model with precedence relation-based decision variables for non-cyclic scheduling of cluster tools. TASE, 22, 2893--2908. (www)

  @article{ , abstract = {Cluster tools, which are extensively used in semiconductor and display manufacturing, offer the capability to perform multiple processing steps within a single tool. Many companies have recently been reducing wafer lot sizes due to diversified customer demands and circuit width reduction, resulting in more frequent transient and non-cyclic periods. We therefore present a novel mixed integer programming (MIP) model that can handle these non-cyclic scheduling problems of cluster tools. Our model is specifically designed to efficiently handle various wafer flows, accommodating both single- and dual-armed robots, while ensuring a shorter computation time, compared to the previous formulations. We first model cluster tools with a timed Petri net (TPN) and develop several precedence relations between transitions by analyzing the characteristics of the scheduling problems. These precedence relations are incorporated into a TPN by introducing additional places and arcs. This modification helps reduce the overall number of decision variables involved in the scheduling problem. We then develop an MIP model which specifies the precedence relations between transitions, as opposed to the position-based decision variables used in the previous studies. The proposed MIP model is capable of handling various flow scenarios encountered in cluster tools, including serial, parallel, concurrent, and re-entrant flows with time window constraints. Note to Practitioners—In response to the diverse demands of customers, the use of larger wafer sizes, and the need for smaller circuit widths, operating cluster tools in a cyclic manner has become increasingly challenging. Cyclic scheduling, where the robot repeats a fixed sequence while assuming identical wafers, is no longer viable in such scenarios. Unfortunately, previous research on cluster tool scheduling has predominantly focused on cyclic scheduling, failing to address the growing importance of non-cyclic scenarios. To address this gap, we propose an innovative and efficient mixed integer programming (MIP) model capable of handling non-cyclic scheduling problems in cluster tools. Our model accommodates various flow types, including serial, parallel, concurrent, and re-entrant flows. Through comparative analysis, we demonstrate that our proposed model outperforms the well-known MIP model that relies on position-based decision variables. We believe that our proposed model holds practical value as it exhibits relatively short computation times. Additionally, widely-used commercial solvers, such as CPLEX or Google OR Tools, can seamlessly implement the model, making it easily accessible and applicable in real-world scenarios.}, archive = {J_TASE}, author = {Jeongsun Ahn and Hyun-Jung Kim}, doi = {10.1109/TASE.2024.3386562}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2893--2908}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel mixed integer programming model with precedence relation-based decision variables for non-cyclic scheduling of cluster tools}, volume = {22}, year = {2025}, }
• (2025). Digital twin assisted economic dispatch for energy internet with information entropy. TASE, 22, 2881--2892. (www)

  @article{ , abstract = {As the percentage of renewable energy in the Energy Internet (EI) gradually increases, how to deal with the uncertainty of renewable energy in the economic dispatch problem (EDP) becomes an important issue. This paper proposes a digital twin (DT) assisted economic dispatch strategy for EI with information entropy. First, we leverage the storage capacity of the DT and an extensive historical data set to provide a theoretical framework for quantifying uncertainty of renewable energy. Second, a renewable energy cost function based on the maximum entropy principle, confidence interval, and penalty factor is proposed to model the renewable energy resources considering the uncertainty. Further, we design a fully distributed Newton-surplus-based optimization algorithm. This algorithm achieves fast second-order convergence to ensure the real-time performance of the DT-assisted economic dispatch framework and overcome the asymmetry caused by the directed communication network. In addition, we give theoretical proof that the Newton-surplus-based algorithm can converge to the global optimal point. Finally, simulations validate the effectiveness of the proposed algorithm.Note to Practitioners—The essence of EDP is to minimize the total costs through optimal resource allocation while ensuring compliance with all operational constraints. With the increasing penetration of renewable energy resources, their strong stochasticity and uncertainty pose challenges to achieve reliable dispatch strategy. To address this issue, this paper presents the DT-assisted economic dispatch framework, model, and method to quantify the uncertainty of renewable energy resources and achieve distributed economic dispatch with fast convergence speed for EI. Our research is beneficial for practitioners to understand how to use the DT and information entropy to deal with the uncertain of renewable energy resources. The theory and simulation results demonstrate the correctness and effectiveness of the proposed method.}, archive = {J_TASE}, author = {Rufei Ren and Yushuai Li and Qiuye Sun and Xiangpeng Xie and Lei Liu and David Wenzhong Gao}, doi = {10.1109/TASE.2024.3386358}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2881--2892}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Digital twin assisted economic dispatch for energy internet with information entropy}, volume = {22}, year = {2025}, }
• (2025). Automatic optimization of tolerance ranges for model-driven runtime state identification. TASE, 22, 2867--2880. (www)

  @article{ , abstract = {For continuously checking and updating the virtual representation of a real system during operation, the continuous sensing and interpretation of raw sensor data is a must. The challenge is to bundle sensor value streams (e.g., from IoT networks) and aggregate them to a higher logical state level to enable process-oriented viewpoints and to handle uncertainties about sensor measurements and state realization precision. To address these uncertainties, so-called “tolerance ranges” must be defined in which logical states are detected during operation with acceptable deviations. Specifying such tolerance ranges manually is a time-consuming, error-prone task and often not feasible due to the huge associated value search space. To tackle this challenge, the problem is turned into an optimization problem in this paper. For this purpose, we present a framework based on meta-heuristic search that enables the automatic configuration of tolerance ranges based on available execution traces of multiple sensor value streams. An exploratory study evaluates the approach. For this purpose, we implemented a lab-sized demonstrator of a five-axis grip arm robot, which we continuously monitored during operation in a simulated environment. The evaluation shows the advantage of using meta-heuristic optimizers such as Harmony Search or Genetic Algorithm to identify stable tolerance ranges automatically for state detection at runtime. Note to Practitioners—Monitoring sensor values streams is nowadays a frequently employed technique in many automation domains. However, combining and mapping single value streams to higher-level state-based representations such as state machines or other design-time related models is a major challenge due to measurement and realization precision uncertainties. Thus, simply mapping monitored raw data to these design descriptions can lead to falsely identified or missed states. To improve this situation, we present an approach that provides a mechanism to continuously analyze data streams during operation by automatically finding appropriate tolerance ranges to detect realized system states. The approach uses a small set of annotated execution traces and meta-heuristic searchers to derive optimal tolerance ranges, which provide high correctness and completeness of the identified system states. This approach represents the basis for building a “vertical bridge” from the operation technology layer considering pure sensor data streams to the IT layer where state-based process views are provided to perform monitoring and analytics, e.g., by using process mining.}, archive = {J_TASE}, author = {Sabine Sint and Alexandra Mazak-Huemer and Martin Eisenberg and Daniel Waghubinger and Manuel Wimmer}, doi = {10.1109/TASE.2024.3386313}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2867--2880}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic optimization of tolerance ranges for model-driven runtime state identification}, volume = {22}, year = {2025}, }
• (2025). Robust tube-based reference tracking nonlinear model predictive control for wind turbines. TASE, 22, 2854--2866. (www)

  @article{ , abstract = {Recently, a research effort to extend nonlinear model predictive control (NMPC) methods from setpoint stabilization to reference tracking has been felt increasingly. On the other hand, applying robust NMPC approaches to a larger class of systems subject to disturbances and uncertainties often results in the control approach’s complexity and a significant increase in computational burden. Therefore, this study proposes a general reference-tracking, robust tube-based NMPC for nonlinear systems. The system is subject to input– and state–dependent uncertainties. Robust recursive feasibility and constraint fulfillment are guaranteed by the employment of suitable terminal components. The proposed control approach includes two algorithms for offline and online computations. In the offline algorithm, the tube shape is formed based on the stabilizing feedback and the Lyapunov function. In the online implementation, only the bounds related to these functions are used, which makes the proposed controller easy to implement. In addition, to facilitate efficient real-time implementation, a parallel Newtonian-type approach is applied to solve the optimization problem of NMPC. The proposed control scheme is designed to control wind turbines in the presence of additive and parametric disturbances and is compared with the nominal tracking NMPC scheme and two other newly developed methods. Simulations are performed for fatigue and extreme load scenarios, and results show that the proposed controller has superior robust performance compared to the other methods. Note to Practitioners—Model predictive control is a widely used industry method for advanced process control. In the field of robust MPC, tube-based MPC provides a suitable compromise between complexity and performance. In addition, in most industrial applications, the purpose of control goes beyond stabilizing set points and tends to include tracking time-varying references. Hence, this paper proposes a general reference-tracking, robust tube-based NMPC for nonlinear systems. In the offline design of the proposed controller, only the calculation of scalar bounds is needed, simplifying the online implementation of the controller. In addition, in the real-time implementation of the proposed controller, a parallel Newtonian-type method is used, which makes it more suitable for practical applications. Among all wind turbine control strategies, NMPC is emerging as a strong choice from the point of view of optimal power extraction and mechanical load reduction. Therefore, a robust NMPC approach is proposed here to track time-varying references dependent on wind speed for wind turbines exposed to external disturbances, static frictions, and model uncertainties.}, archive = {J_TASE}, author = {Mohammad Soleymani and Mehdi Rahmani and Nooshin Bigdeli}, doi = {10.1109/TASE.2024.3385714}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2854--2866}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust tube-based reference tracking nonlinear model predictive control for wind turbines}, volume = {22}, year = {2025}, }
• (2025). A fully-distributed resilient dynamic output feedback consensus control for MASs under markovian DoS attacks. TASE, 22, 2844--2853. (www)

  @article{ , abstract = {This paper is concerned with the resilient leader-following consensus control (LFCC) of multi-agent systems (MASs) subjected to denial-of-service (DoS) attacks. Unlike most existing studies, the considered DoS attacks can independently jeopardize different communication links and are modeled using the Markovian switching phenomenon. First, a cooperative fully-distributed estimator is introduced that enables the follower agents to estimate the leader’s state. The proposed cooperative estimator relaxes the requirement for knowledge about the exact leader state during the occurrence of DoS attacks. The cooperative estimator is then used to solve the resilient LFCC of MASs under two different control protocols. The dynamic state feedback (DSF)-based consensus protocol is first used to solve the fully-distributed resilient LFCC problem with cooperative estimation of the leader state. Then, to overcome the limitation of the follower’s state unavailability, a novel dynamic output feedback (DOF)-based control protocol is introduced to solve the resilient consensus problem. The design conditions for the proposed estimators and controllers are derived using the Lyapunov theory. Unlike existing studies, the proposed strategy 1) employs estimated states for feedback, 2) is fully-distributed, and 3) is independent of bounds on attack frequency or duration. Finally, the applicability of the presented strategies is validated through a simulation example involving the F-18 aircraft model. Note to Practitioners—The purpose of the present work is to provide an output feedback approach for the LFCC of MASs by incorporating cyber attacks. The cyber attacks in the form of DoS while communicating over a network have been considered. The state estimation problem over a network under DoS attacks on communication links has been solved by using the cooperative estimator equation. Then, the adaptive and smoothing coupling weights are updated by applying the local estimation error. The proposed approach employs an adaptive mechanism for handling DoS attacks over the communication topology to deal with unknown connections. This mechanism does not require knowledge of the topology between agents and, consequently, does not require the central unit for implementation. Then, two approaches are used to solve the problem of resilient cooperative control: 1) by employing the estimation of the leader’s state only; 2) by employing both estimates of the leader’s and followers’ states. The proposed approach can be applied practically by solving cooperative and conventional observers along with adaptation weights, recursively, in addition to an algebraic equation for estimated state feedback. Moreover, the presented method can deal with independent attacks on different communication channels, which is a practical scenario. The proposed approach can be applied to consensus and formation control applications under cyber attacks, like the formation of aircrafts, consensus in robots, and synchronization in power systems. The presented approach has been applied to attain consensus in F-18 aircrafts, and simulation results are shown.}, archive = {J_TASE}, author = {Muhammad Ahsen Ali and Muhammad Rehan and Naeem Iqbal and Abdul Basit and Choon Ki Ahn}, doi = {10.1109/TASE.2024.3385489}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2844--2853}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A fully-distributed resilient dynamic output feedback consensus control for MASs under markovian DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Multi-agent deep reinforcement learning for persistent monitoring with sensing, communication, and localization constraints. TASE, 22, 2831--2843. (www)

  @article{ , abstract = {Determining multi-robot motion policies for persistently monitoring a region with limited sensing, communication, and localization constraints in non-GPS environments is a challenging problem. To take the localization constraints into account, in this paper, we consider a heterogeneous robotic system consisting of two types of agents: anchor agents with accurate localization capability and auxiliary agents with low localization accuracy. To localize itself, the auxiliary agents must be within the communication range of an anchor, directly or indirectly. The robotic team’s objective is to minimize environmental uncertainty through persistent monitoring. We propose a multi-agent deep reinforcement learning (MARL) based architecture with graph convolution called Graph Localized Proximal Policy Optimization (GALOPP), which incorporates the limited sensor field-of-view, communication, and localization constraints of the agents along with persistent monitoring objectives to determine motion policies for each agent. We evaluate the performance of GALOPP on open maps with obstacles having a different number of anchor and auxiliary agents. We further study 1) the effect of communication range, obstacle density, and sensing range on the performance and 2) compare the performance of GALOPP with area partition, greedy search, random search, and random search with communication constraint strategies. For its generalization capability, we also evaluated GALOPP in two different environments– 2-room and 4-room. The results show that GALOPP learns the policies and monitors the area well. As a proof-of-concept, we perform hardware experiments to demonstrate the performance of GALOPP. Note to Practitioners—Persistent monitoring is performed in various applications like search and rescue, border patrol, wildlife monitoring, etc. Typically, these applications are large-scale, and hence using a multi-robot system helps achieve the mission objectives effectively. Often, the robots are subject to limited sensing range and communication range, and they may need to operate in GPS-denied areas. In such scenarios, developing motion planning policies for the robots is difficult. Due to the lack of GPS, alternative localization mechanisms, like SLAM, high-accurate INS, UWB radio, etc. are essential. Having SLAM or a highly accurate INS system is expensive, and hence we use agents having a combination of expensive, accurate localization systems (anchor agents) and low-cost INS systems (auxiliary agents) whose localization can be made accurate using cooperative localization techniques. To determine efficient motion policies, we use a multi-agent deep reinforcement learning technique (GALOPP) that takes the heterogeneity in the vehicle localization capability, limited sensing, and communication constraints into account. GALOPP is evaluated using simulations and compared with baselines like random search, random search with ensured communication, greedy search, and area partitioning. The results show that GALOPP outperforms the baselines. The GALOPP approach offers a generic solution that be adopted with various other applications.}, archive = {J_TASE}, author = {Manav Mishra and Prithvi Poddar and Rajat Agrawal and Jingxi Chen and Pratap Tokekar and P. B. Sujit}, doi = {10.1109/TASE.2024.3385412}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2831--2843}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-agent deep reinforcement learning for persistent monitoring with sensing, communication, and localization constraints}, volume = {22}, year = {2025}, }
• (2025). A modular direct description method with low memory usage and execution time for hysteresis modeling and compensation of piezoelectric actuators. TASE, 22, 2819--2830. (www)

  @article{ , abstract = {This paper proposes a modular direct description method based on the relatively weak interdependence among hysteresis properties to establish a hysteresis model that includes the major curves fitter (MaCF) module, the turning points recorder (TPR) module, and the minor curves fitter (MiCF) module. The MaCF module utilizes the Fourier series fitting method to fit the major hysteresis curves, improving model scalability in conjunction with the modular approach. The TPR module is established to manage the turning points. A memory optimization strategy is further proposed to optimize the model in terms of memory consumption and execution time. The MiCF module then employs a two-stage curve fitting strategy to fit the minor hysteresis curves accurately. Finally, three modules are cascaded to construct the hysteresis model. Experiments have been performed to validate the excellent performance of the developed model in terms of model accuracy, memory usage, and execution time. The results indicate that the model only needs 0.28 KB of memory space to achieve accurate prediction and compensation of hysteresis phenomenon, the standardized normalized maximum error ( $NME$ ) of hysteresis curve prediction is about 1%, and the $NME$ value of compensation error is 1.37%. Meanwhile, the average execution time of the model is less than $1~\mu \text{s}$ with a variance of 0.04. Note to Practitioners—The presented modular direct description (MDD) method can accurately describe and compensate hysteresis phenomenon. The MDD-based model offers the advantage of small memory consumption, making it suitable for application in memory-constrained scenarios, such as offline deployment of microcontrollers and embedded systems. Meanwhile, the execution time of this model is short and stable, which can be applied in applications requiring a high computational period, such as the precise position control of stick-slip drive motors, the precise speed control of ultrasonic motors, and other scenarios with the presence of high-frequency inputs. The MDD method has the potential to become a general modeling method for solving hysteresis challenges, and can be used by simply modifying the relevant modules according to the hysteresis rules in different domains, such as in the fields of artificial muscles and shape memory alloys.}, archive = {J_TASE}, author = {Zengsheng Li and Xinjian Fan and Dunfa Long and Zhan Yang and Chaoyang Shi}, doi = {10.1109/TASE.2024.3384591}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2819--2830}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A modular direct description method with low memory usage and execution time for hysteresis modeling and compensation of piezoelectric actuators}, volume = {22}, year = {2025}, }
• (2025). Optimization-based adaptive assistance for lower limb exoskeleton robots with a robotic walker via spatially quantized gait. TASE, 22, 2806--2818. (www)

  @article{ , abstract = {Gait training with human-like gait patterns can be provided by lower limb exoskeletons (LLEs) for patients with gait impairments. For patients with little effort to keep balance, using a mobile robotic walker to assist gait training with LLEs is an effective way. Since gait patterns are varying with walking speeds, it is a critical issue to coordinated control the robotic walker and the exoskeleton to obtain a natural and human-like walking posture. In this paper, a novel adaptive assistance approach named SQG-OPT is proposed to tackle the problem, which comprises of two parts: the Spatially Quantized Gait (SQG) and the optimization. The SQG generates reference joint angles and reference trajectory of the Center Of Mass (COM) for the human-exoskeleton system in space domain. The optimization part is constructed to convert the reference joint angles from the space domain to the time domain, which is based on the dynamics model of the human-exoskeleton-walker system and adaptive to different walking speeds. The proposed approach has been tested on the robot simulation platform CoppeliaSim, the experimental results indicate that the proposed approach can generate human-like gait patterns for different walking speeds from 0 to 0.8 m/s. Additionally, in comparison with other methods, the proposed approach has a better performance on the movement tracking of the COM for a natural walking posture. Note to Practitioners—The coordinated control is important for the exoskeleton robot with a mobile robotic walker, one of the potential challenge is the adaptive coordinated motion planning for these robotic devices. The proposed approach is for the coordinated control of the exoskeleton robots with a robotic walker and adaptive to different walking speeds, which may inspires more extended coordinated control strategies for human-exoskeleton systems in more gait training applications. The proposed approach is also potential for the coordinated motion planning of the other human-centered assistance robots, such as the wheeled walking assistance robots for the elderly.}, archive = {J_TASE}, author = {Chaobin Zou and Zhinan Peng and Long Zhang and Fengjun Mu and Rui Huang and Hong Cheng}, doi = {10.1109/TASE.2024.3384515}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2806--2818}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Optimization-based adaptive assistance for lower limb exoskeleton robots with a robotic walker via spatially quantized gait}, volume = {22}, year = {2025}, }
• (2025). Funnel-based adaptive predefined-time leader-following output-feedback optimal control for second-order nonlinear multi-agent systems. TASE, 22, 2794--2805. (www)

  @article{ , abstract = {This study proposes a novel adaptive funnel-based approach for optimal formation control in second-order nonlinear multi-agent systems (MASs). The nonlinear system considered includes unknown dynamics, with only one measurable output variable. The states, not directly measurable, are reconstructed using fuzzy state observers. In the context of an actor-critic framework, a novel funnel optimal predefined-time formation controller is developed. The primary objective is to achieve practical predefined-time stability in the closed-loop system, confining formation errors within a specified funnel while minimizing the cost of maintaining formation patterns between the leader and followers. Comparative simulations further confirm the advantages of the proposed method. Note to Practitioners—Cooperative control of multi-agent systems (MASs) has garnered considerable interest in diverse applications, including multi-motor synchronization and spacecraft formation flying. Managing complex engineering systems is challenging with a solitary component, prompting a focus on MASs. Formation control, a key aspect of MASs, has become a prominent research area. In practical applications, certain states are challenging to measure, and performance behaviors, such as convergence rate, maximum overshoot, and maximum steady-state error, often require predefined specifications. Simultaneously, achieving control tasks must be balanced with minimizing resource consumption. This study explores funnel-based output-feedback optimal leader-following control for MASs, emphasizing predefined-time convergence.}, archive = {J_TASE}, author = {Jiaxin Zhang and Yue Fu and Jun Fu}, doi = {10.1109/TASE.2024.3384472}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2794--2805}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Funnel-based adaptive predefined-time leader-following output-feedback optimal control for second-order nonlinear multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Noise optimization in artificial neural networks. TASE, 22, 2780--2793. (www)

  @article{ , abstract = {Artificial neural network (ANN) has been widely used in automation. However, the vulnerability of ANN under certain attacks poses a security threat to critical automation systems. Previous research has shown that adding noise to ANNs can enhance robustness. Nonetheless, striking a balance between robustness and task performance remains challenging, as excessive noise improves robustness but hampers performance, while low noise offers minor robustness improvement. In this work, we propose to learn the distribution of optimal injected noise, which improves the robustness as well as maintains the performance. Specifically, we compute the pathwise stochastic gradient estimate with respect to the standard deviation of the Gaussian noise added to each neuron of the ANN and optimize both the noise distribution and model parameters during training with negligible additional computational cost. In numerical experiments, our proposed method can achieve significant performance improvement on the robustness of several popular ANN structures under both black box and white box attacks. We also evaluate the proposed technique on two automation tasks: the classic reinforcement learning task of the cart pole game and a fault detection problem. Our results showed that the proposed technique outperforms a conventional neural network in terms of performance, robustness, and visual explainability. Note to Practitioners—The robustness of artificial neural networks is a critical consideration in automation applications as real-world data is often subject to unforeseen perturbations from the environment, potentially causing AI systems to behave unpredictably and unstably. For example, object detection is a widely employed AI technique in automation applications. However, current object detection systems are vulnerable to noise perturbation. Even small, imperceptible noise can lead the model to malfunction. Our work focuses on improving the robustness of neural networks. We propose a novel technique that can be added to any layer of existing neural networks to enhance robustness. Extensive experiments conducted in various scenarios have verified the effectiveness of the proposed method in enhancing both performance and robustness.}, archive = {J_TASE}, author = {Li Xiao and Zeliang Zhang and Kuihua Huang and Jinyang Jiang and Yijie Peng}, doi = {10.1109/TASE.2024.3384409}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2780--2793}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Noise optimization in artificial neural networks}, volume = {22}, year = {2025}, }
• (2025). Asymptotic leader-following consensus of heterogeneous multi-agent systems with unknown and time-varying control gains. TASE, 22, 2768--2779. (www)

  @article{ , abstract = {This paper investigates the consensus tracking control problem for a class of nonlinear multi-input multi-output (MIMO) heterogeneous multi-agent systems (HMASs), where the dimension of the dynamics of each agent is allowed to be different from each other. In addition, the control gain matrices (CGMs) with unknown time-varying coefficients and actuator faults with unpredictable jumps are also involved in the considered MIMO HMASs, all of which would cause damage for the system performance. First, a novel distributed auxiliary filter by feat of time-varying technology is introduced, which allows the zero-error estimation for the desired trajectory to be achieved and the asymptotic consensus tracking result to be further realized. Then, a cooperative adaptive control solution is proposed to ensure the asymptotical consensus tracking control result, in spite of the inherent unknown time-varying coefficients, unpredictable jumps caused by the unknown actuator faults, unknown disturbances and uncertain system parameters, distinguishing itself from those existing cooperative control works for HMASs where only ultimately uniformly bounded (UUB) result is derived. This is achieved mainly by the introduction of a series of Nussbaum functions and the employment of the adaptive estimation techniques. The effectiveness of the proposed control algorithm is confirmed by the simulation conducted on a group of HMASs involving unmanned aerial vehicles (UAVs) and autonomous surface vessels (ASVs). Note to Practitioners—In large-scale complex communication networks, uncertain HMASs with different structures and functions are capable of exchanging information and collaborating with each other to accomplish more complex and diverse tasks. Simultaneously, the probability of actuator faults within the HMASs increases dramatically, and the fault of a single agent may evolve into the failure of the whole system. Thus, the safety and reliability of HMASs are extremely important. Additionally, the control direction (the symbol of control gain) caused by both CGMs with coupling property of MIMO systems and actuator faults with unpredictable jumps, may not be guaranteed to be known in practical application, such as ship autopilot systems or uncalibrated visual servoing. This will have a considerable influence on the system’s control performance. On account of the threat of nonlinear uncertainties, and unknown control direction yield CGMs and actuator faults to HMASs, an adaptive fault-tolerant control solution based on an effective distributed time-varying auxiliary filter, is developed for MIMO HMASs to guarantee the asymptotical consensus tracking control result. Further, the proposed control scheme has been illustrated to be feasible through simulation experiment conducted on a group of HMASs consisting of UAVs and ASVs.}, archive = {J_TASE}, author = {Dahui Luo and Yujuan Wang and Zeqiang Li and Yongduan Song and Frank L. Lewis}, doi = {10.1109/TASE.2024.3384400}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2768--2779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asymptotic leader-following consensus of heterogeneous multi-agent systems with unknown and time-varying control gains}, volume = {22}, year = {2025}, }
• (2025). Discretized distributed optimization over dynamic digraphs. TASE, 22, 2758--2767. (www)

  @article{ , abstract = {We consider a discrete-time model of continuous-time distributed optimization over dynamic directed-graphs (digraphs) with applications to distributed learning. Our optimization algorithm works over general strongly connected dynamic networks under switching topologies, e.g., in mobile multi-agent systems and volatile networks due to link failures. Compared to many existing lines of work, there is no need for bi-stochastic weight designs on the links. The existing literature mostly needs the link weights to be stochastic using specific weight-design algorithms needed both at the initialization and at all times when the topology of the network changes. This paper eliminates the need for such algorithms and paves the way for distributed optimization over time-varying digraphs. We derive the bound on the gradient-tracking step-size and discrete time-step for convergence and prove dynamic stability using arguments from consensus algorithms, matrix perturbation theory, and Lyapunov theory. This work, particularly, is an improvement over existing stochastic-weight undirected networks in case of link removal or packet drops. This is because the existing literature may need to rerun time-consuming and computationally complex algorithms for stochastic design, while the proposed strategy works as long as the underlying network is weight-symmetric and balanced. The proposed optimization framework finds applications to distributed classification and learning. Note to Practitioners—Inspired by recent advances in cloud-computing and distributed and parallel processing along with embedded low-cost CPUs and wireless communications, this paper considers distributed algorithms for optimization and machine learning over wireless-connected autonomous multi-agent systems (MASs). In contrast to the classical centralized learning methods, which are prone to single-point-of-failure and centralized processing, in cooperative optimization the learning is distributed among a group of data-processing agents (e.g. robots) with communication units. This article provides an efficient algorithm to enable MASs to collaboratively optimize a cost function, e.g., for binary classification and distributed support vector machine (D-SVM). Sampled data systems related to MASs and robotic networks, due to digital communications and discretized control models, use discrete-time algorithms that need to account for intermittent communications and dynamic networking. This requires discretized algorithms over dynamic digraphs, practical in the presence of packet drops (or lost communication channels). Most existing distributed algorithms either are susceptible to change in the communication network or impose computationally inefficient stochastic design. These algorithms need to be rerun, for example, whenever the mobile robotic network changes due to limited communication range. This makes most existing algorithms infeasible in real-time applications. Our proposed method outperforms similar algorithms over dynamic robotic networks and in the presence of link removal (packet drops). We show this efficiency of our distributed optimization method by simulation.}, archive = {J_TASE}, author = {Mohammadreza Doostmohammadian and Wei Jiang and Muwahida Liaquat and Alireza Aghasi and Houman Zarrabi}, doi = {10.1109/TASE.2024.3383894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2758--2767}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discretized distributed optimization over dynamic digraphs}, volume = {22}, year = {2025}, }
• (2025). Prediction of condition monitoring signals using scalable pairwise gaussian processes and bayesian model averaging. TASE, 22, 2746--2757. (www)

  @article{ , abstract = {Predicting condition monitoring signals has become a critical task for health status assessment and monitoring of industrial systems. It is crucial to incorporate correlated historical data when making predictions for the target signal. As a flexible nonparametric approach, the multi-output Gaussian process (MOGP) model can be employed for this nonlinear regression problem. One effective way to construct MOGP is through convolving a latent function drawn from a Gaussian process. While leveraging the convolution process makes MOGP expressive, there are several challenges that remain to be addressed. First, the scalability of MOGP is always an important issue since the computational demands would increase drastically as the dimension of output variables grows. Besides, the negative transfer should be mitigated when the target variable and the source variable share little commonality. In this study, a pairwise structure is adopted by decomposing the full multivariate model into a group of bi-output models. Furthermore, a Bayesian model averaging approach is utilized to combine the prediction results of the bi-output models. A model selection scheme based on Bayes factor is employed to alleviate negative transfer and facilitate model scalability further by discarding the weakly correlated outputs. The key advantage of the proposed model lies in the improved prediction and uncertainty quantification performance compared with traditional MOGP models. The superiority of the proposed method is validated by numerical studies and a case study. Note to Practitioners—This study addresses the challenge of predicting condition monitoring signals in a high-dimensional setting. Existing nonparametric approaches suffer from high computation complexity or ineffective information integration from different signals. We propose a novel approach using a scalable pairwise MOGP model based on Bayesian model averaging. Our method decomposes the full model into bi-output submodels and averages them in a Bayesian way. We also employ a model selection scheme based on the Bayes factor to alleviate negative transfer by discarding weakly correlated outputs. Numerical experiments suggest that this approach can improve prediction accuracy and uncertainty quantification performance for the prediction of condition monitoring signals. Our approach offers a promising solution for condition monitoring signal prediction in automatic and industrial systems where condition monitoring data are readily available.}, archive = {J_TASE}, author = {Jinwen Sun and Shiyu Zhou and Dharmaraj Veeramani and Kaibo Liu}, doi = {10.1109/TASE.2024.3383811}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2746--2757}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prediction of condition monitoring signals using scalable pairwise gaussian processes and bayesian model averaging}, volume = {22}, year = {2025}, }
• (2025). Observer-based adaptive consensus for lipschitz nonlinear multiagent systems via asynchronous event-triggered strategy. TASE, 22, 2734--2745. (www)

  @article{ , abstract = {This article investigates the event-triggered (ET) consensus issue of Lipschitz nonlinear multiagent systems (MASs) with a positive minimum inter-event time (MIET) in the presence of unmeasurable state. To handle this issue, we design a fully distributed control scheme by virtue of asynchronous event-triggered mechanism (ETM) and available local measurement output, in which the knowledge of global information is no longer required by introducing adaptive parameters. It is shown that under the proposed control scheme, the consensus is asymptotically realized, in which the local state observer is proposed to remove the requirement of full-state information. Besides, a dynamic variable is introduced to obtain a positive MIET in the proposed ETM, where each agent asynchronously sends its observer value to its neighbors based on intermittent communication. In light of consensus analysis and MASs dynamics, a novel estimator of the observer is designed to ensure the asymptotic consensus. Note that based on the reformulated Lipschitz property, the linear matrix inequality (LMI) conditions of the control design are less restrictive as compared with the exist results. A practice simulation example is presented to verify the theoretical results. Note to Practitioners—Some practical (such as manipulator and pendulum) control systems can be modeled as Lipschitz nonlinear MASs. If the information exchange determined by ETM is high-frequency, it is unsuitable for physical implementation because each agent has limited communication resource. Thus, it is significant to design ETM with a positive MIET guarantees. Besides, the asynchronous communication among neighbors is more distributed compared with the synchronous communication, which might significantly save the communication resource. The reformulated Lipschitz property is introduced to reduce conservatism of the LMI conditions in the existing results. It might be difficult to measure the full-state information of each agent in some situations owing to physical limitations. Therefore, an observer-based adaptive asynchronous ET control scheme is proposed for Lipschitz nonlinear MASs with unmeasurable state, which can guarantee asymptotic convergence and a positive MIET. Finally, a one-link manipulator with revolute joints is considered to demonstrate the effectiveness of the proposed algorithm.}, archive = {J_TASE}, author = {Wang Yang and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3383726}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2734--2745}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based adaptive consensus for lipschitz nonlinear multiagent systems via asynchronous event-triggered strategy}, volume = {22}, year = {2025}, }
• (2025). SPDP-net: A semantic prior guided defect perception network for automated aero-engine blades surface visual inspection. TASE, 22, 2724--2733. (www)

  @article{ , abstract = {Automated surface defect detection is essential to manufacturing automation. However, automated inspection of aero-engine blades remains challenging due to tiny defects and weak features. To address this issue, we propose a semantic prior guided defect perception network, named SPDP-Net, which is ultimately integrated into an automated system to achieve efficient detection of defects. Firstly, a semantic prior mining module (SPM) is developed to capture finer-grained pixel-level location priors of defects by leveraging image feature mapping relations which facilitates the precise perception of tiny defects. Subsequently, we propose a defect enhancement perception module (DEP) to separate weak defects from complex backgrounds by utilizing defect location priors provided by SPM to enhance the features of defects while suppressing the values of non-defect regions, which makes the weak defects present as more obvious outliers. Finally, the global information extraction module (GIE) extracts the global features of defects, which helps to further improve the predicted results. When equipped with SPM, DEP and GIE, SPDP-Net can accurately identify and locate defects, exhibiting more competitive recognition and feature extraction capabilities for tiny defects and weak defects. To evaluate the effectiveness of our method, we construct an aero-engine blade surface defect detection dataset from real industrial scenarios called ABSDD with the collaboration of senior engineers. We achieve 95.9% precision, 94.0% recall and 94.9% F1 score on ABSDD. In addition, we also achieve state-of-the-art results on two public benchmark datasets, KSDD2 and DAGM. Finally, we have applied the developed SPDP-Net to an automated system and have conducted actual tests in collaboration with a well-known aero-engine production company. Note to Practitioners—At present, the automated detection system for surface defects in industrial manufacturing has not been well developed, which is particularly trailing behind in the field of aero-engine manufacturing. To the best of our knowledge, the surface defect detection of aero-engine blades is still carried out manually. To address this problem, we propose SPDP-Net for the automated detection of surface defects in aero-engine blades. It can accurately perceive and capture tiny and weak defects with excellent performance. SPDP-Net achieves state-of-the-art results on three tasks from different industrial manufacturing fields, which demonstrates its good transfer application capabilities. In addition, we integrate SPDP-Net into an automated detection system consisting of an autonomous imaging system and a high-performance computing server and conduct actual tests in an aero-engine production company. The test achieves remarkable results, indicating the good application prospects of the automated detection system.}, archive = {J_TASE}, author = {Yunfeng Ma and Min Liu and Yiqiong Zhang and Xueping Wang and Yaonan Wang}, doi = {10.1109/TASE.2024.3383451}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2724--2733}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {SPDP-net: A semantic prior guided defect perception network for automated aero-engine blades surface visual inspection}, volume = {22}, year = {2025}, }
• (2025). Adaptive event-triggered output-feedback stabilization with exponential convergence. TASE, 22, 2712--2723. (www)

  @article{ , abstract = {This paper seeks adaptive event-triggered output-feedback control which enables exponential stabilization for nonlinear systems with unknown growth rate. Convergence rate, as an important performance specification, is usually hard to acquire in the context of adaptive control. Besides, the event-triggered architecture could undermine convergence rate, entailing a competent compensation mechanism under reduced execution. As such, we are compelled to pursue a distinctive adaptive event-triggered output-feedback scheme. Specifically, a delicate dynamic gain incorporating exponential-type time-varying information is introduced, which would not only counteract the unknown growth rate, but particularly enable desired convergence. Correspondingly, a compatible event-triggering mechanism, capable of ensuring timely execution for adaptive compensation, is designed by suitably exploiting the gain information. In this way, an adaptive event-triggered output-feedback controller is constructed, which can render exponential convergence for system states, alongside an explicit pre-estimate for inter-execution intervals. Note to Practitioners—In the networked setup, the communication resources are usually shared but limited, entailing efficient resource utilization. Notably, event-triggered control could cater for the efficiency demand, which requires information transmission and control updating only when necessary for systems, unlike sampled-data control with a conservative prescribed rate of sampling/execution. This paper is devoted to developing a distinctive adaptive output-feedback scheme of event-triggered stabilization under the architecture with controller-to-actuator communication reduced. The scheme can not only cope with system nonlinearities with unknown growth rate (resulting from e.g., slowly-but-widely varying parameters and multiplicative disturbances) but particularly enable exponential convergence rate, an objective which is impossible for the existing schemes. Besides, the scheme allows flexible parameter choice, with which the change in the number of executions is illustrated through simulation.}, archive = {J_TASE}, author = {Yaxin Huang and Yungang Liu}, doi = {10.1109/TASE.2024.3383415}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2712--2723}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive event-triggered output-feedback stabilization with exponential convergence}, volume = {22}, year = {2025}, }
• (2025). Multiscale global prompt transformer for EEG-based driver fatigue recognition. TASE, 22, 2700--2711. (www)

  @article{ , abstract = {Driver fatigue is a critical factor that lead to traffic accidents with a high fatality rate. Electroencephalogram (EEG) is one of the most reliable indicators to objectively assess fatigue status, but recognizing fatigue driving status from it is still an essential and challenging problem. In this paper, we propose a multiscale global prompt Transformer (MsGPT) deep learning model, which can automatically recognize driver fatigue end-to-end. First, we construct an intra-inter-scale cascade framework based on Transformer with a multiscale convolutional patch embedding (MC-PatchEmbed), and guide global-local feature interaction by adding a global prompt token throughout. Second, to efficiently integrate intra-scale and inter-scale feature information, we design a mixed token by aggregating the output from the intra-scale, which includes rich low-level feature information for multiscale. Moreover, a novel learnable query is introduced into multi-head self-attention (MSA) to reduce the computational complexity to linear level. Experiments are conducted on the SEED-VIG dataset and the SADT dataset with both intra-subject and inter-subject settings to evaluate the performance of MsGPT, and the results show that MsGPT greatly outperforms various methods in terms of the classification evaluation metrics of EEG-based fatigue driving. Note to Practitioners—This paper considers the use of raw EEG data to recognize the driver fatigue state. Existing methods mainly rely on manually extracted EEG features and convolutional neural network (CNN) based inference. However, the large intra-individual and inter-individual differences greatly limit the extraction of EEG fatigue features. This paper suggests a multiscale global prompt Transformer (MsGPT) deep learning model. This model leverages a shared weighting mechanism to construct an inter- to intra-scale multiscale framework that can capture refined fatigue features not achievable at a single scale, we incorporate a new Transformer of the global prompt mechanism, which facilitates multiscale local-to-global fusion of long-term physiological signals. Our experiments demonstrate the superiority of our method on two datasets with intra-subject and inter-subject settings. The proposed method can be readily deployed in the automatic driving assistance system to alert drivers to avoid or reduce traffic accidents caused by excessive fatigue.}, archive = {J_TASE}, author = {Pengbo Zhao and Cheng Lian and Bingrong Xu and Zhigang Zeng}, doi = {10.1109/TASE.2024.3382892}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2700--2711}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multiscale global prompt transformer for EEG-based driver fatigue recognition}, volume = {22}, year = {2025}, }
• (2025). UL-SLAM: A universal monocular line-based SLAM via unifying structural and non-structural constraints. TASE, 22, 2682--2699. (www)

  @article{ , abstract = {Leveraging structural line features to complement sparse point features has been studied in recent years. However, this approach relies on a Manhattan world assumption and does not incorporate non-structural lines due to the triangulation degeneracy problem and tracking process instability. To address these problems, we propose a general line-based SLAM system that combines points, structural and non-structural lines. First, an efficient line matching algorithm for multi-scale is designed to obtain more accurate matching pairs through a divide-and-conquer approach. In addition, a novel line triangulation strategy utilizing spatial-temporal consistency and degeneracy identification is proposed to improve the quality of line generation in a sliding window. Finally, universal structural constraints based on measurement of the vanishing directions are implemented to complement the information missing from the Plücker line projection in local mapping optimization. Extensive experiments are conducted on the public EuRoC and TUM datasets as well as a self-collected dataset, and the results show that UL-SLAM achieves cutting-edge performance among recent state-of-the-art methods in both accuracy and speed. Ablation experiments also demonstrate that the integration of different line features can improve the robustness and accuracy of a visual SLAM system in challenging scenarios with low texture and weak illumination. Our implementation of the UL-SLAM will be open-sourced to benefit the community (https://github.com/jhch1995/UL-SLAM). Note to Practitioners—This article was motivated by the challenges of visual localization problems in human-made indoor scenes. Visual localization has been widely used in various robotic fields such as self-driving vehicles, augmented reality (AR), and virtual reality (VR). In real-world scenes, the localization accuracy will be significantly decreased because of the sparse and uncertain visual features in low-texture or weak illumination environments, which reduces the robustness of robot tracking. To address this problem, this article proposes a novel universal line-based SLAM system (UL-SLAM) that unifies structural and non-structural constraints within a general framework unrestricted by the strong global Manhattan world assumption. UL-SLAM can not only improve the accuracy of pose estimation due to the proposed methods for the line features but also achieve real-time performance. In addition, for 3D mapping construction, UL-SLAM can also enrich the geometric structure information of the indoor scenes. Extensive experiments are conducted on various indoor datasets for autonomous robots, and the results demonstrate the efficiency, accuracy, and robustness of the proposed system in different complex scenarios.}, archive = {J_TASE}, author = {Haochen Jiang and Rui Qian and Liang Du and Jian Pu and Jianfeng Feng}, doi = {10.1109/TASE.2024.3382770}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2682--2699}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {UL-SLAM: A universal monocular line-based SLAM via unifying structural and non-structural constraints}, volume = {22}, year = {2025}, }
• (2025). Co-design of improved fuzzy switching-type state-FDI estimation and resilient control for DC microgrids under malicious attacks. TASE, 22, 2671--2681. (www)

  @article{ , abstract = {DC microgrids (DC-MGs) consist of multiple units interconnected in a network. This makes the system vulnerable to malicious cyber attacks. Malicious attacks here refer to hybrid attacks of false data injection (FDI) attacks and denial of service (DoS) attacks. The main focus of this paper is to study the state-FDI estimation and resilient control problems of nonlinear DC-MGs under malicious attacks. Since the fuzzy controller/observer based on the parallel distributed compensation method cannot obtain more relaxed satisfactory results. Methods presented in the literature will be reviewed and a method based on multi-mode fuzzy switching mechanism (MMFSM) will be developed to relax the stability analysis of these DC-MGs. Furthermore, a co-design method based on improved fuzzy switching technology is proposed. This method adopts a novel MMFSM, which makes the constraints more relaxed, the state estimation more accurate, and the conservatism of the co-design method is reduced. Finally, the effectiveness of the proposed improved co-design method is verified by comparison and analysis of simulation examples. Note to Practitioners—The motivation of this paper comes from the problem that DC-MG systems, which can realize large-scale application of new energy technology, is vulnerable to malicious cyber attacks in practical tasks. Malicious attacks (such as FDI attacks and DoS attacks) will bring security risks such as data tampering and function destruction to the system. To this end, this paper develops a MMFSM-based co-design method for DC-MGs under malicious attacks. The improved co-design method can deal with sector nonlinearity, state imperfectly measurable, FDI attacks and DoS attacks simultaneously. For this method, the main difficulty in this paper is how to reduce its conservatism. To solve this problem, we adopt a novel MMFSM. By introducing time-varying equilibrium matrix synchronously with the system dynamics of different modes, the improved co-design method can switch accordingly, thereby making the constraints more relaxed and the state estimation more accurate.}, archive = {J_TASE}, author = {Fuyi Yang and Xiangpeng Xie and Engang Tian and Jiayue Sun}, doi = {10.1109/TASE.2024.3382755}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2671--2681}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Co-design of improved fuzzy switching-type state-FDI estimation and resilient control for DC microgrids under malicious attacks}, volume = {22}, year = {2025}, }
• (2025). Advances and challenges in computer vision for image-based plant disease detection: A comprehensive survey of machine and deep learning approaches. TASE, 22, 2639--2670. (www)

  @article{ , abstract = {As advancements in agricultural technology unfold, machine learning and deep learning approaches are gaining interest in robust plant disease identification. Early disease detection, integral to agricultural productivity, has propelled innovations across all phases of detection. This survey paper provides a meticulous examination of plant disease detection systems, elucidating data collection methodologies and underscoring the pivotal role of datasets in model training. The narrative navigates through the complex areas of data and image processing techniques, segueing into an exploration of various segmentation methods. The survey emphasizes the importance of feature extraction and selection techniques, illustrating their efficacy in increasing classification accuracy. It examines the classification process, embracing both traditional machine learning and avant-garde deep learning methods, with a particular spotlight on Convolutional Neural Networks (CNNs). The study examines over one hundred seminal papers, anatomizing their dataset utilizations, feature considerations, and classification strategies. Overall, the paper contemplates the challenges permeating this vibrant field, addressing critical issues such as dataset diversity, model generalization, and real-world applicability. Note to Practitioners—To ensure crop health and yield, timely and precise plant disease detection is crucial. Our research, titled “Advances And Challenges in Plant Disease Detection: A Comprehensive Survey of Machine and Deep Learning Approaches,” examines the critical role of datasets, advanced image processing, and segmentation techniques in disease detection. This paper presents practitioners with a guide to the latest techniques for enhanced disease detection by emphasizing the significance of feature extraction and highlighting the capabilities of convolutional neural networks (CNNs). By understanding the highlighted challenges, such as dataset diversity and model generalization, industry professionals can better equip themselves to integrate these technological advancements into real-world agricultural applications.}, archive = {J_TASE}, author = {Syed Asif Ahmad Qadri and Nen-Fu Huang and Taiba Majid Wani and Showkat Ahmad Bhat}, doi = {10.1109/TASE.2024.3382731}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2639--2670}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Advances and challenges in computer vision for image-based plant disease detection: A comprehensive survey of machine and deep learning approaches}, volume = {22}, year = {2025}, }
• (2025). Informative trajectory planning for air-ground cooperative monitoring of spatiotemporal fields. TASE, 22, 2627--2638. (www)

  @article{ , abstract = {This paper investigates an air-ground cooperative monitoring problem for spatiotemporal fields, such as air pollution, forest fires, oil spills, etc, with unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs). To fully exploit complementarities of these heterogeneous vehicles and improve efficiency of the cooperative monitoring, we design a novel cooperation scheme: each UAV is assigned to loiter over and transmit its observations to a pre-allocated UGV, and the UGV provides guidance on informative trajectories for the UAV and aims to reach a target position as fast as possible. Such a scheme brings challenges to informative trajectory planning of the UGVs, lying in the delayed observations from the UAVs and the cumulative information constraint depending on the unknown field. To overcome them, this work proposes a model-free reinforcement learning (RL)-based trajectory planning method to learn continuous policies for the UGVs, where a field estimator is designed for each UGV to recover observability of the field. In addition, we derive model predictive control (MPC)-based trajectory planners for the UAVs with tailored reference positions, where the uncertain tracking errors can be handled by the RL-based method of the UGVs. Thus, a performance coupling problem of the heterogeneous vehicles is tackled. Simulations illustrate the effectiveness of the proposed trajectory planning methods and the efficiency of the air-ground cooperative monitoring scheme. Note to Practitioners—This article is motivated by cooperative monitoring tasks with UAVs and UGVs in practical applications, such as environmental monitoring, search and rescue after disasters, etc. Due to the complex dynamics of spatiotemporal fields in these tasks, trajectory planning for the cooperative monitoring system is challenging and requires much computations. To resolve the issues, we propose a novel cooperation scheme in this article, where the large computational capability of the UGVs is utilized to solve a minimum-time trajectory planning problem under a cumulative information constraint, and the UAVs only loiter over and transmit measurements about the field to the UGVs. To achieve this scheme, RL-based and MPC-based trajectory planning methods are proposed for the UGVs and the UAVs, respectively. Simulations have validated the effectiveness of the proposed trajectory planning methods and good performance of the cooperative monitoring system.}, archive = {J_TASE}, author = {Zhuo Li and Yunlong Guo and Gang Wang and Jian Sun and Keyou You}, doi = {10.1109/TASE.2024.3382730}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2627--2638}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Informative trajectory planning for air-ground cooperative monitoring of spatiotemporal fields}, volume = {22}, year = {2025}, }
• (2025). LLM-enabled incremental learning framework for hand exoskeleton control. TASE, 22, 2617--2626. (www)

  @article{ , abstract = {It remains a formidable challenge to accurately recognize motion intentions of patients thus to control hand exoskeletons according to their volition. Current methods primarily focus on recognition of limited patient’s motion intentions, with the purpose of controlling preconfigured gestures of a hand exoskeleton for grasping objects. These methods exhibit a marked shortfall when encountering scenarios that are unexpected or not designed in advance, such as non-preprogrammed hand movements and object manipulation tasks. To tackle this issue, large language model (LLM) and speech recognition technology are employed in this study to allow the patient to control a hand exoskeleton at will. In particular, two LLMs are tailored to formulate codes of either generating non-preprogrammed gestures or dealing with unencountered objects. Additionally, an incremental learning framework is proposed to enable patients to perform both predefined and non-predefined operation tasks by integrating a natural language parser with the two LLM-based learners. The natural language parser can directly control the hand exoskeleton to perform predefined operations tasks from prestored command set, while the LLM-based learners can incrementally expand the control command set so as to enhance adaptability of the hand exoskeleton to complex activities over daily use. This study is a pioneering work in the field of hand exoskeletons, which will revolutionize the way to control hand exoskeletons. Furthermore, the proposed framework can be easily generalized to any other robots by modifying the prompt of customized LLMs, which provides a new idea to achieve autonomous learning in robotics. Note to Practitioners—The motivation of this article is to tackle the challenge of intention recognition for performing activities of daily living (ADLs) by stroke patients using a multi-degree of freedom hand exoskeleton. Existing methods for intention recognition so far can only be used for several tasks that are predefined in advance, thus none of them allow patients to control the hand exoskeleton completely at will. To surpass this limitation, an LLM-enabled incremental learning framework that integrates a hand exoskeleton controller with Large Language Model (LLM) is proposed and validated in this study. The framework offers patients an intuitive interface via voice interaction and enables patients to perform not only predefined operation tasks by the hand exoskeleton controller but also non-predefined ones that can be learned from the LLM. As a result, the hand exoskeleton controller continues to learn from the LLM, therefore is gradually able to perform all tasks in daily life. This pioneering study paves a new way in building patient-controlled hand exoskeletons with autonomous intelligence that can deal with non-predefined operation tasks in unstructured environments.}, archive = {J_TASE}, author = {Wenyuan Chen and Guangyong Li and Mingwei Li and Wenxue Wang and Peng Li and Xiujuan Xue and Xingang Zhao and Lianqing Liu}, doi = {10.1109/TASE.2024.3382679}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2617--2626}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {LLM-enabled incremental learning framework for hand exoskeleton control}, volume = {22}, year = {2025}, }
• (2025). Anti-unmanned directional drive-off method under integrated navigation mode. TASE, 22, 2607--2616. (www)

  @article{ , abstract = {To realize the defense task of precise directional drive off of unmanned combat system in the target area, an anti-unmanned directional drive-off method under SINS(Strapdown Inertial Navigation System)/GPS(Global Positioning System) integrated navigation mode is proposed in this paper, which is suitable for the security defense of important places, urban public security protection and anti-terrorism and stability maintenance in sensitive areas. Firstly, this paper constructs a seven dimensional SINS/GPS integrated navigation model of unmanned combat system including horizontal position, velocity and attitude, and uses discrete Kalman filter for data fusion; Then, combined with the coupling relationship between various state parameters of SINS/GPS integrated navigation, the controllability and stability of false satellite signal acting on the position output of integrated navigation are analyzed, and then the feasibility of satellite navigation spoofing under SINS/GPS integrated navigation mode is deduced; Finally, the false satellite signal is designed and constructed to make the position estimation of SINS/GPS Integrated Navigation drive away accurately. The simulation and experimental results verify the effectiveness and correctness of the anti-unmanned directional drive-off method under the integrated navigation mode. Note to Practitioners—This work is motivated by the need for the defense task of precise directional drive off of unmanned combat system in the target area. Most unmanned combat platforms adopt the integrated navigation mode with satellite navigation system as an auxiliary means. It can be challenging for researchers and practitioners to drive off the unmanned combat system which adopt the integrated navigation mode in the target area by the existing spoofing methods. Therefore, this paper presents an accurate directional displacement method against unmanned under the SINS/GPS integrated navigation mode and the effectiveness of the proposed method is verified by UAV spoofing simulation and test. This paper provides researchers and practitioners with comprehensive reference sources in anti-Unmanned directional drive-off method under integrated navigation mode, which can help them develop suitable solutions to drive off the unmanned combat system which adopt the integrated navigation mode in the target area.}, archive = {J_TASE}, author = {Yan Guo and Juliang Cao and Kanghua Tang and Kaixin Luo and Xingshou Geng}, doi = {10.1109/TASE.2024.3382360}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2607--2616}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Anti-unmanned directional drive-off method under integrated navigation mode}, volume = {22}, year = {2025}, }
• (2025). Safety-certified self-triggered cooperative path following control via data-driven learning and neurodynamic optimization. TASE, 22, 2595--2606. (www)

  @article{ , abstract = {This paper investigates the safety-certified cooperative path following problem for second-order nonlinear systems with multiple-input multiple-output strict-feedback form subject to safety, state, and input constraints. A safety-certified learning control approach is proposed to achieve collision-free cooperative path following based on command optimization, online learning, and self-triggered communication. Specifically, an extended-state-observer-aided learning neural predictor is developed to simultaneously identify nonlinear functions and unknown input gains without measuring state derivatives. Control barrier functions are then employed to ensure safety through forward invariant sets. Next, command optimization is utilized to generate the optimal virtual control signals that satisfy safety constraints, state constraints, and input constraints. A neurodynamic optimization technique is employed to solve the quadratic optimization problem in real time. Additionally, a self-triggered mechanism is introduced in path variable coordination to reduce the listening and triggering times. By using the proposed safety-certified cooperative path following approach, a safe formation is guaranteed for input-to-state safety. Simulation results are provided to illustrate the effectiveness of the proposed method. Note to Practitioners—This paper addresses the safety-certified cooperative path following problem of multi-agent systems, which has practical implications in various applications. These applications include formation patrol, cargo transportation, search and rescue missions, swarm robotics, and agricultural tasks. By coordinating the movements of multiple agents, cooperative path following enhances efficiency in these scenarios. The challenges posed by safety, state, and input constraints are commonly encountered in practical. To address these challenges, the command optimization approach is proposed in this paper. The optimization problem is efficiently solved in real-time using neurodynamic optimization technique. Furthermore, to enhance feasibility in a limited communication environment, a self-triggered mechanism is introduced to reduce the communication burden. Therefore, the aforementioned effective scheme is suitable for implementation in industrial applications.}, archive = {J_TASE}, author = {Siming Cong and Dan Wang and Nan Gu and Tieshan Li and Zhouhua Peng}, doi = {10.1109/TASE.2024.3382323}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2595--2606}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safety-certified self-triggered cooperative path following control via data-driven learning and neurodynamic optimization}, volume = {22}, year = {2025}, }
• (2025). Motion planning for aerial pick-and-place with geometric feasibility constraints. TASE, 22, 2577--2594. (www)

  @article{ , abstract = {This paper studies the motion planning problem of the pick-and-place of an aerial manipulator that consists of a quadcopter flying base and a Delta arm. We propose a novel partially decoupled motion planning framework to solve this problem. Compared to the state-of-the-art approaches, the proposed one has two novel features. First, it does not suffer from increased computation in high-dimensional configuration spaces. That is because it calculates the trajectories of the quadcopter base and the end-effector separately in Cartesian space based on proposed geometric feasibility constraints. The geometric feasibility constraints can ensure the resulting trajectories satisfy the aerial manipulator’s geometry. Second, collision avoidance for the Delta arm is achieved through an iterative approach based on a pinhole mapping method, so that the feasible trajectory can be found in an efficient manner. The proposed approach is verified by five experiments on a real aerial manipulation platform. The experimental results show the effectiveness of the proposed method for the aerial pick-and-place task. Note to Practitioners—Aerial manipulators have attracted increasing research interest in recent years due to their potential applications in various domains. In this paper, we particularly focus on the motion planning problem of the pick-and-place of aerial manipulators. We propose a novel partially decoupled motion planning framework, which calculates the trajectories of the quadcopter base and the end-effector in Cartesian space, respectively. Geometric feasibility constraints are proposed to coordinate the trajectories to ensure successful execution. Five experiments on a real aerial manipulator platform demonstrate the effectiveness of the approach. In future research, we will address the motion planning problem of aerial manipulators in complex environments.}, archive = {J_TASE}, author = {Huazi Cao and Jiahao Shen and Cunjia Liu and Bo Zhu and Shiyu Zhao}, doi = {10.1109/TASE.2024.3382296}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2577--2594}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Motion planning for aerial pick-and-place with geometric feasibility constraints}, volume = {22}, year = {2025}, }
• (2025). Development and quantitative evaluation of a novel autonomous in situ bioprinting surgical robotic framework for treatment of volumetric muscle loss injuries. TASE, 22, 2565--2576. (www)

  @article{ , abstract = {In situ bioprinting has been identified as a promising tissue engineering technique for treating volumetric muscle loss (VML) injuries. However, the success of this procedure significantly depends on the uniform and precise deposition of cells contributing to the regeneration of muscles. To address this critical need, in this work, we present design and quantitative evaluation of a novel autonomous in situ bioprinting surgical robotic framework that can be used with a generic bioprinting material. The proposed framework consists of three main components: (i) a bioprinting tool integrated with a seven-degree-of-freedom robotic manipulator to perform a precise autonomous bioprinting procedure; (ii) a unique 3D visual measurement framework comprised of a high-accuracy structured light camera with complementary 2D/3D computer vision algorithms- to enable online and accurate measurement and reconstruction of the bioprinted constructs; and (iii) a quantitative evaluation module with novel assessment metrics- to characterize and evaluate the performance of the bioprinting process toward finding optimal bioprinting parameters. To ensure the biological functionality of a printed construct using our robotic system, we performed 90 experiments and identified optimal bioprinting parameters using the proposed novel assessment metrics. Note to Practitioners—This paper was motivated by the problem of volumetric muscle loss treatment using an in situ bioprinting procedure but it also can be applied for treatment of skin and cartilage injuries. Existing approaches to perform in situ bioprinting is limited to either manual handheld bioprinting devices– that suffer from poor manual control and inaccurate printing constructs– or robotic systems– that have been developed without (i) considering a realistic surgical workflow and (ii) quantitatively evaluating the quality of printed constructs. To collectively address these issues, in this paper, we propose a novel autonomous in situ robotic bioprinting framework. We also introduce unique and complementary quantitative assessment metrics to characterize and evaluate the performance of the bioprinting process. Experiments suggest that the proposed framework can robustly identify optimal bioprinting parameters to ensure the biological functionality of a printed construct using our robotic system.}, archive = {J_TASE}, author = {Shuojue Yang and Hansoul Kim and Omid Rezayof and Jeff Bonyun and Johnson V. John and Mehmet R. Dokmeci and Ali Khademhosseini and Farshid Alambeigi}, doi = {10.1109/TASE.2024.3382200}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2565--2576}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development and quantitative evaluation of a novel autonomous in situ bioprinting surgical robotic framework for treatment of volumetric muscle loss injuries}, volume = {22}, year = {2025}, }
• (2025). Mode substitution and constraint implementation in complex dynamic process regulation: A solution for performance self-recovery. TASE, 22, 2552--2564. (www)

  @article{ , abstract = {With the increase of wastewater treatment volume and the insufficiency of preventive maintenance measures, the potential safety hazards of wastewater treatment process (WWTP) are gradually emerging. The unplanned sensor failures may result in false information feedback and reduce the reliability of the control system. This paper attempts to formulate a working mode substitution-based performance self-recovery control (WMS-based PSRC) strategy for the WWTP against sensor failures. Therein, a substitution indication-based switching function scheme is developed to automatically activate different sensor working modes. The transient and steady-state process regulation performance after one working mode substitution are guaranteed by establishing the prescribed-time performance function and error transformation dynamics. Then, an adaptive performance self-recovery control is developed to ensure the desired regulation level of the WWTP. Extensive studies on a recognized WWTP platform illustrate that the proposed control scheme can guarantee the working mode substitution and desired regulation performance while mitigating the failure detriment. Note to Practitioners—The vulnerability of industrial control systems in terms of policy, architecture and platform, as well as the hysteresis and subjectivity of manual operations on sensor maintenance, will result in unstable process performance, loss of critical control data, unnecessary workload and economic losses. In this paper, a WMS-based PSRC is presented for the WWTP with sensor failures. Three main aspects are contained: mode substitution, performance guarantee and WMS-based PSRC structure. The total sensor working modes are divided, meanwhile the switching function index is updated or maintained based on the mode substitution indication. After reconstructing performance function, the guaranteed performance technology is developed and applied to improve response rate and regulation accuracy. Then, a WMS-based PSRC structure is designed to comprehensively analyse the monitoring and control process. Finally, the industrial application results of WWTP demonstrate that the WMS-based PSRC can optimize process operation. This strategy is, therefore, useful for practitioners to achieve the fast performance self-recovery after the abnormal conditions.}, archive = {J_TASE}, author = {Peihao Du and Weimin Zhong and Xin Peng and Linlin Li}, doi = {10.1109/TASE.2024.3382113}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2552--2564}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mode substitution and constraint implementation in complex dynamic process regulation: A solution for performance self-recovery}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy resilient fixed-time bipartite consensus tracking control for nonlinear MASs under sensor deception attacks. TASE, 22, 2542--2551. (www)

  @article{ , abstract = {This paper studies the adaptive fuzzy resilient fixed-time bipartite consensus tracking control problem for a class of nonlinear multi-agent systems (MASs) under sensor deception attacks. Firstly, in order to reduce the impact of unknown sensor deception attacks on the nonlinear MASs, a novel coordinate transformation technique is proposed, which is composed of the states after being attacked. Then, in the case of unbalanced directed topological graph, a partition algorithm (PA) is utilized to implement the bipartite consensus tracking control, which is more widely applicable than the previous control strategies that only apply to balanced directed topological graph. Moreover, the fixed-time control strategy is extended to nonlinear MASs under sensor deception attacks, and the singularity problem that exists in fixed-time control is successfully avoided by employing a novel switching function. The developed distributed adaptive resilient fixed-time control strategy ensures that all the signals in the closed-loop system are bounded and the bipartite consensus tracking control is achieved in fixed time. Finally, the designed control strategy’s validity is demonstrated by means of a simulation experiment. Note to Practitioners—Currently, there are many practical application scenarios for nonlinear MASs, such as intelligent transportation, unmanned aerial vehicle cluster formation, etc. This paper investigates the adaptive fuzzy resilient fixed-time bipartite consensus tracking control problem for a class of nonlinear MASs under sensor deception attacks. In practice, these two situations are common: 1) The topology graph describing the communication relationship of nonlinear MASs is unbalanced 2) The nonlinear MASs is subjected to external malicious cyber attacks. Therefore, a novel coordinate transformation technique is proposed to reduce the impact of sensor deception attacks on the nonlinear MASs, and a partition algorithm is employed to implement bipartite consensus tracking control based on the unbalanced communication topology graph. At the same time, the nonsingular fixed-time control strategy can significantly improve the convergence of the studied nonlinear MASs. Furthermore, the nonlinear nonstrict-feedback model and backstepping design method used in this paper are general and practical.}, archive = {J_TASE}, author = {Ben Niu and Zihao Shang and Guangju Zhang and Wendi Chen and Huanqing Wang and Xudong Zhao and Ding Wang}, doi = {10.1109/TASE.2024.3381568}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2542--2551}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy resilient fixed-time bipartite consensus tracking control for nonlinear MASs under sensor deception attacks}, volume = {22}, year = {2025}, }
• (2025). Large-scale continuous-time crude oil scheduling: A variable-length evolutionary optimization approach. TASE, 22, 2526--2541. (www)

  @article{ , abstract = {Evolutionary algorithms (EAs) have significantly contributed to addressing large-scale crude oil scheduling problems (COSPs) that exceed the capabilities of mathematical programming. However, current research using EAs to resolve COSPs is limited to discrete-time models that assume uniform operation durations. While this simplification transforms COSPs into fixed-length optimization problems suitable for conventional EAs, it fails to adequately tackle the challenge posed by variable operation durations in COSPs. This limitation results in a compromised solution quality for EAs. To bridge the gap between large-scale COSPs and real-world refineries, this paper introduces the formulation of a continuous-time model for a large-scale COSP, accounting for varying operating durations and practical requirements. In an effort to enhance the effectiveness of EAs in optimizing COSPs represented in continuous time, we propose a variable-length constrained evolutionary algorithm (VLCEA). The variable-length encoding in VLCEA, tailored to the specific problem, facilitates the generation of high-quality solutions with variable lengths. This feature addresses the challenge of determining solution length in advance based on a priori knowledge using conventional methods. Additionally, VLCEA incorporates a sorting strategy and a niching-based selection to expedite the search for promising regions, aiming to achieve better solutions in less time. Experimental results on practical cases not only verify that the continuous-time model produces higher quality solutions compared to the discrete-time model but also demonstrate that VLCEA significantly outperforms five state-of-the-art large-scale EAs with fixed length when applied to the continuous-time model. Note to Practitioners—Crude oil scheduling is one of the most challenging problems in the refinery scheduling process. With the increasing size of scheduling tasks, traditional methods have become inefficient and inaccurate. To address this, evolutionary algorithms (EAs) have emerged as powerful tools for solving large-scale crude oil scheduling problems (COSPs). However, existing EA-based approaches often rely on a fixed number of decisions, resulting in suboptimal solutions when confronted with insufficient prior knowledge. To overcome these obstacles, we have developed a comprehensive continuous-time model specifically designed to address COSPs in real-world refineries. Additionally, we propose a variable-length constrained evolutionary algorithm (VLCEA) as an effective approach for resolving these complex scheduling problems. Computational results demonstrate that the VLCEA achieves superior performance compared to three state-of-the-art EAs in practical large-scale cases. Consequently, the proposed VLCEA exhibits significant potential for solving real-world refinery scheduling problems and can be extended to tackle other scheduling problems with an unfixed number of decisions.}, archive = {J_TASE}, author = {Wanting Zhang and Wenli Du and Wei Du and Renchu He and Yaochu Jin}, doi = {10.1109/TASE.2024.3380744}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2526--2541}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Large-scale continuous-time crude oil scheduling: A variable-length evolutionary optimization approach}, volume = {22}, year = {2025}, }
• (2025). A composite high-speed and high-precision positioning approach for dual-drive gantry stage. TASE, 22, 2514--2525. (www)

  @article{ , abstract = {As an important part of the motion system in high-end manufacturing equipment, dual-drive gantry stage urgently requires higher performance. Planar rapid positioning is a classical application of the gantry stage, while unmodeled dynamics, coupled dynamics of mechanism, and conflict between rapidity and precision bring many difficulties into controller design. In this article, a coupled dynamic model is established and transformed for positioning to provide better guidance for controller design under the rotation mode of the cross beam. A composite positioning control method consisting of a high-speed moving section and a high-precision positioning section is proposed for linearmotors on the gantry to achieve both fast response and precise control simultaneously. In addition, a synchronization scheme is proposed for the dual-driven cross beam with coupled dynamics to enhance the positioning precision of the workbench and balance of the beam. Experiments are conducted on the $X$ -axis with one linearmotor for the composite positioning method and $Y$ -axis with two linearmotors for the synchronization scheme combined with the composite positioning control separately, through which the effectiveness and validity of the proposed method are verified. Note to Practitioners—Planar positioning of dual-drive gantry stage is a significant technical problem in industrial applications, which pursues the ultimate speed and precision performance. In this paper, a dynamic model with rotational mode is established and transformed for positioning in an intuitive way. In addition, a composite positioning method is proposed for linearmotors on the stage to achieve high-speed and high-precision performance in the case of common nonlinear dynamics and disturbances. And then, to achieve high positioning performance of the cross beam while maintaining the balance, a synchronous scheme incorporated with the composite positioning method is presented, based on the transformed model. In industrial scenarios, the proposed method is very practical since a precise, rapid and stable point-to-point planar motion can be obtained only with conventional dynamic models and parameters, and it is crucial for production efficiency and quality.}, archive = {J_TASE}, author = {Weichao Sun and Jianing Liu and Huijun Gao}, doi = {10.1109/TASE.2024.3380646}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2514--2525}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A composite high-speed and high-precision positioning approach for dual-drive gantry stage}, volume = {22}, year = {2025}, }
• (2025). A hierarchical multi-action deep reinforcement learning method for dynamic distributed job-shop scheduling problem with job arrivals. TASE, 22, 2501--2513. (www)

  @article{ , abstract = {The Distributed Job-shop Scheduling Problem (DJSP) is a significant issue in both academic and industrial fields. In real-world production, uncertain disturbances such as job arrivals are inevitable. In the paper, the DJSP with job arrivals is addressed with a Multi-action Deep Reinforcement Learning (MDRL) method. Firstly, a multi-action Markov Decision Process (MDP) is formulated, where a hierarchical multi-action space combining operation set and factory set is proposed. The reward function is related to the machine idle time. Additionally, the state transition is also elaborately designed, which includes four typical cases based on job arrival times. Then, a scheduling policy with two decision networks is proposed, where the Graph Neural Network (GNN) is applied to extract the intrinsic information of the scheduling scheme. A Proximal Policy Optimization (PPO) with two actor-critic frameworks is designed to train the model to achieve intelligent decision-making with hierarchical action selections. Extensive experiments are conducted based on 1350 instances. The comparison among 17 composite rules, 3 closely-rated DRL methods, and 2 metaheuristics has proven the outperformance of the proposed MDRL. The application of the MDRL in an automotive engine manufacturing company has demonstrated its engineering value in the industrial field. Note to Practitioners—The DJSP with job arrivals is a common challenge faced by equipment manufacturers, specifically in the electronic device manufacturing industry. These manufacturers are located in different areas and have varying facility configurations and operation trajectories. To address this challenge, a machine learning-based method can be applied for scheduling daily production tasks. This method divides the DJSP into two subproblems, namely job assigning and job sequencing, and uses two decision networks based on DRL to solve them. To address the uncertainty caused by job arrivals, the rescheduling process and the state update mechanism are carefully designed. A GNN is used for feature extraction at each decision point, and it feeds the decision networks with the extracted features to make the optimal selection. The proposed method has the ability of self-learning and self-adapting, and its effectiveness has been proven through experiments on 1350 test instances. Its practical application has been demonstrated in the production scenarios of an automotive engine manufacturing company. In the future, the method can be adopted to solve more complex distributed manufacturing problems that have constraints such as transportation costs and machine breakdowns.}, archive = {J_TASE}, author = {Jiang-Ping Huang and Liang Gao and Xin-Yu Li}, doi = {10.1109/TASE.2024.3380644}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2501--2513}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hierarchical multi-action deep reinforcement learning method for dynamic distributed job-shop scheduling problem with job arrivals}, volume = {22}, year = {2025}, }
• (2025). Robust depth and heading control system for a novel robotic dolphin with multiple control surfaces. TASE, 22, 2485--2500. (www)

  @article{ , abstract = {For field tasks, it is quite challenged to operate in a complex environment for the underwater robots, especially for those with multiple control surfaces due to different response and gain characteristics. To this end, this paper develops a highly integrated robotic dolphin followed by a robust motion control system. For better maneuverability and fault-tolerant capabilities, a newly-designed robotic dolphin is presented, owning a wide array of sensors and multiple control surfaces, in which passive flukes are particularly applied. On this basis, a robust motion control system is proposed, including a depth controller based on velocity-related allocation strategies and a heading controller based on clearance compensation. In detail, considering the degradation of motion performance caused by passive flukes, a sliding mode controller for gain uncertainty and an allocation-related parameter tuning strategy for inputs response characteristics are designed. Extensive simulations and aquatic experiments are conducted, and the obtained results demonstrate the satisfied maneuverability of the designed prototype and the effectiveness of the proposed methods. This study can lay a foundation for further development of robotic dolphins with a robust motion system to execute complex tasks in the field. Note to Practitioners—This paper is inspired by the issue of robust motion control system for a newly-designed practical robotic dolphin that possesses a passive tail and redundant control surfaces. The traditional methods are usually susceptible to uncertainties in the passive tail gain, exhibiting degraded control performance. Moreover, control oscillations and slow convergence speed often occur caused by neglecting the characteristics of different control surfaces, including response patterns and clearance. This paper suggests a robust depth controller based on velocity-related allocation strategies and a robust heading controller based on clearance compensation. Specifically, an allocation-related parameter tuning strategy is given by considering inputs response characteristics, including response speed, saturations, and hydrodynamic force variation patterns. To guarantee fine regulations of heading control, a nonlinear disturbance observer (NDOB)-based clearance compensation is proposed. Extensive aquatic experiments on the newly-designed robotic dolphin verified the effectiveness of the proposed methods. It is envisioned that all these presented results can provide valuable engineering practice insights for industry practitioners.}, archive = {J_TASE}, author = {Sijie Li and Zhengxing Wu and Jian Wang and Changlin Qiu and Min Tan and Junzhi Yu}, doi = {10.1109/TASE.2024.3380394}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2485--2500}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust depth and heading control system for a novel robotic dolphin with multiple control surfaces}, volume = {22}, year = {2025}, }
• (2025). Promoting automatic detection of road damage: A high-resolution dataset, a new approach, and a new evaluation criterion. TASE, 22, 2472--2484. (www)

  @article{ , abstract = {Using deep learning to detect road damage can significantly improve the effectiveness of road maintenance. To promote the development of road damage detection, we construct a high-resolution road damage data set named Asphalt Road Surface Disease Dataset(ARSDD), comprising 2297 images used in a real-world project. The annotation process is under the guidance of the road maintenance department, and it has more accurate labels and appropriate damage types. Most current road damage detection models are anchor-based, where one anchor corresponds to one sample. Therefore, these models are primarily limited by the setting of pre-defined anchors. Road damages have more extreme aspect ratios and scales than natural objects, and the general settings of anchors are inappropriate for road damage. In this paper, we propose a road damage detection model based on an improved adaptive training sample selection strategy, which can reduce manual anchor settings and is suitable for road damage detection. Moreover, as slight road damages tend to lose information during down-sampling, a cross-layer attention feature pyramid network is designed to compensate for this degradation in the spatial dimensions. While testing the ARSDD dataset, we find that the evaluation criterion for general object detection is unsuitable for road damage detection and propose a new post-processing method and diagonal-based evaluation criterion according to the characteristics of road damage. We validate our model using the 2018 Road Damage Dataset and our proposed dataset, and the results demonstrate the superiority of our model in road damage detection. Note to Practitioners—This paper was motivated by the problem of road damage detection, which is the key of intelligent road maintenance system. We comprehensively analyze the shortcomings of current publicly available content about road damage detection, including detection datasets, algorithms and evaluation metrics. Correspondingly, we first construct a road damage dataset from the real-world project, which contains six common categories of disease and is annotated under the guidance of the road maintenance department. Then we propose a detection model based on computer vision technology to improve the accuracy of damage detection. Finally, we propose a new post-processing method and a diagonal-based evaluation criterion based on the detection boxes. Any practitioner working on pavement inspection systems can use our released dataset and method to build a better-performing automated inspection system. In future research, we will try more refined inspection schemes and evaluation indicators, such as meshing the road surface images and evaluating the quality of the road based on the inspection results of each mesh.}, archive = {J_TASE}, author = {Tianxiang Yin and Wei Zhang and Jinqiao Kou and Ningzhong Liu}, doi = {10.1109/TASE.2024.3379945}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2472--2484}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Promoting automatic detection of road damage: A high-resolution dataset, a new approach, and a new evaluation criterion}, volume = {22}, year = {2025}, }
• (2025). Automatic identification of individual african leopards in unlabeled camera trap images. TASE, 22, 2460--2471. (www)

  @article{ , abstract = {This article describes an algorithm to solve the real-world animal identification problem, i.e., determine the unknown number of K individual animals in a dataset of N unlabeled camera-trap images of African leopards, provided by Panthera. To determine the leopards’ IDs, we propose an effective automated algorithm, that consists of segmenting leopard bodies from images, scoring similarity between image pairs, and clustering followed by verification. To perform clustering, we employ a modified ternary search that uses a novel adaptive k-medoids++ clustering algorithm. The best clustering is determined using an expanded definition of the silhouette score. A new post-clustering verification procedure is used to further improve the quality of a clustering. The algorithm was evaluated using the Panthera dataset that consists of 677 individual leopards taken from 1555 images, and resulted in a clustering with an adjusted mutual information score of 0.958 as compared to 0.864 using a baseline k-medoids++ clustering algorithm. Note to Practitioners—We proposed an effective automated algorithm to solve the real-world animal identification problem: identifying K unknown individual animals in N images of a given species, with most animals only represented by a single image. This algorithm is different from other methods that assume all images in a dataset are from known individuals and thus regard the animal ID problem as a retrieval identification task. Our approach consists of a new adaptive k-medoids++ clustering algorithm and a novel post-clustering verification procedure. The clustering is performed based on the degree of similarity between all image pairs in the dataset with the result validated using an expanded definition of the silhouette score. The accuracy of our algorithm was demonstrated on a real-world image dataset of African leopards, a small dataset with a relatively large ratio of $K/N$ , provided by Panthera. Code has been made available at: https://github.com/obaiga/Automatic-individual-animal-identification.}, archive = {J_TASE}, author = {Cheng Guo and Agnieszka Miguel and Anthony A. Maciejewski}, doi = {10.1109/TASE.2024.3379553}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2460--2471}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic identification of individual african leopards in unlabeled camera trap images}, volume = {22}, year = {2025}, }
• (2025). Robust attitude control with fixed exponential rate of convergence and consideration of motor dynamics for tilt quadrotor using quaternions. TASE, 22, 2445--2459. (www)

  @article{ , abstract = {In the existing literature on the robust control design of UAV systems, the controllers are designed without considering motor dynamics. Hence, if these controller gains are not correctly tuned, the system undergoes oscillation and may even go unstable. We have demonstrated this through an experiment in this work. Here, we propose a novel control strategy that considers actuator parameter uncertainties, including motor dynamics for a tilt quadrotor. This strategy is based on the traditional two-loop control scheme where the inner loop controls the angular velocity, and the outer loop controls the vehicle’s attitude based on quaternions. In the quaternion-based controller, usually, the convergence rate increases when the quaternion starts closer to its equilibrium point, thus making it challenging to design a linear controller for the inner loop. To overcome this, we propose a nonlinear control with a varying gain for the outer loop that ensures the quaternion has a fixed convergence rate. We propose the control design of the inner loop, which consists of a disturbance observer (DOB) and a linear controller. The DOB is optimally designed to minimize external disturbances in the presence of model uncertainties. With the DOB, a linear controller is designed for the inner loop, guaranteeing robust stability and performance against the model and actuator parameter uncertainties. The results of experimental flights are reported in this paper, and the corresponding videos are at https://youtu.be/dS9WUR0yLhA. Note to Practitioners—This paper was motivated by the potential application of the tilt quadrotors in industry, such as surveillance in confined spaces and landing on high-speed moving targets, due to their independent control of position and orientation. We demonstrate that a controller designed without consideration of motor dynamics can result in unwanted oscillatory or unstable behaviour in the case of a UAV. So, the proposed method considers these dynamics during the design. In most of the existing control designs based on quaternion, the convergence rate depends on the initial condition. This paper proposes a nonlinear control with a varying gain to achieve a faster response that ensures the quaternion has a fixed convergence rate. The stability and convergence of this nonlinear controller are analyzed mathematically and validated through experiments. In UAVs, the precise moment-of-inertia and motor parameters are difficult to estimate, and external disturbances such as wind gusts are always present in outdoor environments. Therefore, this paper designs a robust control scheme which guarantees robust stability and performance against inertia and motor parameter uncertainties. Secondly, a disturbance observer (DOB) whose filter coefficient is optimally tuned such that the effect of external disturbance is minimized in the presence of model uncertainties. The proposed approach is validated on a tilt quadrotor in an outdoor environment and has a faster response than the existing ones having same maximum convergence rate. Thus, this proposed approach has potential applications, like landing a tilt quadrotor on a ship while the ship is in motion under the influence of waves and winds.}, archive = {J_TASE}, author = {Sathyanarayanan Seshasayanan and Souradip De and Soumya Ranjan Sahoo}, doi = {10.1109/TASE.2024.3379430}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2445--2459}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust attitude control with fixed exponential rate of convergence and consideration of motor dynamics for tilt quadrotor using quaternions}, volume = {22}, year = {2025}, }
• (2025). A voltage sensorless robust control scheme for LCL grid-connected inverters based on GESO approach. TASE, 22, 2433--2444. (www)

  @article{ , abstract = {The traditional current control of the LCL grid-connected inverters often requires additional sensors for resonance damping, harmonic suppression and grid synchronization. A single current sensor control scheme is presented in this article to reduce the number of sensors and the risk of sensor failure on the basis of maintaining the similar performance as the traditional dual sensor control. The proposed control scheme is based on a generalized extended state observer (GESO) used to estimate the equivalent disturbance, while the estimation is used for disturbance rejection and phase-locked loop (PLL) synchronization. Aiming at the problem of current overshoot at startup time of sensorless strategy, a voltage pre-estimation strategy is propos-ed to reduce the current overshoot phenomenon. In addition, a low-pass filter is added to the current tracking loop to improve the system adaptability for grid impedance. Then the system is equivalent to three loops for the stability analysis and the parameter constraints of the controller are given. Moreover, it was found that the parameters in GESO are smaller than the nominal parameters of the inverter, which could improve the robust stability of the system. Finally, the effectiveness and feasibility of the proposed control scheme are verified through the experimental tests. Note to Practitioners—The state observer has been used in this paper to estimate the voltage and disturbance of the grid, which are used to control the grid-connected current and suppress the disturbance. The proposed method not only can reduce the number of sensors and improve the robust stability of the system, but also can reduce the risk of power grid instability caused by sensor failure on the premise of ensuring similar performance to traditional dual sensors. It ensures the stable operation of the power grid, and has a certain significance for expanding to the three-phase gr id-connected system.}, archive = {J_TASE}, author = {Danyun Li and Changsi Li and Dashuang Zhao and Kun Song and Zihan Zeng}, doi = {10.1109/TASE.2024.3379424}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2433--2444}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A voltage sensorless robust control scheme for LCL grid-connected inverters based on GESO approach}, volume = {22}, year = {2025}, }
• (2025). Autonomous navigation of magnetic microrobots with improved planning and control in complex environments. TASE, 22, 2421--2432. (www)

  @article{ , abstract = {Magnetic field-driven microrobots have shown high potential in the field of medical applications. Autonomous navigation is a crucial concern for magnetic microrobots, however, the path planning, actuation and control of magnetic microrobots still remain challenging, especially for complex and large-workspace human body environments. Depending on the specific task and environmental conditions, it is important to employ appropriate planning and control architectures for the magnetic navigation systems. In light of this objective, this paper introduces a novel navigation framework, using an improved path planning and following control method. An evolutionary strategy based RRT (ES-RRT) planner is designed to achieve a shorter, smoother and safer path. Furthermore, an extended state observer (ESO)-based controller is specifically designed for the path tracking process of the microrobots. This controller enables the microrobots to accurately follow the computed path. Experiments demonstrate the effectiveness of the proposed strategy: Feasible path in different conditions and environments can be obtained with short and smooth enough, and autonomous navigation following of microrobots is realized with satisfactory path tracking control accuracy. Note to Practitioners—In contrast to macroscale robots, microrobots face challenges when it comes to integrating onboard components such as processors and power sources. Consequently, alternative methods have been developed, including optical, chemical, and biological actuation. Among these approaches, the utilization of magnetic fields is particularly favorable due to its ability to penetrate deep tissues while ensuring high safety. Additionally, magnetic fields offer diverse propulsion options, such as rotating fields and oscillating fields, along with excellent controllability. Despite significant advancements in the fabrication, functionalization, and locomotion of magnetic microrobots, autonomous navigation remains an area that requires further development. In the medical application, the planning path of the microrobots needs to be short and smooth enough. Besides, considering the unknown dynamics and external disturbances of the system, it is extremely important for microrobots to complete the precise path following control of the planned path. The motivation of this work is to develop an effective navigation scheme for microrobots which consisting of a path planner and motion controller. A shorter and smoother path will be planned based on a novel improved RRT planner. The precise control of the path following then will be achieved by adopting an extended state observer (ESO) controller. The proposed methods would enable microrobots to act safely and greatly enhance robots’ capabilities.}, archive = {J_TASE}, author = {Yueyue Liu and Haoyu Wang and Xiaoyu Wu and Juntian Qu and Xinyu Liu and Qigao Fan}, doi = {10.1109/TASE.2024.3379364}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2421--2432}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous navigation of magnetic microrobots with improved planning and control in complex environments}, volume = {22}, year = {2025}, }
• (2025). Robust motion learning for musculoskeletal robots based on a recurrent neural network and muscle synergies. TASE, 22, 2405--2420. (www)

  @article{ , abstract = {Musculoskeletal robots with human-like joints, muscles, and actuation mechanisms are characterized by exceptional dexterity, compliance, and versatility. However, existing reinforcement learning methods for such robots rely on precise and sufficient state observation, rendering them vulnerable to perturbations. To address this limitation, this paper proposes a robust motion learning method based on a recurrent neural network (RNN) and muscle synergy. First, the proposed method utilizes task-joint-muscle space states to create an RNN-based neuromuscular controller. Furthermore, a motion learning method with a synergistic constraint of muscles is developed. Additionally, theoretical analysis confirms that the RNN-based controller is more robust to perturbations of state observation than a Multilayer Perceptron (MLP) based controller. The proposed method is evaluated on a simulated musculoskeletal robot and demonstrates superior robustness to other MLP-based reinforcement learning methods. Furthermore, the proposed method is also validated on a musculoskeletal robot hardware system, indicating its potential for real-world applications. Note to Practitioners—Musculoskeletal robots have shown promising potential in various applications, but their development and application have been restricted by the limited robustness of existing reinforcement learning methods. These methods work well in simulation with precise and sufficient state observation, but exhibit a considerable degradation of performance in real-world environments with perturbed and insufficient state observation. In this article, we propose a novel method to improve the robustness of motion learning for musculoskeletal robots using a recurrent neural network and muscle synergy. The proposed method is theoretically and experimentally validated to perform well not only under precise and sufficient state observation but also in the presence of perturbed and insufficient state observation. Our results demonstrate the effectiveness of the proposed method and motivate the application of reinforcement learning methods to musculoskeletal robots. This research contributes to the advancement of robust motion learning for musculoskeletal robots and paves the way for wider adoption in real-world applications.}, archive = {J_TASE}, author = {Jiahao Chen and Yaxiong Wu and Chaojing Yao and Xiao Huang}, doi = {10.1109/TASE.2024.3379247}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2405--2420}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust motion learning for musculoskeletal robots based on a recurrent neural network and muscle synergies}, volume = {22}, year = {2025}, }
• (2025). Automatic control of magnetic helical microrobots for on-plane docking. TASE, 22, 2395--2404. (www)

  @article{ , abstract = {Docking is a common step in magnetic helical microrobot operations. However, the boundary effect causes helical microrobots to drift laterally near the container bottom. The mismatch between orientation and movement direction further complicates manual steering. In this paper, an automatic control method for on-plane docking of magnetic helical microrobots is proposed, which divides the docking process into preparation, alignment, and insertion stages. First, the docking path is planned according to the position and orientation of the helix and the target. Second, a steering-based path-following controller guides the helix swim to the ready position for alignment. Third, another path-following controller realizes the position alignment through rotational-frequency-controlled movement direction while locking the helix orientation aligned with the target. Finally, the docking is completed in the insertion stage. Experiments verified the automatic docking of helical microrobots with static targets, including inserting into a slender gap/tube, assembling with a micro-shaft axially, and leaning against a rod radially. This method can eliminate the negative impact of operator’s fatigue and inexperience on docking operations and has the potential to be applied in biomedical and microoperations. Note to Practitioners—This paper is motivated by the problem of docking magnetically driven helical-shaped swimming microrobots with target objects on Petri dishes, which is a common step in capture, loading, assembly, and delivery operations. Here, two closed-loop path-following controllers based on steering and rotational frequency control guide the microrobot to approach and align with the target respectively along a calculated path. In alignment, the non-ideal lateral drift of helical microrobots is controlled by rotational frequency for position alignment. Experiments verified the automatic docking of micro-helixes with static targets, including inserting into a slender gap/tube, connecting with a micro-shaft, and leaning against a rod.}, archive = {J_TASE}, author = {Fu Zhao and Weibin Rong and Lefeng Wang and Lining Sun}, doi = {10.1109/TASE.2024.3379226}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2395--2404}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic control of magnetic helical microrobots for on-plane docking}, volume = {22}, year = {2025}, }
• (2025). Discrete-time neural control with flexible prescribed performance for constrained systems with fragility relaxation. TASE, 22, 2386--2394. (www)

  @article{ , abstract = {This article investigates prescribed performance neural autonomous control of discrete-time nonlinear systems subject to actuator saturation. Unlike existing sliding-mode-control (SMC) driven structure, we aim to construct a more general and user-friendly framework that achieves prescribed behaviors by indirectly stabilizing transformed errors in the discrete-time domain. We propose two new approaches to accomplish such objective. The first one is to compensate actuator saturation by developing a modification system, the state of which is further applied to design flexible terms that endow an ability to resiliently adjust prescribed envelopes according to saturation condition, which constitutes the second approach. In comparison with current discrete-time prescribed performance control (PPC), the proposed method avoids the singularity problem and consequently remedies the fragility defect. In addition, an adaptive neural back-stepping procedure is used to devise discrete-time PPC protocols that limit tracking errors inside the developed flexible prescribed envelopes to satisfy both transient and steady-state properties. Finally, the efficiency of design is illustrated via Lyapunov synthesis, and is further verified by numerical simulation. Note to Practitioners—The motivation arises from the requirement for autonomous tracking control of discrete-time nonlinear systems with actuator saturation and fragility relaxation. Unfortunately, existing SMC-driven discrete-time PPC isn’t able to achieve expected prescribed behaviors in the presence of actuator saturation. To break through this bottleneck, we firstly develop a new design framework, being different from existing SMC-based ones, to impose prescribed performance on tracking errors by indirectly stabilizing transformed errors, and then we define a new modification system which effectively compensates the actuator saturation so that the fragility defect can be remedied. On this basis, a novel discrete-time tracking control approach with flexible prescribed performance is addressed for constrained systems while relaxing the fragility. The obtained results in this study are of great significance for providing a more general and user-friendly framework for discrete-time PPC synthesis.}, archive = {J_TASE}, author = {Xiangwei Bu and Ju H. Park and Humin Lei}, doi = {10.1109/TASE.2024.3379224}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2386--2394}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Discrete-time neural control with flexible prescribed performance for constrained systems with fragility relaxation}, volume = {22}, year = {2025}, }
• (2025). Attack-resistant sliding mode control for markov jump systems with redundant channels: A novel two-layer mapping approach. TASE, 22, 2375--2385. (www)

  @article{ , abstract = {A security control problem is addressed for a kind of networked Markov jump system subject to two-side stochastic denial-of-service (DoS) attacks, in which the attack occurrence situation obeys the Markov chain. To resist the attack effects, a redundant channel protocol is adopted, wherein measurement/control signals are sent to the controller/actuator via the redundant channel when the primary channel suffers from cyber-attacks. According to the above redundant channel protocol, both measurement and input models are established to represent the received signals by the controller and actuator, respectively. Then, a novel two-layer mapping strategy via logical operations is proposed to describe attack occurrences and system jumping. These facilitate the design of a security sliding mode controller, under which the exponential mean-square stability of the networked Markov jump systems subject to DoS attacks and the corresponding sufficient conditions are derived. Eventually, the simulation results via the DC chopper circuit system are provided to illustrate the proposed redundant-channel-based security control method. Note to Practitioners—This paper is motivated by the attack resistance problem in networked control systems. With the revolutionary evolution of wireless communication technology, in many practical systems, such as smart power grids, unmanned aerial vehicles (UAVs), and unmanned surface vehicles (USVs), the physical devices including sensors, actuators, controllers, and other intelligent devices are interconnected and transmitted the data through a network infrastructure. The inevitable DoS attacks may cause reliability degradation of data transmission, which in practice results in shaky grid frequency, failed UAV formation, and inaccurate USV detection. Distinguished from the previous investigation results from a passive perspective, this paper proposes a redundant-channel-based method to solve the reliability degradation problem caused by DoS attacks from a proactive perspective. The developed redundant channel method against Markov-chain-based DoS attacks can improve the reliability of data transmission, and the proposed novel two-layer mapping strategy via logical operations can increase the freedom of security controller design. A DC chopper circuit is used to verify the effectiveness of the proposed novel control method and comparison simulation results show the advantages of the redundant channel method. How to improve the reliability of vulnerable channels is a knotty problem in communication. With the groundbreaking research of this paper, against other attacks, such as deception attacks, reply attacks, and frequency- and duration-constrained DoS attacks, the attack resistance problems via proactively strengthening the reliability of data transmission should be further explored in the future.}, archive = {J_TASE}, author = {Zhiru Cao and Yugang Niu and Ju H. Park and Kaiqun Zhu}, doi = {10.1109/TASE.2024.3379192}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2375--2385}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attack-resistant sliding mode control for markov jump systems with redundant channels: A novel two-layer mapping approach}, volume = {22}, year = {2025}, }
• (2025). DIBNN: A dual-improved-BNN based algorithm for multi-robot cooperative area search in complex obstacle environments. TASE, 22, 2361--2374. (www)

  @article{ , abstract = {Aiming at the area search task of a multi-robot system in an unknown complex obstacle environment, we propose a cooperative area search algorithm based on a dual improved bio-inspired neural network (DIBNN). First, we improve the BNN model to reduce the interference of the complex obstacle environment on robot decision making. Each robot generally chooses the neuron with the largest sum of surrounding activity values among adjacent neurons as its next movement position. Then, we propose a collaborative search mechanism. When a robot falls into a local deadlock state in the complex obstacle environment, the mechanism will guide the robot to quickly find unsearched areas. Finally, we conduct multi-robot area search simulation experiments under different obstacle environments and compare them with three baseline algorithms in this field. The simulation results verify that the proposed algorithm can efficiently guide the multi-robot to complete the area search task in the complex obstacle environment. Note to Practitioners—The motivation of this article arises from the need to develop fast and effective area search algorithms for practical applications such as UAV swarm reconnaissance and multiple mobile robots area search and rescue. The algorithms based on BNN has been widely used in search tasks under unknown environments due to its good scalability and efficiency. However, the efficiency of area search in complex obstacle environments cannot be guaranteed. In order to achieve efficient area search in unknown complex obstacle environments, the DIBNN algorithm is proposed. It utilizes a cooperative search mechanism and achieves better performance. DIBNN can also be applied to multi-robot systems in different scenarios, demonstrating strong scalability.}, archive = {J_TASE}, author = {Bo Chen and Hui Zhang and Fangfang Zhang and Yiming Jiang and Zhiqiang Miao and Hongnian Yu and Yaonan Wang}, doi = {10.1109/TASE.2024.3379166}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2361--2374}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DIBNN: A dual-improved-BNN based algorithm for multi-robot cooperative area search in complex obstacle environments}, volume = {22}, year = {2025}, }
• (2025). Safe learning by constraint-aware policy optimization for robotic ultrasound imaging. TASE, 22, 2349--2360. (www)

  @article{ , abstract = {Ultrasound-based medical examination usually requires establishing proper contact between an ultrasound probe and a human body that ensures the quality of ultrasound images. The scanning skills are quite challenging for a robot to learn primarily due to the complex coupling between the applied force profile and the resulting ultrasound image quality. While reinforcement learning appears as a powerful tool for learning complex robot skills, the deployment of these algorithms in medical robots demands special attention due to the evident safety concerns that arise from physical probe-tissue interactions. In this paper, we explicitly consider external constraints on the force magnitude when searching for the optimal policy parameters to enhance safety during ultrasound-guided robotic interventions. In particular, we study policy optimization under the framework of a constrained Markov decision process. The resulting gradient-based policy update is then subject to the involved constraints, which can be readily addressed by the primal-dual interior-point technique. In addition, upon the observation that policy update requires consecutive policies to be close to each other to have stable and robust performance with reinforcement learning algorithms, we design the learning rate of policy gradient from an imitation perspective. The performance of the proposed constraint-aware policy optimization method is validated with experiments of robotic ultrasound imaging for spinal diagnosis. Note to Practitioners—This paper was motivated by the problem of safely learning the optimal interaction force strategy to facilitate robotic ultrasound imaging. Existing approaches to robotic ultrasound imaging usually empirically set a constant value for the scanning force, despite the fact the force strategy plays an important role in the quality of the ultrasound images. This paper suggests the usage of reinforcement learning to identify the optimal interaction force due to the complex acoustic coupling between the force and the ultrasound image quality. Specifically, we propose constraint-aware reinforcement learning in view of the safety-critical issues as a result of physical human-probe interaction. We then conduct a theoretical analysis of the proposed safe reinforcement learning, including monotonic improvement and policy value bound under mild assumptions. Preliminary real experiments on ultrasound imaging of the spine of a phantom for scoliosis assessment suggest that the proposed approach can safely learn the optimal scanning force without violating the prescribed force threshold. In the future, we would like to apply our approach to learning the optimal scanning force on different organs of interest of human subjects.}, archive = {J_TASE}, author = {Anqing Duan and Chenguang Yang and Jingyuan Zhao and Shengzeng Huo and Peng Zhou and Wanyu Ma and Yongping Zheng and David Navarro-Alarcon}, doi = {10.1109/TASE.2024.3378915}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2349--2360}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safe learning by constraint-aware policy optimization for robotic ultrasound imaging}, volume = {22}, year = {2025}, }
• (2025). Automated point-of-care semen analysis using smartphone imaging and occlusion-aware multi-object tracking. TASE, 22, 2335--2348. (www)

  @article{ , abstract = {This paper presents an automated point-of-care semen analysis method that uses smartphone imaging to visually measure sperm concentration and motility of semen samples. The proposed method follows the same visual tracking scheme as laboratory semen analysis systems, aiming to match clinical standards while being suitable for point-of-care use. A boundary-sensitive segmentation network is developed to identify and distinguish sperm from impurities in raw semen. A novel occlusion-aware multi-sperm tracking algorithm is proposed to tackle challenges posed by smartphone imaging and undiluted raw semen samples. For automated motility measurement, an occlusion-awareness module is proposed to robustly track multiple sperm during frequent sperm crossover/occlusion. The module combines the segmented contour and kinematic-based probabilistic modeling to determine the occlusion status of both targets and measurements, facilitating fundamental improvement to feasible joint event enumeration to enable robust data association. The proposed method achieved a high success rate of 95.14% for tracking occluded sperm, with low mean errors for sperm concentration (2.03 million/ml) and motility (1.58%), outperforming existing multi-sperm tracking methods. In clinical tests involving 50 participants, our method exhibited good agreement with clinical standards (Spearman rank correlation coefficients of 0.94 for concentration and 0.89 for motility) even when used by inexperienced users. Note to Practitioners—Semen analysis is the gold standard method for assessing male reproductive capacity. Clinical semen analysis routinely uses professional computer-assisted semen analysis (CASA) systems to examine sperm concentration and motility; however, clinical visits for semen analysis are not always feasible due to the unavailability of such professional systems and the mental stress brought by clinical visits. This work provides a point-of-care semen analysis method. For hardware, a smartphone microscopic imaging modality was developed to enable clear visualization of sperm with the built-in smartphone camera. For software, an occlusion-aware multi-sperm tracking algorithm was proposed to automatically measure sperm concentration and motility. In addition to intensively validating the proposed point-of-care method against clinical CASA systems, this work also analyzed different types of tracking failure and quantified their effects on the automated evaluation of sperm concentration and motility. The techniques pave the way for further improvement in both point-of-care and clinical semen analysis. Overall, this work offers an accessible and reliable tool for automated male fertility evaluation.}, archive = {J_TASE}, author = {Aojun Jiang and Miao Hao and Yiqian Li and Jiaqi Wang and Chunfeng Yue and Zongjie Huang and Ying Yang and Meng Ma and Rongan Zhai and Junhui Zhu and Changhai Ru and Qifeng Lyu and Yu Sun and Zhuoran Zhang}, doi = {10.1109/TASE.2024.3378702}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2335--2348}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automated point-of-care semen analysis using smartphone imaging and occlusion-aware multi-object tracking}, volume = {22}, year = {2025}, }
• (2025). Adaptive non-gaussian cubature filter based on GS-MCC with correlated multiplicative noises. TASE, 22, 2318--2334. (www)

  @article{ , abstract = {Due to complex environmental factors, the position measurement data of mobile robots are susceptible to pollution from multiplicative noise and non-Gaussian characteristics, which poses a challenge for the existing adaptive filtering methods to obtain high-performance pose estimation. Based on the comprehensive consideration of non-Gaussian multiplicative noise and correlated noise, a non-Gaussian adaptive Cubature Kalman Filter design method based on Gaussian sum and maximum correntropy techniques is proposed. Firstly, an enhanced Gaussian-sum and Cubature Kalman Filter method is designed to re-derive the innovation covariance and cross-variance for the multiplicative noise system. Secondly, under the framework of the maximum correntropy filter, the design strategy of the Gaussian kernel function weight matrix is improved, and a Cubature Kalman Filter method based on the improved maximum correntropy criterion is proposed. Then, based on consideration of the correlation between multiplicative noise and measurement noise, a Gaussian sum maximum correntropy Cubature Kalman Filter method with a known correlation coefficient is established. Finally, an adaptive Gaussian sum maximum correntropy Cubature Kalman Filter method that can dynamically estimate the correlation coefficient of multiplicative noise in real time is proposed, which can perform joint high-performance estimation of mobile robot’s pose and correlation coefficient. Note to Practitioners—This paper focuses on nonlinear multiplicative noise systems which are non-Gaussian. On one hand, non-Gaussian multiplication noise is pervasive in all types of systems, seriously affecting signal transmission and data acquisition. On the other hand, the correlation between multiplicative noise and measurement noise will seriously affect filter accuracy and create new challenges. Particularly, as it is difficult to obtain the correlation coefficient in the current system, the method to estimate the correlation coefficient between non-Gaussian noise is a challenging problem. To address these issues. Firstly, the Gaussian sum technique is used to approximate the non-Gaussian noise. Secondly, the correntropy technique is used to further reduce the influence of non-Gaussian noise on the filter accuracy. Then, the problem of correlated noises is solved by improving the cost function. Finally, the problem of inaccurate noise correlation coefficient in the real system is solved by the principle of covariance correspondence. This method is suitable for systems with correlated multiplicative noise. If the accuracy of sensor measurement data is too low, it will affect estimation performance. In future research, we will address the issue of reduced filtering accuracy in various correlated noise environments.}, archive = {J_TASE}, author = {Kai Lin and Quanbo Ge and Hong Li and Sheng Chen}, doi = {10.1109/TASE.2024.3378391}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2318--2334}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive non-gaussian cubature filter based on GS-MCC with correlated multiplicative noises}, volume = {22}, year = {2025}, }
• (2025). Command filtered adaptive asymptotic tracking control of nonlinear systems with unknown control directions via logic-based switching. TASE, 22, 2309--2317. (www)

  @article{ , abstract = {This article focuses on the command-filtered-based asymptotic tracking control problem of uncertain nonlinear systems with mismatching disturbances. Different from the previous command filtered control results, the control directions are assumed to be unknown, and a novel logic-based switching method is presented to handle them instead of the classical Nussbaum function method. Besides, some smooth functions are introduced into the virtual controllers to compensate the unknown disturbances. Then, an adaptive tracking controller with dynamic parameter is proposed for the considered nonlinear system. With the aid of the supervisory function, a logic-based switching algorithm is developed to regulate the dynamic controller parameter. Furthermore, it is shown that asymptotic tracking control can be achieved for the considered system with finite switchings. Finally, a simulation example is given to verify the effectiveness of the novel design scheme. Note to Practitioners—This paper aims at presenting a novel asymptotic tracking control scheme for uncertain nonlinear systems with unknown control signs and external disturbances, which can model most practical systems such as spacecraft with actuation sign errors. Although considerable command filter control results have been obtained for nonlinear systems, it is still challenging and difficult to realize the command filter asymptotic tracking control of nonlinear system with multiple unknown control signs. To this end, switching functions with dynamic parameter are introduced to design the compensation signals and controller. Besides, an effective switching algorithm is proposed to regulate the dynamic parameters. Based on the proposed control scheme, the command filter control problem can be handled for nonlinear system with multiple unknown control signs. Besides, some smooth functions are also utilized to compensate the mismatching disturbances such that the asymptotic tracking control can be realized. It is worth pointing out that our proposed control method has potential application in practical systems including spacecraft and autonomous underwater vehicles.}, archive = {J_TASE}, author = {Jiali Ma and Jiaqi Wang and Shengyuan Xu}, doi = {10.1109/TASE.2024.3378328}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2309--2317}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Command filtered adaptive asymptotic tracking control of nonlinear systems with unknown control directions via logic-based switching}, volume = {22}, year = {2025}, }
• (2025). A survey of object goal navigation. TASE, 22, 2292--2308. (www)

  @article{ , abstract = {Object Goal Navigation (ObjectNav) refers to an agent navigating to an object in an unseen environment, which is an ability often required in the accomplishment of complex tasks. Though it has drawn increasing attention from researchers in the Embodied AI community, there has not been a contemporary and comprehensive survey of ObjectNav. In this survey, we give an overview of this field by summarizing more than 70 recent papers. First, we give the preliminaries of the ObjectNav: the definition, the simulator, and the metrics. Then, we group the existing works into three categories: 1) end-to-end methods that directly map the observations to actions, 2) modular methods that consist of a mapping module, a policy module, and a path planning module, and 3) zero-shot methods that use zero-shot learning to do navigation. Finally, we summarize the performance of existing works and the main failure modes and discuss the challenges of ObjectNav. This survey would provide comprehensive information for researchers in this field to have a better understanding of ObjectNav. Note to Practitioners—This work was motivated by the increased interest in real-world applications of mobile robots. Object Goal Navigation (ObjectNav), which is an important task in these applications, requires an agent to find an object in an unseen environment. To accomplish that, the agent needs to be equipped with the capability to move in the environment, decide where to go, and recognize the object categories. So far, most works on ObjectNav have been done in a simulation environment. We present an overview of the existing works in ObjectNav and introduce them in three categories. Additionally, we analyze the current performance of ObjectNav and the challenges for future research. This paper provides researchers and practitioners with a comprehensive overview of the developed methods in ObjectNav, which can help them to have a good understanding of this task and develop suitable solutions for applications in the real world.}, archive = {J_TASE}, author = {Jingwen Sun and Jing Wu and Ze Ji and Yu-Kun Lai}, doi = {10.1109/TASE.2024.3378010}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2292--2308}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A survey of object goal navigation}, volume = {22}, year = {2025}, }
• (2025). Dual optimization-based distributed tracking control under completely unknown dynamics. TASE, 22, 2281--2291. (www)

  @article{ , abstract = {Unlike existing results on output tracking control in multi-agent systems, which mainly focus on the relative state among agents, in some cases the state of the system may not be available or measurable. In this paper, we investigate the model-free learning-based distributed optimal tracking control for heterogeneous multi-agent systems with dynamic output feedback under a switching reinforcement learning algorithm. First, a relative output-based distributed observer is developed without exchanging state information, which can significantly reduce the interaction load and broaden the range of applications. Then, a distributed feedback-feedforward controller is proposed, where the optimal feedback and feedforward gain matrices can be learned online by solving two optimization problems, employing policy iteration-based reinforcement learning instead of relying on the leader’s state as in existing studies. Subsequently, the policy iteration algorithm (PI) is modified into the value iteration (VI) learning algorithm, which can relax the requirement for an initial stabilizing control policy and does not depend on the known dynamical model. Additionally, a switching reinforcement learning algorithm is put forward by fully integrating the merits of the previously mentioned methods. The new algorithm not only overcomes the initial stabilizing assumption, but also ensures the convergence of the algorithm in a model-free fashion. Finally, a simulation example is provided to illustrate the theoretical analysis. Note to Practitioners—This paper investigates the model-free learning-based distributed optimal tracking control for heterogeneous multi-agent systems with dynamic output feedback under a switched reinforcement learning algorithm. The proposed distributed cooperative control algorithms can be applied to multiple ground vehicles, air vehicles, and underwater vehicles. Unlike existing results on output tracking control, the obtained results rely on accurate system dynamics and ignore the transient performance, which makes the designed controller far from optimal in potential applications. Moreover, the optimal learning algorithm strictly relies on the initial stabilizing control policy related to accurate system dynamics, and it may be helpless when the system model is completely unknown. To overcome those issues, we developed a new data-driven algorithm or data-based model-free reinforcement learning algorithm to study distributed output tracking control with relative output information by collecting system data instead of using accurate system dynamics. The new algorithm not only overcomes the initial stabilizing assumption but also ensures the algorithm convergence in a model-free fashion. Potential applications of the proposed control algorithms include cooperative formation control, and secondary control of microgrids.}, archive = {J_TASE}, author = {Di Mei and Jian Sun and Yong Xu and Lihua Dou}, doi = {10.1109/TASE.2024.3377894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2281--2291}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual optimization-based distributed tracking control under completely unknown dynamics}, volume = {22}, year = {2025}, }
• (2025). Autonomous landing of the quadrotor on the mobile platform via meta reinforcement learning. TASE, 22, 2269--2280. (www)

  @article{ , abstract = {Landing a quadrotor on a mobile platform moving with various unknown trajectories presents special challenges, including the requirements of fast trajectory planning/replanning, accurate control, and the adaptability for different target trajectories, especially when the platform is non-cooperative. However, previous works either assume the platform moves along a predefined trajectory or decouple planning from control which may cause a delay in tracking. In this work, we integrate planning and control into a unified framework and present an efficient off-policy Meta-Reinforcement Learning (Meta-RL) algorithm that enables a quadrotor (agent) to land on a mobile platform with various unknown trajectories autonomously. In our approach, we disentangle task-specific policy parameters by a separate adapter network to shared low-level parameters and learn a probabilistic encoder to extract common structures across different tasks. Specifically, during meta-training, we sample different trajectories from the task distribution, and then the probabilistic encoder accumulates the necessary statistics from past experience into the latent variables that enable the policy to perform the task. At meta-testing time, when the quadrotor is faced with an unseen trajectory, the latent variables can be sampled according to past interactions between the quadrotor and the mobile platform and held constant during an episode, enabling rapid trajectory-level adaptation. We assume similar tasks share a common low-dimensional structure in the representation of the policy network and the task-specific information is learned in the head of the policy. Accordingly, we further propose a separate adapter net as a supervised learning problem. The adapter net learns the weights of the policy’s output layer for each meta-training task given by the environment interactions from the agent. When adapting to a new task during meta-testing, we fix the shared model layers and predict the head weights for the new task using the trained adapter network. This ensures that the pretrained policy can efficiently adapt to different tasks, which boosts the out-of-distribution performance. Our method can directly control the pitch, roll, yaw angle, and thrust of the quadrotor, yielding a fast response to the trajectory change. Simulation results show the superiority of our method both in success rate and adaptation efficiency over other RL algorithms on meta-testing tasks. The real-world experimental results compared with traditional planning and control algorithms demonstrate the satisfactory performance of our autonomous landing method, especially its robustness in adapting to unknown dynamics. Note to Practitioners—Given the challenge posed by the motion uncertainty when a quadrotor lands on a mobile platform with an unknown trajectory, there hasn’t been a well-established solution, as far as we know. This paper introduces meta-reinforcement learning, incorporating a latent variable encoder to extract common features from training tasks, and designing an adapter network to enhance the ability of policy networks to adapt to new tasks, thereby enhancing the landing performance of the agent. The proposed method demonstrates promising results in both simulation and experiments.}, archive = {J_TASE}, author = {Qianqian Cao and Ziyi Liu and Hai Yu and Xiao Liang and Yongchun Fang}, doi = {10.1109/TASE.2024.3377810}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2269--2280}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autonomous landing of the quadrotor on the mobile platform via meta reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Event-triggered FCS-MPC with sliding mode observer for permanent magnet synchronous motor servo motion systems. TASE, 22, 2257--2268. (www)

  @article{ , abstract = {This work proposes an event-triggered finite control set model predictive control (FCS-MPC) strategy with a sliding mode observer (SMO) for permanent magnet synchronous motor (PMSM) drives to reduce calculation burden and switching losses, which arise from redundant control actions, without sacrificing the servo tracking performances. A sliding mode observer is utilized to identify load torque to enhance the robustness against loading disturbance for the PMSM control system. The estimation outputs are utilized as feed-forward compensation for the references, which are determined or constructed by the outer-loop controller. By introducing an electrical angular velocity region and decoupling the time-varying system function, a developed mathematical description for the PMSM model is reconstructed. An event-triggered FCS-MPC scheme with the triggering condition derived from the proposed PMSM model is carried out to deal with the calculation burden problem, which is the main challenge for applying the FCS-MPC theory to power electronics. Extensive comparative experiments are completed to validate the stability and efficiency of the proposed strategy. Note to Practitioners—This study is motivated by the difficulties in practical applications for the FCS-MPC methodology to PMSM drives. In recent years, a large amount of FCS-MPC strategies have been investigated with respect to the servo motion and tracking performance for PMSM. Realized by enumeration-based predictions, the control algorithm brings exponentially increased calculation burden for long prediction horizons, which are grown to benefit the important closed-loop performance. The efficient traditional FCS-MPC algorithm for PMSM drives puts forward considerable demands on hardware computation resources, which leads to high expense and harmful time delay. To our practical knowledge, there exist repetitive optimized control inputs provided by the FCS-MPC controller for PMSM drives, such that the resulting redundant actions in electronic circuits contain challenges in energy consumption and heat dissipation. Furthermore, coupling disturbances are inevitable for industrial applications of PMSM, among which the load torque has a major impact on control performances. To address these problems, an event-triggering mechanism and a slide mode observer (SMO) are introduced into the FCS-MPC scheme for PMSM servo systems. The triggering condition is designed with guaranteed input-to-state stability. The estimation results from the observer are used for the feed-forward channel, and the enumeration-based prediction for FCS-MPC is processed only when the triggering condition is satisfied. The experimental results demonstrate the effectiveness in reducing the computational burden and switching losses, as well as strengthening system robustness against unknown loading disturbances.}, archive = {J_TASE}, author = {Wei Shen and Luwei Shao and Dongchen Liu and Junzheng Wang and Chuanyu Ge}, doi = {10.1109/TASE.2024.3377639}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2257--2268}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered FCS-MPC with sliding mode observer for permanent magnet synchronous motor servo motion systems}, volume = {22}, year = {2025}, }
• (2025). Data-driven methods applied to soft robot modeling and control: A review. TASE, 22, 2241--2256. (www)

  @article{ , abstract = {Soft robots show compliance and have infinite degrees of freedom. Thanks to these properties, such robots can be leveraged for surgery, rehabilitation, biomimetics, unstructured environment exploring, and industrial grippers. In this case, they attract scholars from a variety of areas. However, nonlinearity and hysteresis effects also bring a burden to robot modeling. Moreover, following their flexibility and adaptation, soft robot control is more challenging than rigid robot control. In order to model and control soft robots, a large number of data-driven methods are utilized in pairs or separately. This review first briefly introduces two foundations for data-driven approaches, which are physical models and the Jacobian matrix, then summarizes three kinds of data-driven approaches, which are statistical method, neural network, and reinforcement learning. This review compares the modeling and controller features, e.g., model dynamics, data requirement, and target task, within and among these categories. Finally, we summarize the features of each method. A discussion about the advantages and limitations of the existing modeling and control approaches is presented, and we forecast the future of data-driven approaches in soft robots. A website (https://sites.google.com/view/23zcb) is built for this review and will be updated frequently. Note to Practitioners—This work is motivated by the need for a review introducing soft robot modeling and control methods in parallel. Modeling and control play significant roles in robot research, and they are challenging especially for soft robots. The nonlinear and complex deformation of such robots necessitates specific modeling and control approaches. We introduce the state-of-the-art data-driven methods and survey three approaches widely utilized. This review also compares the performance of these methods, considering some important features like data amount requirement, control frequency, and target task. The features of each approach are summarized, and we discuss the possible future of this area.}, archive = {J_TASE}, author = {Zixi Chen and Federico Renda and Alexia Le Gall and Lorenzo Mocellin and Matteo Bernabei and Théo Dangel and Gastone Ciuti and Matteo Cianchetti and Cesare Stefanini}, doi = {10.1109/TASE.2024.3377291}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2241--2256}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven methods applied to soft robot modeling and control: A review}, volume = {22}, year = {2025}, }
• (2025). Distributed coverage control of constrained constant-speed unicycle multi-agent systems. TASE, 22, 2225--2240. (www)

  @article{ , abstract = {This paper proposes a novel distributed coverage controller for a multi-agent system with constant-speed unicycle robots (CSUR). The work is motivated by the limitation of the conventional method that does not ensure the satisfaction of hard state- and input-dependent constraints and leads to feasibility issues for multi-CSUR systems. In this paper, we solve these problems by designing a novel coverage cost function and a saturated gradient-search-based control law. Theoretical proofs are provided to guarantee that the CSURs ultimately move to the optimal coverage configuration without moving out of the covered domain. The controller is implemented in a distributed manner based on a novel communication standard among the agents. A series of simulation studies are conducted to validate the correctness of our theory by showing the efficacy of the proposed coverage controller in different initial conditions and with various control parameters. A comparison study in simulation reveals the advantage of the proposed method over the conventional method in terms of avoiding infeasibility. The experimental study verifies the applicability of the method to real robots. The development procedure of the method from theoretical analysis to experimental validation provides a novel framework for multi-agent system coordinate control with complex dynamics. Note to Practitioners—This paper gives a novel method to effectively cover a polygonal area using multiple constant-speed unicycle robots (CSUR) like wheeled robots and fixed-wing unmanned aerial vehicles (fUAV). Compared to the conventional approaches, our method allows these robots to cover a target region using circular orbits without departing the covered region. Also, the method satisfies common control saturation constraints in practice and can be implemented in a reliable distributed scheme. While the efficacy and correctness of the proposed method are rigorously proved using control theory, we also provide necessary interpretive elucidations to explain its underlying mechanism and selection rationale. The method is validated to be effective for wheeled robots in experimental studies, although it can also be applied to fUAVs in theory.}, archive = {J_TASE}, author = {Qingchen Liu and Zengjie Zhang and Nhan Khanh Le and Jiahu Qin and Fangzhou Liu and Sandra Hirche}, doi = {10.1109/TASE.2024.3377131}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2225--2240}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed coverage control of constrained constant-speed unicycle multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Parallel self-assembly for a multi-USV system on water surface with obstacles. TASE, 22, 2213--2224. (www)

  @article{ , abstract = {Parallel self-assembly is an efficient approach to accelerate the assembly process for modular robots. However, these approaches cannot accommodate complicated environments with obstacles, which restricts their applications. We in previous work consider the surrounding stationary obstacles and propose a parallel self-assembly planning algorithm. With this algorithm, modular robots can avoid immovable obstacles when performing docking actions, which adapts the parallel self-assembly process to complex scenes. The algorithm was simulated in 25 distinct maps with different obstacle configurations and shows a significantly higher success rate, which is more than 80%, compared to the existing parallel self-assembly algorithms. For verification in real-world applications, we in this paper develop a multi-agent hardware testbed system. The algorithm is successfully deployed on four omnidirectional unmanned surface vehicles, CuBoats. The navigation strategy that translates the high-level discrete plan to the continuous controller on the CuBoats is presented. The algorithm’s feasibility and flexibility were demonstrated through successful self-assembly experiments on 5 maps with varying obstacle configurations. Note to Practitioners—This paper addresses deploying of self-assembly technologies for modular robots in practical environments with obstacles to facilitate overwater construction tasks or collective transportation systems. Stationary obstacles may severely influence the assembly planning and robot routing processes. Moreover, efficient task coordination, robot navigation, and structure formation are required for large-scale assembly tasks. The algorithm in this work allows all participating robots to navigate online and connect simultaneously to promote efficiency. The strategy presented here endows the robots’ assembly with obstacle-avoidance capability in dense environments. This work will interest those pursuing efficient assembly in scenes with surrounding obstacles. Our hardware experiments demonstrate a concept system and verify the real-time performance of the algorithm under limited computing power. The approach introduced here is not applicable to robots with heterogeneous shapes, three-dimensional target structures, or overcrowded environments with too many obstacles.}, archive = {J_TASE}, author = {Lianxin Zhang and Yihan Huang and Zhongzhong Cao and Yang Jiao and Huihuan Qian}, doi = {10.1109/TASE.2024.3376981}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2213--2224}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Parallel self-assembly for a multi-USV system on water surface with obstacles}, volume = {22}, year = {2025}, }
• (2025). Off-line programming framework for sorting task based on human-demonstration. TASE, 22, 2199--2212. (www)

  @article{ , abstract = {Sorting tasks are typical applications in the product and industrial domains. When facing new settings, such as different types of objects and their positions, the system has to be reprogrammed by experienced engineers, which decreases production efficiency and increases labour and downtime costs. In this paper, an off-line programming framework for sorting tasks based on human demonstration has been developed, taking advantage of machine vision and programming-by-demonstration concepts. Only one eye-to-hand depth camera has been used to record both the demonstrated trajectory and the corresponding relationship of the sorting task. A feature generalization algorithm has also been proposed to ensure that the system is able to reproduce the sorting task using both the key features of the trajectory and the sorting features. Finally, tracking, object and target matching, and generalization experiments have been conducted using the JAKA Zu7 robot with different types of fruit models and target boxes. The results show that the proposed framework is able to record the sorting trajectory at a frequency of 10 Hz, acquire and maintain the sorting relationship after new demonstration, and automatically reproduce the sorting task in new settings. Note to Practitioners—This paper is motivated by a goal to accomplish intuitive and user-friendly off-line programming for sorting task. The programming framework should be usable by non-experts and be able to complete programming tasks with a few demonstrations when faced with new scenarios, such as different types of objects, different correspondence relationships of the object, and different positions of the object. By observing the human demonstrated sorting tasks, the above-mentioned features can be effectively extracted, so as to complete the convenient teaching. The underlying principle of this paper is to let the robot observe human demonstrated sorting task by a depth camera, tracking the relative trajectory between the object and box as well as recording their matching information. The key features of the trajectory are extracted and further generalized combing the matching information to automatically reproduce the executable path in different settings. The proposed framework allows the users without prior programming knowledge to program the sorting task in new setting by simply conducting the pick and place process with their hands.}, archive = {J_TASE}, author = {Wei Du and Cheng Ding and Jianhua Wu and Zhenhua Xiong}, doi = {10.1109/TASE.2024.3376712}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2199--2212}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Off-line programming framework for sorting task based on human-demonstration}, volume = {22}, year = {2025}, }
• (2025). Edge-assisted multi-robot visual-inertial SLAM with efficient communication. TASE, 22, 2186--2198. (www)

  @article{ , abstract = {The integration of cloud computing and edge computing is an effective way to achieve global consistent and real-time multi-robot Simultaneous Localization and Mapping (SLAM). Cloud computing effectively solves the problem of limited computing, communication and storage capacity of terminal equipment. However, limited bandwidth and extremely long communication links between terminal devices and the cloud result in serious performance degradation of multi-robot SLAM systems. To reduce the computational cost of feature tracking and improve the real-time performance of the robot, a lightweight SLAM method of optical flow tracking based on pyramid IMU prediction is proposed. On this basis, a centralized multi-robot SLAM system based on a robot-edge-cloud layered architecture is proposed to realize real-time collaborative SLAM. It avoids the problems of limited on-board computing resources and low execution efficiency of single robot. In this framework, only the feature points and keyframe descriptors are transmitted and lossless encoding and compression are carried out to realize real-time remote information transmission with limited bandwidth resources. This design reduces the actual bandwidth occupied in the process of data transmission, and does not cause the loss of SLAM accuracy caused by data compression. Through experimental verification on the EuRoC dataset, compared with the current most advanced local feature compression method, our method can achieve lower data volume feature transmission, and compared with the current advanced centralized multi-robot SLAM scheme, it can achieve the same or better positioning accuracy under low computational load. Note to Practitioners—The purpose of this paper is to reduce the communication load of a Cloud-Edge-Robot system by compressing and transmitting of keyframes and non-keyframes, respectively, which is suitable for a multi-robot SLAM system and can realize multi-robot joint localization and sparse map reconstruction under efficient communication. Currently, remote SLAM or centralized multi-robot SLAM is usually implemented by transferring the whole image or the features and descriptors of the image. In this paper, lightweight SLAM optical flow tracking based on pyramid IMU prediction is implemented to track non-keyframes. At the edge server, tracking between non-keyframes is realized only by transmitting keypoints. For keyframes, the pose estimation is realized by transmitting compressed features and descriptors. Multi-robot localization and map fusion are realized in the cloud through key frame feature information. Experiments on public datasets show that this method is feasible and can achieve high-precision joint positioning with a low amount of transmitted data. In future studies, we will apply this framework to more real-world systems, while achieving rich, accurate map fusion with more advanced features.}, archive = {J_TASE}, author = {Xin Liu and Shuhuan Wen and Jing Zhao and Tony Z. Qiu and Hong Zhang}, doi = {10.1109/TASE.2024.3376427}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2186--2198}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Edge-assisted multi-robot visual-inertial SLAM with efficient communication}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time extended state observer-based bipartite formation control of vehicle multi-agent systems. TASE, 22, 2174--2185. (www)

  @article{ , abstract = {Formation control is one of critical topics in cooperative control. In this paper, we design a prescribed-time bipartite formation control strategy for vehicle multi-agent systems with external disturbances and internal unknown dynamics. A prescribed-time extended state observer (PTESO) is developed to compensate for total disturbances in a prescribed time, and the convergence time can be set in advance for different initial conditions. In order to make the formation error converge in a prescribed time, two novel time functions are introduced and applied to the control strategy. A prescribed-time control strategy with PTESO is developed for bipartite vehicle formation, and the relationship of the convergence time between PTESO and formation error is illustrated. Stability analysis shows the observation error and bipartite formation error are steered to converge their prescribed time, respectively. Furthermore, simulations are conducted to demonstrate the effectiveness of the prescribed-time bipartite formation control strategy. Note to Practitioners—This paper presents a prescribed-time bipartite formation control strategy for vehicle multi-agent systems with external disturbances and internal unknown dynamics. The total disturbance of each agent is compensated by a prescribed-time extended state observer (PTESO). In this PTESO, the convergence time is adjustable in advance for different initial conditions, and the peak phenomenon, caused by large gains, is avoided owing to adjustable and smaller time-varying gains. Two novel prescribed-time functions are proposed for vehicle MASs. The proposed functions remove the requirement that calculating gains mathematically, and the issue of jumping variation, caused by modifying coefficients in traditional prescribed-time functions arbitrarily, is also addressed. This strategy provides a feasible strategy for industrial applications.}, archive = {J_TASE}, author = {Yang Yang and Tianqi Yang and Shicai Zhou and Wei Sun and Defeng Wu}, doi = {10.1109/TASE.2024.3376081}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2174--2185}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time extended state observer-based bipartite formation control of vehicle multi-agent systems}, volume = {22}, year = {2025}, }
• (2025). Comprehensive quantitative tuning method for multiple LADRCs in high order plus time delay processes. TASE, 22, 2159--2173. (www)

  @article{ , abstract = {In response to the challenge posed by the limitations of current active disturbance rejection controllers (ADRCs) in parameter tuning and restricted control efficacy for $K{e^{-\tau s}}/{\left ({Ts+1 }\right)^{n}}$ -type high-order processes, two improved linear active disturbance rejection controllers (LADRCs) are proposed in this paper: model-aided delay LADRC (MDLADRC) and compound LADRC (CLADRC). Utilising frequency domain analysis, comprehensive quantitative parameter tuning rules for LADRCs are derived, based on principles of robust stability. The aim of the rules is to address the issue of parameter tuning for both existing LADRCs and the two newly proposed variants. The simulation and comparison results demonstrate the superiority of the improved LADRCs and the tuning rules, and reveal the relationship between the order of LADRCs and their time domain indices. Experimental results on the Peltier temperature system, water tank, and superheated steam temperature (SST) system further verify the engineering applicability of the proposed method. Furthermore, these results illustrate the promising potential of the proposed method for application in process control practice. Additionally, the quantitative parameter tuning rules are also applicable to high-order inertia and first-order plus time delay (FOPTD) processes. Note to Practitioners—In this study, an investigation is conducted into linear active disturbance rejection control (LADRC) with improved controllers for high-order plus time-delay (HOPTD) systems and the parameter tuning method. Consequently, the innovations are characterized by their improved structures and quantitative parameter tuning rules. From the perspective of the controllers’ structures, the large inertia and time delay of HOPTD process pose significant challenges for the design of many LADRCs and other types of controllers. In order to overcome the aforementioned problems, based on the existing 5 types of LADRCs for high-order processes, model-aided delay LADRC (MDLADRC) and compound LADRC (CLADRC) are designed. Compared with the existing improved LADRCs, the disturbance observation accuracy has been found to be higher, with a faster system response speed. From the perspective of the quantitative parameter tuning rules, the relationship between the initial value of the open-loop system in the frequency domain and the controllers’ parameters is determined by means of open-loop frequency domain analysis. At the same time, general parameter tuning rules based on the robust stability of the system are proposed to comprehensively solve the parameter tuning problems of the described 7 types of LADRCs. Meanwhile, the maximum sensitivity is adopted to constrain the parameter range of the controllers. Finally, the application of various experimental scenarios are applied to verify the superiority of the two proposed improved LADRCs and the universality of the parameter tuning rules. Notably, the proposed method is compatible with other types of controllers and processes, as well as backwards compatible with high-order inertial and first-order plus time delay (FOPTD) processes. In the future, further investigations should be conducted into the proposed parameter optimisation rules and the low lag filter.}, archive = {J_TASE}, author = {Chang-Liang Liu and Chao-Jie Yang and Wei-Liang Liu and Lu-Yang Zhao and Shuai Liu and Zi-Qi Wang}, doi = {10.1109/TASE.2024.3375931}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2159--2173}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Comprehensive quantitative tuning method for multiple LADRCs in high order plus time delay processes}, volume = {22}, year = {2025}, }
• (2025). Dual control of exploration and exploitation for auto-optimization control with active learning. TASE, 22, 2145--2158. (www)

  @article{ , abstract = {The quest for optimal operation in environments with unknowns and uncertainties is highly desirable but critically challenging across numerous fields. This paper develops a dual control framework for exploration and exploitation (DCEE) to solve an auto-optimization problem in such complex settings. In general, there is a fundamental conflict between tracking an unknown optimal operational condition and parameter identification. The DCEE framework stands out by eliminating the need for additional perturbation signals, a common requirement in existing adaptive control methods. Instead, it inherently incorporates an exploration mechanism, actively probing the uncertain environment to diminish belief uncertainty. An ensemble based multi-estimator approach is developed to learn the environmental parameters and in the meanwhile quantify the estimation uncertainty in real time. The control action is devised with dual effects, which not only minimizes the tracking error between the current state and the believed unknown optimal operational condition but also reduces belief uncertainty by proactively exploring the environment. Formal properties of the proposed DCEE framework like convergence are established. A numerical example is used to validate the effectiveness of the proposed DCEE. Simulation results for maximum power point tracking are provided to further demonstrate the potential of this new framework in real world applications. Note to Practitioners—In numerous engineering applications, it is highly desirable to operate a system to improve the efficiency, enhance performance or save energy. However, attaining this optimal control is a challenging task, due to the presence of unknown system and/or environment parameters. We develop a principled approach to balance between exploration and exploitation, involving active learning to estimate unknown parameters and tracking the optimal operational condition based on current estimation. This paper provides a unified framework to solve general auto-optimization control problems. The simulation results demonstrate that the proposed method outperforms existing methods in terms of efficiency and optimality for maximum power point tracking problem, and it can be readily implemented for many other engineering problems. Future research include generalizing the proposed method to nonlinear systems, as well as exploring novel applications to facilitate the widespread adoption of our method.}, archive = {J_TASE}, author = {Zhongguo Li and Wen-Hua Chen and Jun Yang and Yunda Yan}, doi = {10.1109/TASE.2024.3375373}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2145--2158}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dual control of exploration and exploitation for auto-optimization control with active learning}, volume = {22}, year = {2025}, }
• (2025). A joint ship detection and waterway segmentation method for environment-aware of USVs in canal waterways. TASE, 22, 2132--2144. (www)

  @article{ , abstract = {The canal waterways of China still play an important role in the logistics and transportation industry. Unmanned technology helps to reduce costs and improve the safety of navigation. Real-time environmental awareness is vital to making unmanned surface vehicles (USVs) come true. This paper proposes a new lightweight environmental awareness method based on deep convolutional neural networks (DCNN) and a mixed attention mechanism for USVs in canals, which can simultaneously perform ship detection, segmentation, and surface and background segmentation tasks. The features of the ships, surface, and background are extracted by a shared feature extraction backbone network and hybrid attention mechanism, which improves the efficiency of visual environmental awareness. In addition, a dataset named USV-Canal is constructed to enrich the features of canal waterways for environmental awareness, which contains typical canal scenes and 3443 ship objects. To improve the generalization, multiple public datasets are mixed with the USV-Canal dataset to build an integrated dataset to train our model, which boasts diversity in scene types and ship classes. The comparative and field experiments’ results show that 40.9% of mAP, 95.8% of mIoU, and 5 frames per second (FPS) inference speed can be achieved, and have good generalization, which can meet the requirements of environmental awareness of low-speed ships in canal waterways. Note to Practitioners—The trained and validated model can ultimately be deployed on unmanned surface vehicles, and the required hardware platform is NVIDIA’s Jetson Nano, which is used for real-time perception of surrounding ships and navigable surfaces during navigation. The information can be integrated into the guidance, navigation, and control (GNC) system of USVs, achieving obstacle avoidance and ensuring safe navigation. It is vital to make autonomous navigation come true.}, archive = {J_TASE}, author = {Xiaofei Yang and Hongwei She and Mengmeng Lou and Hui Ye and Jun Guan and Jianzhen Li and Zhengrong Xiang and Hao Shen and Bin Zhang}, doi = {10.1109/TASE.2024.3375300}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2132--2144}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A joint ship detection and waterway segmentation method for environment-aware of USVs in canal waterways}, volume = {22}, year = {2025}, }
• (2025). Graph representation learning and optimization for spherical emission source microscopy system. TASE, 22, 2118--2131. (www)

  @article{ , abstract = {Emission source microscopy (ESM) technique can be utilized for the localization of electromagnetic interference (EMI) sources in electronic systems, its performance greatly depends on the scanner accuracy and back-propagation method. In this paper, we introduce a novel spherical ESM system driven by 6-DOF manipulator, and investigate the back-propagation based on sphere wave expansion and robot control strategy. For spherical scanning aperture, we fuse the robot kinematics model and measurement constraints, and propose solving the optimal scanning grid with nonlinear programming method. For manipulator control, we present a graph-based learning framework (Gash-LKH) that combines sparse graph neural network with Lin-Kernighan heuristic (LKH) solver. This framework adopts the gated single-head attention module and parallel sparse graph feature abstracting channels, it can produce high-qualified edge candidate set that help subsequent LKH solver generate optimal scanning path with lower memory cost and less computation. Extensive experiments are conducted to validate the performance of Gash-LKH and Spherical ESM system, the results have demonstrated the feasibility and superiority of spherical ESM system in providing accurate microscopy and localization in EMI measurement. Note to Practitioners—The motivation of this paper is to develop an automated ESM system that realizes the spherical aperture scanning and pattern reconstruction of radiation source. Since adoption the back-propagation method based on sphere wave expansion, this system is supposed to achieve better microscopy performance and lower truncation error than other scanner. In this paper, we employ 6-DOF manipulator as scanner, and propose a complete spherical aperture generation method that produces the discrete and even scanning grid based on any source, frequency band and measurement constraints. Furthermore, we propose an end-to-end learning framework, Gash-LKH, to solve the optimal scanning path for given scanning aperture. The achieved accuracy and time-consumption of Gash-LKH is satisfactory for solving large-scaled and high-density scanning path planning. Note that the entire framework can be trained through random 3D-TSP instance dataset, and can be transferred to handle various radiation sources operating in microwave band. We have demonstrated the feasibility of proposed methodology and system through the experiments using Elfin-5 manipulator and benchmark sources. The results have offered the possibilities of achieving satisfactory localization and characterization in EMI measurement.}, archive = {J_TASE}, author = {Xiaorui Liu and Zijie Li and Weihua Zong and Hang Su and Peng Liu and Shuzhi Sam Ge}, doi = {10.1109/TASE.2024.3375024}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2118--2131}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Graph representation learning and optimization for spherical emission source microscopy system}, volume = {22}, year = {2025}, }
• (2025). Backstepping-based tracking control of nonlinear systems with concurrent state-triggering mechanism. TASE, 22, 2108--2117. (www)

  @article{ , abstract = {This paper investigates tracking control of nonlinear systems in state-triggering setting. Since the trigger mechanism may result in system states being discontinuous throughout the whole control operation, the controller design process and stability analysis algorithms based on conventional backstepping design are not applicable for such discontinuous systems due to the non-differentiability of the virtual control. To tackle this dilemma, the first-order filter is introduced to smooth the discontinuous virtual control. Nonlinear impulse dynamics technique is adopted to verify the boundedness of the error variables generated by the filter in stability analysis. Furthermore, the adaptive parameters are solely updated at triggering points to alleviate the communication burden. Particularly, all states are required to be updated synchronously with the established concurrent state-triggering mechanism. It is indicated that with the developed backstepping-based state-triggering control scheme, the entire closed-loop signals are bounded, the system output tracks a prescribed reference trajectory and Zeno behaviour is absent. Finally, the simulation study illustrates the applicability of the proposed control scheme. Note to Practitioners—The study of tracking control for nonlinear systems is prevalent in various practical engineering applications, such as the gripping operation of a single-link manipulator, the orbiting of a spacecraft, and the cruising process of an unmanned aerial vehicle. Moreover, with the rapid development of networking, communication resources are often limited. Continuous information transmission in the sensor-to-controller channel can generate communication blockage and degrade transmission efficiency. Therefore, it is crucial to utilize the limited bandwidth resources to solve the tracking control problem of nonlinear systems. In this paper, we propose a backstepping-based state-triggering control scheme, where the triggered states and triggered adaptive parameters are used for the whole control design. The proposed scheme accomplishes the tracking control task well, especially in applications with limited bandwidth and energy. It is more consistent with requirements of the actual engineering.}, archive = {J_TASE}, author = {Xu Yuan and Rencheng Jin and Bin Yang and Xudong Zhao}, doi = {10.1109/TASE.2024.3374960}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2108--2117}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Backstepping-based tracking control of nonlinear systems with concurrent state-triggering mechanism}, volume = {22}, year = {2025}, }
• (2025). 3D model-free visual localization system from essential matrix under local planar motion. TASE, 22, 2090--2107. (www)

  @article{ , abstract = {Visual localization plays a critical role in the functionality of low-cost autonomous mobile robots. Contemporary leading methods for precise visual localization are predominantly 3D scene-specific, necessitating extra computational and memory overhead to construct a 3D scene model in novel environments. An alternative approach of directly using a database of 2D images for visual localization offers more flexibility. However, such methods currently suffer from limited localization accuracy. In this paper, we propose an accurate and robust multiple checking-based 3D model-free visual localization system to address the aforementioned issues. To ensure high accuracy, our focus is on estimating the pose of a query image relative to the retrieved database images using 2D-2D feature matches. Theoretically, by incorporating the local planar motion constraint into both the estimation of the essential matrix and the triangulation stages, we reduce the minimum required feature matches for absolute pose estimation, thereby enhancing the robustness of outlier rejection. Additionally, we introduce a multiple-checking mechanism to ensure the correctness of the solution throughout the solving process. The efficacy of our approach is substantiated through both qualitative and quantitative assessments on simulated and two real-world datasets evidencing significant improvements in accuracy and robustness provided by our 3D model-free visual localization system. Note to Practitioners—The motivation of this article stems from the need to develop an accurate visual localization system with simplicity and flexibility of map construction and easy adaption to new environments. Such a system holds great practical value for a range of applications, including warehouse robots, service robots, and countless others. Existing visual localization systems that achieve high accuracy are dependent on a pre-built accurate 3D scene map, which pose challenges in terms of map construction and consume significant storage resources onboard, particularly for large scenes. And the aforementioned efforts need to be repeated when changing to a new scene. In this article, an accurate and robust 3D model-free visual localization system is proposed to handle this problem. The map construction is simplified to build a set of database images with associated camera poses, which is trivial as it amounts to adding posed images to a database. The core idea for achieving high accuracy and robustness is to model the local planar motion characteristic of general ground-moving robots into both essential matrix estimation and triangulation stages to obtain two minimal solutions. The proposed localization system simplifies the task of switching between different application scenarios for the robot, reducing additional workload and lowering the difficulty of use.}, archive = {J_TASE}, author = {Yanmei Jiao and Binxin Zhang and Peng Jiang and Chaoqun Wang and Haojian Lu and Rong Xiong and Yue Wang}, doi = {10.1109/TASE.2024.3374823}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2090--2107}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {3D model-free visual localization system from essential matrix under local planar motion}, volume = {22}, year = {2025}, }
• (2025). Multi self-supervised pre-finetuned transformer fusion for better vehicle detection. TASE, 22, 2075--2089. (www)

  @article{ , abstract = {Vehicle detection is an important task in intelligent transportation, which can provides services for road condition monitoring and highway toll collection systems. However existing detection methods are limited by two aspects. First, there is a difference between the model knowledge pre-trained on large-scale datasets and the knowledge required for target task. Second, most detection models follow the pattern of single-source learning, which limits the learning ability. To address these problems, we propose a Multi Self-supervised Pre-finetuned Transformer Fusion (MSPTF) network, consisting of two steps: self-supervised pre-finetuning domain knowledge learning and multi-model fusion target task learning. In the first step, we introduced self-supervised learning methods into transformer pre-finetuning to reduce data costs and alleviate knowledge gap. In the second step, we take feature information differences between different model architectures and pre-finetuning tasks into account and propose Multi-model Semantic Consistency Cross-attention Fusion (MSCCF) network to combine different transformer features by considering channel semantic consistency and feature vector semantic consistency, which obtain more complete and proper fusion features for detection task. We experimented the proposed method on two vehicle detection datasets and achieved 1.1%, 5.5% improvement compared with baseline and 0.7%, 1.8% compared with sota, which proved the effectiveness of our method. Note to Practitioners—Vehicle detection can provide basic services for many intelligent transportation scenarios, but the results of training and adjusting existing models in vehicle detection scenarios with limited data are often not ideal. This article proposes MSPTF to improve model performance from two aspects. First, we randomly collect additional vehicle pictures, and train the model without the need for annotation through annotation-free methods such as image feature extraction and reconstruction, so that the model can be trained on more vehicle-related pictures with low data cost. In addition, we trained multiple models in the above way and proposed a fusion method to combine the knowledge of multiple models for vehicle detection. The results show that our Our model is able to learn from additional unlabeled vehicle images as well as multiple trained models information to achieve better vehicle detection.}, archive = {J_TASE}, author = {Juwu Zheng and Jiangtao Ren}, doi = {10.1109/TASE.2024.3374759}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2075--2089}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi self-supervised pre-finetuned transformer fusion for better vehicle detection}, volume = {22}, year = {2025}, }
• (2025). A novel guided deep reinforcement learning tracking control strategy for multirotors. TASE, 22, 2062--2074. (www)

  @article{ , abstract = {This paper presents an intelligent control scheme for multirotors, where accurate trajectory tracking, strong robustness and reliable generalization are guaranteed by the dual-feedback sliding-mode (DFSM) guided deep reinforcement learning (RL). Different from current solutions, the proposed method explores optimal learning strategy on the sliding surface according to the DFSM demonstrations, where the elegantly designed parallel evaluation takes full advantage of model knowledge and learning exploration. Specifically, the intelligent tracking control is achieved in a two-step design. First, the DFSM algorithm is designed for multirotors, where the linear and nonlinear feedback terms work cooperatively. Second, the DFSM-guided deep RL is put forward to achieve intelligent switching on the sliding surface, where position and velocity errors are both considered to generate accurate switching decisions. In the framework, explorations and the DFSM demonstrations are evaluated in parallel, where only the explorations that are better than the DFSM baseline, are kept for policy improvement. In this way, the DFSM algorithm keeps pushing the RL policy to explore better strategy, where the unavoidable bad experiences arisen from exploration are identified accurately. Practical comparative experimental results are included to verify the effectiveness of the proposed strategy. Note to Practitioners—This paper is motivated by the practical problem of controlling multirotor system in uncertain environments. Up until now, most existing approaches are proposed without taking full advantage of model knowledge and deep learning techniques simultaneously, which lacks of reliability in practical application. To deal with the problem, a new dual-feedback sliding-mode (DFSM) guided deep reinforcement learning (RL) strategy is proposed, where the dual feedback and guided RL are designed to achieve satisfactory tracking control and simultaneously handle uncertainties. Specifically, by introducing double-check framework, the RL strategy explores optimal switching policy on the sliding surface according to the DFSM demonstrations, guaranteeing strong robustness and reliable generalization of the obtained RL policy in uncertain environments. The key feature of the framework is that the DFSM-driven training guarantees practice-oriented tracking control in a DFSM-RL cooperative manner. Comparative experiments are implemented to verify the tracking performance of the proposed intelligent control strategy.}, archive = {J_TASE}, author = {Hean Hua and Yaonan Wang and Hang Zhong and Hui Zhang and Yongchun Fang}, doi = {10.1109/TASE.2024.3374752}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2062--2074}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel guided deep reinforcement learning tracking control strategy for multirotors}, volume = {22}, year = {2025}, }
• (2025). Quasi-synchronization of heterogeneous hybrid stochastic delayed networks via pinning intermittent discrete observation control. TASE, 22, 2051--2061. (www)

  @article{ , abstract = {This paper focuses on the quasi-synchronization of heterogeneous hybrid stochastic delayed networks (HHSDNs), in which pinning aperiodically intermittent discrete observation control (PAIDOC) strategy is imposed on the networks. During control intervals, PAIDOC is based on discrete-time state observations and is implemented to a fraction of the network nodes, which is an extension of existing literature. In light of stochastic analysis theory and Lyapunov method, we establish improved quasi-synchronization criteria. Compared with previous quasi-synchronization results on HHSDNs, the piecewise continuous, time-varying, and indefinite functions replace the constant coefficient of the upper bound estimation for the operator of a Lyapunov function in the available literature, i.e., it can be positive or negative along time evolution, which shows wider applications of our results. Especially, our results can be directly applied to examine the quasi-synchronization problem via PAIDOC based on the average control rate. Subsequently, some easily-verified quasi-synchronization criteria are presented and a detailed algorithm about how to design suitable PAIDOC is provided. In the end, the theoretical result is applied to spring-mass-damper systems and numerical examples are presented. Note to Practitioners—This paper was motivated by existing results on pinning aperiodically intermittent control about the quasi-synchronization. The existing results mainly require the control method to be continuously sampled and control all nodes of the network, which is difficult to be implemented in practical application. This paper designs a novel hybrid control method named pinning aperiodically intermittent discrete observation control. In addition, we established a lower conservative differential inequality first, which is significantly conducive to analyzing the quasi-synchronization of complex networks. The results obtained are applied on spring-mass-damper systems, demonstrating their effectiveness and it is expected that the proposed approach can be extended to more practical physical engineering systems.}, archive = {J_TASE}, author = {Dongsheng Xu and Yang Zhang and Wenxue Li}, doi = {10.1109/TASE.2024.3374721}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2051--2061}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Quasi-synchronization of heterogeneous hybrid stochastic delayed networks via pinning intermittent discrete observation control}, volume = {22}, year = {2025}, }
• (2025). Dynamic event-triggered control for a class of uncertain strict-feedback systems via an improved adaptive neural networks backstepping approach. TASE, 22, 2041--2050. (www)

  @article{ , abstract = {This article focuses on a dynamic event-triggered adaptive neural networks backstepping control for a class of uncertain strict-feedback systems with communication constraints. The uncertain terms including external disturbances and unknown nonlinear functions are approximated by radial basis function neural networks, in which the weight update laws are obtained via the gradient descent algorithm, ensuring the local boundedness of the approximation error of neural networks. Then, to enhance the transmission efficiency of control signals, a dynamic event-triggered mechanism is introduced, which enables the dynamic adjustment of threshold parameters in response to the actual tracking performance. It is strictly proved via the Lyapunov stability criterion that the tracking error can converge to a desired small neighborhood of the origin, and all signals in the closed-loop system are bounded. Finally, the validity of the control strategy is demonstrated through a simulation example. Note to Practitioners— In practical network control systems, control signals are typically transmitted continuously or periodically to devices through the communication network in the form of data packets. As communication networks are usually shared by various system nodes, and resources such as communication channel bandwidth and computational capabilities are limited, improving the transmission efficiency of control signals becomes a crucial design problem for controllers in network control systems. Therefore, This study introduces a control method via event-triggered sampling, aiming to enhance sampling efficiency while ensuring the stability and reliability of the system. The proposed control method is suitable for a broad category of strict-feedback nonlinear systems with communication constraints, offering notable advantages such as low-complexity design and straightforward implementation.}, archive = {J_TASE}, author = {Ning Xu and Xiang Liu and Yulin Li and Guangdeng Zong and Xudong Zhao and Huanqing Wang}, doi = {10.1109/TASE.2024.3374522}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2041--2050}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic event-triggered control for a class of uncertain strict-feedback systems via an improved adaptive neural networks backstepping approach}, volume = {22}, year = {2025}, }
• (2025). RDNet-KD: Recursive encoder, bimodal screening fusion, and knowledge distillation network for rail defect detection. TASE, 22, 2031--2040. (www)

  @article{ , abstract = {Rail defect detection (RDD) plays a crucial role in ensuring rail transportation safety. Recently, bimodal algorithms have become mainstream; however, the asymmetry in the information of RGB and depth makes it difficult to find a suitable bimodal information fusion algorithm. In addition, it is difficult to deploy most of the existing methods on mobile devices. To solve these problems, we propose a recursive encoder and bimodal information screening fusion with a knowledge distillation network (RDNet-KD) for RDD. First, we propose the recursive encoder-based depth information augmentation (REDA) algorithm. It recursively learns to expand the channel depth information to alleviate the quality problem of depth information. Second, we propose a similarity-driven bimodal information screening fusion (SICF) module. This evaluates the complementarity of information from two modalities by computing the similarity of their hierarchical feature maps to screen useful information for fusion. Third, we introduce the global location and interrelation-based dual contextual knowledge distillation method to enhance the performance of the compact model. Therefore, it is possible to deploy the network on mobile devices. Based on the extensive experiments performed on the RGB-D rail defect dataset NEU RSDDS-AUG, we validate the competitiveness of our RDNet-KD, considering the prediction quality and operational efficiency relative to 12 state-of-the-art methods. The RDNet-KD code and results are available at https://github.com/legendfantasy/RDNet-KD. Note to Practitioners—This study introduces a recursive encoder and bimodal information screening fusion with a knowledge distillation network (RDNet-KD) for RDD in RGB-D images. Our method enhances depth information quality and effectively selects valuable information from both modalities using the similarity as a coefficient to evaluate the complementary capabilities of the modal information. Furthermore, to compress the model, we introduce knowledge distillation (KD) to balance the number of parameters and detection results and propose a novel KD method that transfers knowledge from the teacher network to the student network.}, archive = {J_TASE}, author = {Wujie Zhou and Jinxin Yang and Weiqing Yan and Meixin Fang}, doi = {10.1109/TASE.2024.3374387}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2031--2040}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {RDNet-KD: Recursive encoder, bimodal screening fusion, and knowledge distillation network for rail defect detection}, volume = {22}, year = {2025}, }
• (2025). Adaptive tracking control of high-order nonlinear systems with time-varying delays under asymmetric output constraints. TASE, 22, 2020--2030. (www)

  @article{ , abstract = {This paper addresses the problem of adaptive tracking control of high-order nonlinear systems with time-varying delays under asymmetric output constraints. Unlike the existing researches, the studied system exhibits constantly changing unknown powers, which makes the existing Lyapunov Functionals invalid. Therefore, the Lyapunov Functionals are carefully designed to be suitable for the changing powers, where the changing powers include low power ( $0 < p_{i}\le 1$ ) and high power ( $p_{i} > 1$ ). Besides, we also incorporate the well-constructed dynamic gain signals into the Lyapunov Functionals to enhance the robustness, accuracy, and response speed of the system. Further, we cleverly deal with the time-varying delays by combining the novel dynamic gain signals and Lyapunov-Krasovskii (L-K) Functionals in the controller design process. For the system performance requirement, the asymmetric output constraints are considered by introducing a nonlinear transformation for the output signal $x_{1}$ and the reference signal $y_{r}$ . While the derivative of this nonlinear transformation generates a gain term in the form of function, which makes the controller design more difficult. We utilize the adding one power integrator technique to design the desired controller and solve the design difficulty caused by the gain term. Under the proposed controller, all the signals in the closed-loop system remain bounded, the asymmetric output constraints are not violated as well as the tracking error stays in a small neighborhood of the origin. Finally, a simulation example is given to demonstrate the effectiveness of the present strategy. Note to Practitioners—In the article, we consider a class of high-order nonlinear systems with time-varying delays under asymmetric output constraints, and their models are widely utilized in the engineering field, such as transportation, manufacturing sector and so on. Further, the tracking problem is regarded as a popular topic in the control field, so designing a desired tracking controller for high-order nonlinear systems is a significant research topic. Since the considered system has the changing powers and time-varying delays, the existing methods cannot be applied to the studied problem, so we provide some new strategies by the introduction of the well-constructed dynamic gain signals, the improved L-K functionals and the nonlinear transformation. From what has been mentioned above, an effective adaptive tracking control strategy is developed for the studied system, where the proposed strategy has great application prospect.}, archive = {J_TASE}, author = {Yi Niu and Yulong Yang and Huanqing Wang and Ben Niu and Zihao Shang}, doi = {10.1109/TASE.2024.3374242}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2020--2030}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive tracking control of high-order nonlinear systems with time-varying delays under asymmetric output constraints}, volume = {22}, year = {2025}, }
• (2025). Physics-informed weakly-supervised learning for quality prediction of manufacturing processes. TASE, 22, 2006--2019. (www)

  @article{ , abstract = {In manufacturing processes, a multitude of sensors are typically deployed to collect data of process parameters. While this provides an opportunity to better predict and control the quality of the final product, the relationship between the process variables and the desired final quality is often not well-understood. To establish this relationship, machine learning (ML) models can be used. However, collecting large, labeled datasets to train the ML model can be difficult, as creating such datasets typically involves costly or destructive end-of-line quality testing of the products. To overcome this challenge, we propose a novel framework called Physics-informed Weakly-supervised Learning (PWL) that integrates physics-based models with data-driven ML models. By leveraging physical knowledge and using the outputs of physics-based models as weak labels, PWL offers an alternative to traditional methods that require large, labeled datasets. Our approach simultaneously optimizes the data-driven ML model, as well as the discrepancy and calibration parameters of the physics-based model, resulting in superior predictive performance compared to either model used alone. We demonstrate the effectiveness of PWL through simulation experiments, comparisons with existing methods, and two real-world case studies, highlighting its potential for improving quality prediction in various manufacturing systems. Note to Practitioners—The proposed method in this paper, Physics-informed Weakly-supervised Learning (PWL), addresses a common problem in manufacturing processes, where obtaining large, labeled datasets can be costly or not always possible. By integrating physics-based models with data-driven ML models, PWL leverages available physical knowledge to improve the predictive performance of product quality in manufacturing systems. This enables practitioners to better understand and optimize their manufacturing processes even with limited labeled datasets, leading to improved product quality and reduced production costs.}, archive = {J_TASE}, author = {Dhari F. Alenezi and Michael Biehler and Jianjun Shi and Jing Li}, doi = {10.1109/TASE.2024.3374098}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {2006--2019}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Physics-informed weakly-supervised learning for quality prediction of manufacturing processes}, volume = {22}, year = {2025}, }
• (2025). High-speed obstacle avoidance of a large-scale underactuated autonomous underwater vehicle under a finite field of view. TASE, 22, 1996--2005. (www)

  @article{ , abstract = {This paper addresses the problem of high-speed waypoint tracking and real-time obstacle avoidance for large-scale underactuated autonomous underwater vehicles (AUVs) in the vertical plane. Specifically, a robust nonlinear model predictive control (RNMPC) scheme is proposed, considering different types of constraints including the scale of the AUV, the finite field of view of the sensor, the input saturation, the physical limits on system state, and the influence of the vertical underactuated velocity. To navigate in the completely unknown environment with nonconvex obstacles, a dynamic sensing and collision avoidance scheme is proposed so that the collision avoidance can be properly formulated into convex constraints in the RNMPC optimization problem. Recursive feasibility and closed-loop stability are proved rigorously. Through the high-fidelity simulations with graph and data visualization techniques, the proposed algorithm has higher waypoint tracking accuracy, safer obstacle avoidance ability, and better multiple constraints handling capability than the existing dynamic virtual AUV (DVA) technique. Note to Practitioners—This article was motivated by the problem of high-speed waypoint tracking and obstacle avoidance for large-scale underactuated autonomous underwater vehicles (AUVs) in an unknown environment. In practical engineering, the obstacles’ useful information (e.g., shapes and positions) cannot be directly reflected by the raw data (distance or acoustic image) of multi-beam sonars. And the multi-beam sonars are also limited by detection range and angle. Another practical problem is that such type of AUV suffers from different constraints including the scale of the vehicle, the physical limits on system poses and velocities. In addition, the influence of the underactuated velocity cannot be ignored due to the high speed of the vehicle. Considering the above practical engineering problems, we design a robust nonlinear model predictive control (RNMPC) scheme from the kinematic level of the AUV to achieve high-speed waypoint tracking and obstacle avoidance. And then we developed a dynamic sensing and collision avoidance scheme to formulate nonconvex collision avoidance into convex constraints in the RNMPC optimization problem. We demonstrate the effectiveness of the proposed control method in a high-fidelity simulation environment with large size terrain and shipwrecks as obstacles. In this environment, the AUV can dynamically read the distance data from obstacles to the sensor through the virtual multi-beam sonar equipped in the head of the vehicle, which can simulate the actual engineering experiment with high-fidelity. This work could be applied to other underactuated systems, such as unmanned aerial vehicles, unmanned surface vehicles, etc. In future research, we will consider more complex practical working cases, such as the presence of dynamic obstacles in an unknown environment and the disturbance of current.}, archive = {J_TASE}, author = {Lin Yu and Lei Qiao and Chao Shen}, doi = {10.1109/TASE.2024.3373607}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1996--2005}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {High-speed obstacle avoidance of a large-scale underactuated autonomous underwater vehicle under a finite field of view}, volume = {22}, year = {2025}, }
• (2025). Observer-based nonlinear control for dual-arm aerial manipulator systems suffering from uncertain center of mass. TASE, 22, 1984--1995. (www)

  @article{ , abstract = {The unmanned aerial manipulator system has shown great application potential in rotor blade repairing, bridge inspection, and goods delivery. Allowing additional alternatives in tasks, dual-arm always provides more flexibility, versatility, and manipulability compared to a single arm. Unfortunately, the inherent defects of unignorable nonlinearities and complex dynamic coupling between the multirotor and the manipulator have limited the practical application of dual-arm aerial manipulator systems. It is noteworthy that the dynamic coupling between the multirotor UAV (unmanned aerial vehicle) and the dual-arm manipulator is much more complicated than the case of the single-arm, and may degrade the control performance significantly as the CoM (center of mass) of the system changes with the movement of the manipulator. To this end, this paper presents a novel control method based on dual-arm movement compensation. Specifically, the kinematic and dynamic model of the system is first established, based on which the force effect of the manipulator exerting on the multirotor UAV is estimated by the disturbance observer and then compensated. By using Lyapunov techniques, it is proven that the error signal can converge asymptotically. As far as we know, this paper presents the first controller design for dual-arm aerial manipulator systems with rigorous stability analysis. Finally, the effectiveness and robustness of the proposed method are verified through a significant number of comparison experiments, and the results of these experiments demonstrate that the proposed method can reduce the positioning error obviously compared to the comparison methods. Taking the PID method as the benchmark for comparison, it is obvious that the proposed method exhibits the greatest reduction in both maximum and average errors compared to the baseline method, indicating superior control precision than the other comparison methods. For the result of the proposed method in $x$ -direction, one can find a substantial reduction ranging from 16.69% to 38.57% for the maximum error, and 22.24% to 45.66% for the mean error. Shifting focus to the $y$ -direction, the error reduction for the proposed method is even more remarkable, ranging from 68.10% to 81.80% at maximum, and 66.31% to 86.33% for the mean error. As for the $z$ -direction, the error reduction by the proposed method remained significant, ranging from 50.67% to 86.38% at maximum, and with a mean error reduction of 33.11% to 80.39%. Note to Practitioners—This paper is motivated by the problem of executing such tasks as load transportation and coordinate manipulation for aerial robots in flight. By integrating the dual-arm manipulator, the flexibility, versatility, and manipulability of the unmanned aerial manipulator system is further extended. However, the uncertain center of mass of the system during operation may badly increase the control difficulty of the dual-arm aerial manipulator system. Moreover, the strong nonlinearity and complex coupling existing between the multirotor and the manipulator also induce urgently solved problems in practical aerial manipulation tasks. To this end, this paper proposes a novel dual-arm movement compensation based control scheme by utilizing an elaborately designed disturbance observer to deal with the unestimated part of disturbance exerting on the multirotor by arm operation. With rigorous theoretical analysis, the convergence of the error signal is proven. Additionally, groups of hardware experiments further verify the effectiveness and robustness of the suggested control method. In future studies, we will improve the autonomy level of the system by integrating onboard sensors.}, archive = {J_TASE}, author = {Xiao Liang and Yang Wang and Hai Yu and Zhaopeng Zhang and Jianda Han and Yongchun Fang}, doi = {10.1109/TASE.2024.3373107}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1984--1995}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based nonlinear control for dual-arm aerial manipulator systems suffering from uncertain center of mass}, volume = {22}, year = {2025}, }
• (2025). Adaptive terminal sliding mode control for high-speed EMU: A MIMO data-driven approach. TASE, 22, 1970--1983. (www)

  @article{ , abstract = {In this study, a novel data-driven discrete-time sliding mode control (DSMC) approach is designed for an electric multiple unit (EMU) velocity tracking control system. First, the input/output (I/O) data of the EMU are equivalently modeled as a full-format dynamic linearized (FFDL) data model to facilitate the generation of a data-driven control scheme. Subsequently, a discrete terminal sliding mode function and a new hyperbolic reaching-law are introduced to simultaneously achieve fast convergence and alleviate chattering. Based on the designed sliding mode function and reaching-law, an improved discrete-time terminal sliding mode control (iDTSMC) approach is derived using the FFDL model. The proposed approach considers the error feedback, parameter estimation errors, and total uncertainties for compensation, to achieve better control performance. The key advantages of this approach include its sole utilization of input-output data from the EMU system, low controller order, robust parameter adaptability, and anti-interference capabilities. After providing the stability analysis of the proposed method, the iDTSMC scheme is compared and tested on a simulated CRH380A high-speed train experimental platform in a laboratory. The simulation results show that the velocity tracking errors of each power unit of the EMU under the proposed control scheme are within [−0.112 km/h, 0.118 km/h], and the control forces and accelerations are within [−51 kN, 43 kN] and [−0.952 m/s2, 0.827 m/s2], respectively, with stable fluctuations. Comparative experimental results demonstrate the effectiveness and superiority of the proposed strategy, which remains robust in the presence of disturbances.Note to Practitioners—This study is inspired by the problem of EMU operation control, however, it is also applicable to other systems with trajectory tracking characteristics. Existing automatic train driving control methods are usually based on a dynamics model of the train, which is easily affected by the external environment. In this study, a new data-driven control method is proposed. A discrete terminal sliding mode control function and a new hyperbolic reaching-law were designed based on a dynamic linearized data model equivalent to the EMU operation process. In contrast to similar existing works, the control scheme does not require an accurate EMU dynamic model and is easy to implement. Preliminary simulation experiments show that the method is feasible, however, it has not been incorporated into the train’s physical system, nor has it been tested in production. In future research, we plan to promote a combination of control theory and manufacturing practices.}, archive = {J_TASE}, author = {Liang Zhou and Zhongqi Li and Hui Yang and Chang Tan and Yating Fu}, doi = {10.1109/TASE.2024.3373037}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1970--1983}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive terminal sliding mode control for high-speed EMU: A MIMO data-driven approach}, volume = {22}, year = {2025}, }
• (2025). Wind turbine fault-tolerant control via incremental model-based reinforcement learning. TASE, 22, 1958--1969. (www)

  @article{ , abstract = {A reinforcement learning (RL) based fault-tolerant control strategy is developed in this paper for wind turbine torque & pitch control under actuator & sensor faults subject to unknown system models. An incremental model-based heuristic dynamic programming (IHDP) approach, along with a critic-actor structure, is designed to enable fault-tolerance capability and achieve optimal control. Particularly, an incremental model is embedded in the critic-actor structure to quickly learn the potential system changes, such as faults, in real-time. Different from the current IHDP methods that need the intensive evaluation of the state and input matrices, only the input matrix of the incremental model is dynamically evaluated and updated by an online recursive least square estimation procedure in our proposed method. Such a design significantly enhances the online model evaluation efficiency and control performance, especially under faulty conditions. In addition, a value function and a target critic network are incorporated into the main critic-actor structure to improve our method’s learning effectiveness. Case studies for wind turbines under various working conditions are conducted based on the fatigue, aerodynamics, structures, and turbulence (FAST) simulator to demonstrate the proposed method’s solid fault-tolerance capability and adaptability. Note to Practitioners—This work achieves high-performance wind turbine control under unknown actuator & sensor faults. Such a task is still an open problem due to the complexity of turbine dynamics and potential uncertainties in practical situations. A novel data-driven and model-free control strategy based on reinforcement learning is proposed to handle these issues. The designed method can quickly capture the potential changes in the system and adjust its control policy in real-time, rendering strong adaptability and fault-tolerant abilities. It provides data-driven innovations for complex operational tasks of wind turbines and demonstrates the feasibility of applying reinforcement learning to handle fault-tolerant control problems. The proposed method has a generic structure and has the potential to be implemented in other renewable energy systems.}, archive = {J_TASE}, author = {Jingjie Xie and Hongyang Dong and Xiaowei Zhao and Shuyue Lin}, doi = {10.1109/TASE.2024.3372713}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1958--1969}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Wind turbine fault-tolerant control via incremental model-based reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Distribution-agnostic probabilistic few-shot learning for multimodal recognition and prediction. TASE, 22, 1941--1957. (www)

  @article{ , abstract = {In industrial scenarios with insufficient sensor data, intelligent few-shot failure mode recognition and remaining useful lifetime (RUL) prediction are critically essential for effective prognostics and health management. Existing few-shot learning (FSL) methods focus on either the failure mode recognition as a classification problem or the RUL prediction as a regression problem, failing to capture the dependence between failure modes and RUL given that units under different failure modes present distinct degradation characteristics. To address the issue, this paper proposes a distribution-agnostic probabilistic FSL method for multimodal recognition and prediction of operating units. The proposed model establishes a neural network with prototypes to solve a few-shot classification- and regression-integrated problem. To fully capture the uncertainty caused by limited sensor data, we develop multimodal Bayesian model-agnostic meta-learning (MBMAML) for the probabilistic modeling of failure modes and the RUL under multiple failure modes. We construct the loss function based on probabilistic modeling that captures the interaction between failure modes and RUL for model training. Finally, the proposed model adaptively learns the approximate distributions of failure modes and RUL for a new operating unit. We evaluate the proposed model performance through a case study on the degradation of aircraft gas turbine engines. Note to Practitioners—Failure mode recognition and RUL prediction are essential in prognostics health management (PHM) to avoid unexpected failures of units in industrial systems, such as aircraft gas turbine engines. However, insufficient sensor data are quite common issue in industrial scenarios due to expensive sensor deployment, the difficulty of installing sensors to certain special mechanical equipment, and so on. This paper aims to develop a FSL method to jointly recognize the failure mode and predict the RUL of a unit based on insufficient sensor data. The four steps to implement the proposed method in practice are as follows: First, collect sensor signal data, RUL data, and failure mode data of units. Second, construct the model framework via the proposed MBMAML. Third, formulate the loss function based on the probability distributions of failure modes and RUL, and train the model using collected data. Fourth, adaptively recognize the failure mode and predict the RUL of a new operating unit. The proposed method is expected to be applicable to many practical few-shot industrial scenarios due to its data-driven neural network with flexible model structure.}, archive = {J_TASE}, author = {Di Wang and Xiaochen Xian and Haidong Li and Dong Wang}, doi = {10.1109/TASE.2024.3372711}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1941--1957}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distribution-agnostic probabilistic few-shot learning for multimodal recognition and prediction}, volume = {22}, year = {2025}, }
• (2025). A joint fault-tolerant and fault diagnosis strategy for multiple actuator faults of full-vehicle active suspension systems. TASE, 22, 1928--1940. (www)

  @article{ , abstract = {In this paper, a joint fault-tolerant and fault diagnosis strategy is proposed for handling multiple actuator faults in full-vehicle active suspension systems. Different from traditional methods where fault detection and isolation must precede the fault-tolerant control to provide the latter with certain fault information, our proposed scheme performs both jointly over the whole operation process with guaranteed suspension performance. For the fault-tolerant control, we develop high gain filters under a fast timescale to estimate the variations of the integrated control inputs caused by actuator faults, and the equivalent fault-free parts of the integrated control inputs are selected as the target control variables. In such a decomposition, we are able to determine the control law of each actuator without using the information of the faulty actuators. Under the framework of the fault-tolerant algorithm, we construct a bank of adaptive fault diagnosis observers to online identify the force constants of the actuators, where information of the heave, pitch, and roll motions are all utilized to ensure persistent excitation and enhance identification accuracy. In particular, actuator faults can be detected via the estimation of the force constants, in which both the location and severity of the actuator faults can be further identified. Subsequently, control commands are reassigned to alleviate the use of faulty actuators and thus protect them from further damage. The effectiveness of our proposed method is validated via extensive simulation results. Note to Practitioners—Satisfactory performance of vehicle active suspension systems is founded upon good reliability and safety of the suspension systems, which motivates us to conduct deep research into fault diagnosis and fault-tolerant problems to improve the significant system reliability. In this paper, a joint fault-tolerant and fault diagnosis (joint FTFD) scheme is proposed for addressing multiple actuator faults of full-vehicle active suspension systems. The proposed method can be used to guarantee suspension performance over the whole operation period even if multiple actuator faults occur, and the location and severity of the actuator faults can be identified at the same time. Different from the traditional methods where fault diagnosis precedes the fault-tolerant control to provide certain fault information, in the proposed joint FTFD scheme, fault diagnosis is carried out under the always-on fault-tolerant framework to ensure system reliability and performance over the whole operation process. The multi-timescale technique is used to compensate for actuator failure and achieve fault-tolerant control. Under the fault-tolerant framework, actuator faults can be detected and both the location and severity of the faults are decided via monitoring the online identification results of the motor force constants. However, one limitation of the proposed method is the influence of sensor noise on high gain filters in real world application, which might degrade fault-tolerant performance. For the future research, one significant work is to carry out real vehicle tests to promote the proposed algorithm to practical application. Also, fault diagnosis and fault-tolerant control for suspension sensor faults is another valuable topic which deserves our efforts. Given that the proposed fault-tolerant framework is independent on a particular control system or a specific type of nominal controller, the fault-tolerant design idea can be transplanted to a broad range of systems besides addressing suspension actuator faults.}, archive = {J_TASE}, author = {Shuai Yan and Weichao Sun and Yuanqing Xia}, doi = {10.1109/TASE.2024.3372626}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1928--1940}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A joint fault-tolerant and fault diagnosis strategy for multiple actuator faults of full-vehicle active suspension systems}, volume = {22}, year = {2025}, }
• (2025). Intelligent marine survey: Lightweight multi-scale attention adaptive segmentation framework for underwater target detection of AUV. TASE, 22, 1913--1927. (www)

  @article{ , abstract = {Accurate and automatic underwater target recognition is a compelling challenge for autonomous underwater vehicles (AUVs) in intelligent marine surveys. This study proposed a seabed target correction model based on side-scan sonar (SSS) images and combined the navigation information of AUV to achieve pixel-level geocoding. Moreover, a lightweight multi-level attention adaptive segmentation framework ( ${\mathrm{M}}{{\mathrm{A}}^{\mathrm{2}}}{\mathrm{Net}}$ ) was proposed to achieve fine-grained recognition. It contains three new modules: 1) The lightweight attention network (LAN) is designed as the baseline to obtain dense feature maps and focus on interesting features based on a balanced attention mechanism. 2) the multi-scale feature pyramid (MASPP) was then constructed to capture the context of SSS images and extract rich semantic information at high levels. 3) Finally, the adaptive feature fusion module (AFF) effectively incorporates feature maps of MASPP and spatial information to improve the learned representations further. Extensive experiments are verified on six SSS categories and show the remarkable performance of the ${\mathrm{M}}{{\mathrm{A}}^{\mathrm{2}}}{\mathrm{Net}}$ compared with state-of-the-art methods. Furthermore, real sea trials were conducted by deploying ${\mathrm{M}}{{\mathrm{A}}^{\mathrm{2}}}{\mathrm{Net}}$ to the autonomous target recognition (ATR) system of AUV, which can achieve 29.7 fps and 81.23% MIoU for a ( $512\times 512$ ) input on a single Nvidia Jetson Xavier. Note to Practitioners—This paper aims to provide a real-time semantic segmentation model for the autonomous target detection of AUV, which is suitable for the autonomous detection of underwater targets by underwater robots (ROV, AUV, ARV, et al). This paper proposes a lightweight, multi-scale attention-adaptive segmentation framework ( ${\mathrm{M}}{{\mathrm{A}}^{\mathrm{2}}}{\mathrm{Net}}$ ) incorporating pixel-level seabed targets rectification methods. The algorithm has high segmentation accuracy and fast operation speed. It can identify seabed targets in high-resolution sonar images online and realize precise positioning of small seabed targets, which is conducive to improving the intelligence level of marine survey unmanned equipment. This paper details the design of ${\mathrm{M}}{{\mathrm{A}}^{\mathrm{2}}}{\mathrm{Net}}$ and the hardware structure of the autonomous target recognition system (ATR). Plenty of simulation experiments and sea trials have proved the efficiency and practicability of the method for the autonomous detection of different seabed targets (sand waves, coral reefs, metal balls, threads, and artificial reefs). Future research will verify the generalization of the algorithm in more seabed targets.}, archive = {J_TASE}, author = {Qi Wang and Yixiao Zhang and Bo He}, doi = {10.1109/TASE.2024.3371963}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1913--1927}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intelligent marine survey: Lightweight multi-scale attention adaptive segmentation framework for underwater target detection of AUV}, volume = {22}, year = {2025}, }
• (2025). TCRNet: Transparent object depth completion with cascade refinements. TASE, 22, 1893--1912. (www)

  @article{ , abstract = {Transparent objects are commonly found in real life and industrial production. Unlike opaque objects, transparent objects are not easily identifiable in RGB images and often require depth information to determine their position in the image. However, due to the influence of other environmental factors such as reflection and refraction, the depth information of transparent objects is often inaccurate. This leads to difficulties for robots in grasping transparent objects, as incorrect depth information can result in the robot being unable to predict or predict incorrectly the grasping pose. Therefore, it is necessary to complete the depth information for transparent objects. Previous methods for depth completion of transparent objects often struggle to balance accuracy and real-time performance simultaneously. To achieve this goal, in this paper, we propose a transparent object depth completion network called TCRNet based on a cascade refinement structure, which balances accuracy and real-time performance simultaneously. First, the network incorporates a cascade refinement structure in the decoding stage to refine features multiple times, improving the accuracy of depth information. Additionally, an attention module is designed to adjust the extracted features, enabling the network to focus on depth information features in transparent object regions. Finally, a transformer-based error module is implemented in the network’s final output stage to predict and adjust the error between the depth image and the ground truth. TCRNet is trained and tested on three datasets: ClearGrasp, Omniverse Object, and TransCG. It outperforms previous methods in terms of performance. Furthermore, TCRNet is applied to existing grasp detection methods to conduct grasping experiments on transparent objects using a real Baxter robot. Note to Practitioners—With the development of RGB-D camera technology, RGB-D cameras are now widely used in various scenarios such as industrial production, autonomous driving, and robot grasping. However, in certain situations where the camera faces transparent or highly reflective objects, the depth information captured by the camera is often not accurate enough, which can lead to subsequent accidents. Therefore, it is necessary to repair and complete the depth images to achieve accurate understanding of the scene’s depth information. In recent years, with the advancement of deep learning, deep learning-based depth image processing and restoration techniques have been widely applied. In this paper, we propose a high-accuracy network for repairing depth images of transparent objects, which can accurately restore and estimate the depth information of transparent objects in various scenarios. Moreover, experimental results demonstrate that our proposed method can generalize well to other unknown scenes, achieving excellent results.}, archive = {J_TASE}, author = {Di-Hua Zhai and Sheng Yu and Wei Wang and Yuyin Guan and Yuanqing Xia}, doi = {10.1109/TASE.2024.3371583}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1893--1912}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {TCRNet: Transparent object depth completion with cascade refinements}, volume = {22}, year = {2025}, }
• (2025). Hybrid task scheduling in cloud manufacturing with sparse-reward deep reinforcement learning. TASE, 22, 1878--1892. (www)

  @article{ , abstract = {Cloud manufacturing (CMfg) converts the traditional manufacturing system into an Internet-of-things-enabled (IoT-enabled) manufacturing system, where both manufacturing and computational tasks must be scheduled among distributed and heterogeneous resources. Deep reinforcement learning (DRL) has recently become a promising idea for task scheduling in CMfg. However, existing DRL-based methods depend heavily on problem-specific reward engineering and struggle to represent hybrid decision variables. To this end, this paper proposed the sparse-reward deep reinforcement learning (SDRL) method to solve the hybrid task scheduling problem in CMfg. First, the hybrid task scheduling model in CMfg is constructed to minimize the makespan. We reformulate the studied problem as a partially observable Markov decision process (POMDP). Then, the objective hindsight experience replay (objective HER) mechanism is proposed to alleviate the sparse reward issue, through which the scheduling policy can be effectively trained without problem-specific reward engineering. The continuous action space is defined to represent hybrid decision variables, and the implicit action-selection mapping is utilized to alleviate the boundary effect. Numerical experiments validated the effectiveness and superiority of our method compared to eleven popular scheduling algorithms including evolutionary algorithms and DRL. Compared to mainstream DRL scheduling methods, the proposed SDRL outperforms the second-best one at most by 23.6% regarding generalization, and a scheduling solution can be generated in 0.5 seconds. Note to Practitioners—With the intelligentization of the CMfg platform, hybrid tasks, including manufacturing and computational tasks, need to be scheduled simultaneously. However, this hybrid task scheduling problem is rarely considered by existing works. DRL exhibits many benefits in addressing scheduling problems, but the strong dependency on problem-specific reward engineering limits its application. Additionally, most DRL-based scheduling algorithms are discrete-action DRL, restricting their capacity to effectively represent hybrid decision variables. The studied problem originates from the CMfg platform, but the proposed method holds potential for broader application. The scheduling framework and the POMDP modeling can be applied to similar problems, including hybrid, manufacturing, or computational task scheduling problems. The proposed objective HER serves as a general approach to addressing challenges associated with sparse rewards, which can be extended to diverse combinatorial optimization problems aimed at optimizing an objective. We will open-source our codes to help others to apply the method to other fields.}, archive = {J_TASE}, author = {Xiaohan Wang and Yuanjun Laili and Lin Zhang and Yongkui Liu}, doi = {10.1109/TASE.2024.3371250}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1878--1892}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hybrid task scheduling in cloud manufacturing with sparse-reward deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). An unmanned system-guided crowd evacuation method in complex and large-scale evacuation environments. TASE, 22, 1864--1877. (www)

  @article{ , abstract = {With the continuous expansion of the city scale and urbanization, urban road networks are becoming increasingly complex. Moreover, severe and extreme weather events, earthquakes, and other natural disasters occur frequently. Therefore, how to effectively and quickly evacuate urban crowd in dynamic environments is an urgent issue. To carry out the above objective, an unmanned system-guided crowd evacuation method is proposed in the current study. In the proposed method, the robot can perceive the environment in a timely and accurate manner to generate the evacuation map via advanced information technologies such as the Internet of Things or urban brain. Subsequently, an improved elliptic tangent graph approach based on global and local information (ETG-GLI) is utilized to plan a feasible and short evacuation path in large-scale scenarios. Finally, a novel crowd evacuation model based on the social force model is proposed to simulate the actual crowd evacuation process in complex and large-scale environments. To test the performance of the proposed path planning method, 25 different scenarios are proposed to simulate complex urban crowd evacuation environments. The experimental results show that the proposed algorithm outperforms other competitors in terms of path planning ability and computational time. Three actual evacuation cases with 324 pedestrians are modeled to further test the performance of the proposed algorithm. The simulation results demonstrate that the unmanned system-guided crowd evacuation method can find a shorter evacuation path for reducing the evacuation time in three complex and large-scale environments when compared with three other methods. Therefore, the proposed algorithm is a highly effective and promising approach to provide useful decision support and guidance for actual urban planning and urban emergence management. Note to Practitioners—In modern cities, the population density is high and the road network is complex. To evacuate the crowd in a timely and safe manner, planning feasible and short paths in large-scale and complex environments is a critical and challenging task. Therefore, the present study aims to provide a novel method to plan high-quality evacuation routes to guide the pedestrian flow. The performance of the proposed approach is validated in 25 test scenarios and 3 real-world instances. Experimental results demonstrate that the proposed algorithm performs well in terms of path length and computation time. Moreover, the proposed crowd evacuation model can simulate the actual process of crowd evacuation.}, archive = {J_TASE}, author = {Tianrui Wu and Jun Yu and Qingchao Jiang and Qinqin Fan}, doi = {10.1109/TASE.2024.3371102}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1864--1877}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An unmanned system-guided crowd evacuation method in complex and large-scale evacuation environments}, volume = {22}, year = {2025}, }
• (2025). A biarc fitting algorithm for concentric cable-driven manipulators working in oropharyngeal swab samplings. TASE, 22, 1853--1863. (www)

  @article{ , abstract = {A biarc fitting algorithm is proposed for the designed concentric cable-driven manipulator working in oropharyngeal (OP) swab samplings. This algorithm makes the concentric cable-driven manipulator composed of discrete nodes have a high degree of fit with the continuous curve planned by the biarc method. Three advantages of this algorithm are as follows. First of all, by adjusting the tangent at the end point, the problem of large deviation fitting between the actual path and the planned path caused by the fixed tangent at the end point during path tracking is solved. In addition, by adjusting the given starting point, the problem of large deviation fitting between the concentric cable-driven manipulator and biarc curve caused by non-integer number of nodes during configuration fitting is solved. Further, by adjusting the proportionality coefficient, the problem of large deviation fitting between the concentric cable-driven manipulator and biarc curve caused by the odd-even variation of the number of moving nodes during configuration fitting is solved. Tracking the planned paths of OP-swab sampling, the simulation and experiment verify the effectiveness of the biarc fitting algorithm for concentric cable-driven manipulators working in OP-swab samplings. Note to Practitioners—This paper is motivated by the requirement of effective planning for a concentric cable-driven manipulator composed of discrete nodes working in automatic oropharyngeal-swab samplings. Existing planning methods make it difficult to fit a concentric cable-driven manipulators composed of discrete nodes to the planned continuous curve. In this paper, we propose a biarc fitting algorithm to improve the degree of fit between cable-driven manipulator composed of discrete nodes and the planned continuous curve. This algorithm makes the concentric cable-driven manipulator composed of discrete nodes have a high degree of fit with the continuous curve planned by the biarc method. Three key factors of this algorithm to be optimized are the tangent at the end point, given starting point and proportionality coefficient. The simulation and experiment verify the effectiveness of the biarc fitting algorithm. In the future work, we will study the modelling based on variable-stiffness control and force feedback control of concentric cable-driven manipulators. This method can be used not only for concentric cable-driven manipulators working in the medical field, but also for cable-driven manipulators composed of discrete nodes inspecting and maintaining in aerospace, nuclear power plant and other similar fields.}, archive = {J_TASE}, author = {Yuming Gao and Chengjiang Wang and Lihui Jia and Hao Lv and Guikun Lv and Zonggao Mu}, doi = {10.1109/TASE.2024.3371023}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1853--1863}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A biarc fitting algorithm for concentric cable-driven manipulators working in oropharyngeal swab samplings}, volume = {22}, year = {2025}, }
• (2025). Integrated task assignment and trajectory planning for a massive number of agents based on bilayer-coupled mean field games. TASE, 22, 1833--1852. (www)

  @article{ , abstract = {Aiming at the problem of integrated task assignment and trajectory planning of a massive number of agents in the scenario with different priority task nodes and multiple static obstacles, this paper proposes a general framework based on bilayer-coupled mean field games, which couples the minimum cost of trajectory planning of an agent in the task assignment process to achieve a reasonable, globally optimal, and targeted adjustable assignment result. In the proposed general framework, firstly, the multi-population mean field game is used to plan the optimal trajectory of an agent between each pair of priority adjacent task nodes, and the minimum costs are calculated. Then, based on the discrete time finite state space mean field game, a task assignment model in the discrete task space is constructed, and the minimum costs obtained in the trajectory planning are coupled into the model as a reference, the task assignment strategies are finally obtained. Moreover, we give a specific example of the proposed general framework and prove the existence of equilibrium solutions of two mean field games. The effectiveness of the proposed general framework is demonstrated through simulation experiments and results analysis. Note to Practitioners—In multi-agent decision-making and control, task assignment and trajectory planning are two fundamental problems that coexist in many scenarios. Examples include the collaborative exploration of multiple task areas by UAV swarm, and the lane selection and efficient driving of autonomous vehicles. There are many methods for dealing with the integrated task assignment and trajectory planning. However, they have difficulties in dealing with large-scale agent problems, mainly due to the significant increase in communication and computation costs as the number of agents increases. In response to this problem, based on the characteristic of mean field game that transforms the game between individuals into a game between an individual and the whole, this paper proposes a general framework of bilayer-coupled mean field games for the scenario with different priority task nodes and multiple static obstacles. The multi-population mean field game is used to plan the optimal trajectory of agents, and the discrete time finite state space mean field game is utilized for task assignment, in which the cost of trajectory planning between task nodes is considered. We propose a specific example model and theoretically prove the existence of an equilibrium solution of this model. The effectiveness of the general framework is verified by simulation experiments and results analysis.}, archive = {J_TASE}, author = {Zijia Niu and Wang Yao and Yuxin Jin and Sanjin Huang and Xiao Zhang and Langyu Qian}, doi = {10.1109/TASE.2024.3370619}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1833--1852}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated task assignment and trajectory planning for a massive number of agents based on bilayer-coupled mean field games}, volume = {22}, year = {2025}, }
• (2025). NUMA-aware virtual machine placement: New MMMK model and column generation-based decomposition approach. TASE, 22, 1817--1832. (www)

  @article{ , abstract = {The efficiency and profitability of cloud data centers are significantly influenced by virtual machine (VM) placement. However, the Non-Uniform Memory Access (NUMA), which has been practically applied to reduce the memory bandwidth competition, is often neglected in the existing research. Actually, the incorporation of NUMA may change the traditional resource allocation mechanism, and demands for a new VM placement model. Hence, considering the multi-NUMA architecture, this paper studies the NUMA-aware VM placement (NAVMP) problem in a cloud computing system, where the resource pool is composed of enormous number of heterogeneous servers with diverse multi-resource remains. The NAVMP problem is analytically formulated as an integer program (IP). Also, for the first time, the incarnations of VM types are introduced to simplify the VM deployment rules originated from complex NUMA architecture. We aim to maximize the VM provision ability (VPA) of the resource pool, and thus propose a novel Value Function to describe servers’ VPA. The resulting formulation, which is a new variant of the multiple-choice multiple multi-dimensional knapsack (MMMK) problem, is of significant computational challenges. So we customize a decomposition approach based on Column Generation (CG) to support the offline optimization. Numerical experiments on a practical dataset demonstrate the validity and scalability of the customized CG-based approach. Our approach outperforms a professional IP solver, i.e., Cbc, and a popular meta-heuristic algorithm, i.e., genetic algorithm (GA), and can efficiently address large-scale NAVMP instances with ten thousands of VM demands and servers. Note to Practitioners—This paper proposes a novel IP model for NAVMP. To cope with the complicated deployment logic associated with the complex multi-NUMA architecture of modern multi-core systems, we present an NAVMP formulation from the perspective of incarnations of VM types. Different from the traditional VM placement problem that aims to minimize the number of activated servers, i.e., the vector bin packing (VBP)-based model, we adopt the objective that maximizes the VPA of a resource pool for further improving the resource utilization. The resulting formulation is an MMMK problem, which is computational very challenging for a practical scale resource pool. Hence, to mitigate the computation burden, we design and implement a CG-based decomposition approach to support the offline optimization for NAVMP. Parallelization scheme and nontrivial heuristic strategies are applied to promote the computation efficiency. According to our numerical experiments, the proposed decomposition approach demonstrates a much superior solution capacity to the Cbc solver and GA. In particular, to achieve a comparable solution precision with Cbc, the computing time can be reduced by orders of magnitude. Also the CG-based approach outperforms GA in both the solution quality and computation time for large-scale instances. Besides, compared to the VBP model, our MMMK-based NAVMP model has improved the VPA up to 44.39%. Practically, the proposed offline approach can be leveraged to guide online VM allocation decisions, and perform efficient results evaluation.}, archive = {J_TASE}, author = {Xunhang Sun and Xiaoyu Cao and Qiaozhu Zhai and Haisheng Tan and Jianchen Hu and Lei Zhu and Li Su and Wenli Zhou and Feng Gao and Xiaohong Guan}, doi = {10.1109/TASE.2024.3370392}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1817--1832}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {NUMA-aware virtual machine placement: New MMMK model and column generation-based decomposition approach}, volume = {22}, year = {2025}, }
• (2025). The stack loading problem with load-bearing limit. TASE, 22, 1801--1816. (www)

  @article{ , abstract = {The stack loading problem has been studied in recent years for its great impact on the container loading and unloading operations. Among different objectives of the problem considered, minimizing the total number of unordered stackings and minimizing the total number of used stacks are the two important ones, which ensure efficient loading and unloading schedules, as well as reduce storage costs, respectively. The load-bearing setting, where each container has its own weight and bearing weight, is frequently considered in box packing operations but rarely in the existing studies on the stack loading problem. However, the load-bearing constraint on containers is very important for stack loading, because safety is of paramount importance. This paper is the first study on the stack loading problem with the load-bearing constraint with an aim to minimize the number of stacks and the number of unordered stackings. We show that this problem is strongly $\mathcal {NP}$ -hard even when the number of stacks is given and equals 2. For the case where the number of stacks is given and jobs on the bottom tiers are fixed, we show that the problem can be solved by dynamic programming in pseudo-polynomial time. For the general problem, based on a two-index integer linear programming formulation and a tabu search heuristic, we develop a binary-search based matheuristic. Our experimental results demonstrate the efficiency and effectiveness of the newly developed matheuristic. Note to Practitioners—This paper is motivated by the stack loading problem and is the first study on the load-bearing limit case. The load-bearing limit is a fundamental constraint but has not been taken into account in studies in the stack loading problem. Based on ISO Standard 1496, the corner posts and corner fittings of ISO Series I containers can bear a certain amount of weight. If the total weight of the containers above exceeds the load-bearing limit of the lower container, it will hazard the load-bearing safety. This paper proposes two problem formulations: three-index formulation and two-index formulation. The three-index formulation adds the load-bearing limit to the existing stack loading problem formulation. It turns out that the traditional three-index formulation of the stack loading problem is not efficient when being used in solving the problem with load-bearing constraints. Therefore, we propose a new two-index formulation. Apart from the theoretical results, this paper proposes a matheuristic solution framework: firstly, using binary search with greedy matheuristic for feasibility checking to minimize the number of stacks, and secondly, using tabu search matheuristic to minimize the number of unordered stackings. In future research, we will apply the matheuristic to different types of container scenarios and the parallel stack loading case.}, archive = {J_TASE}, author = {Xinbo Zhang and Minming Li and Zhou Xu and Yingchao Zhao}, doi = {10.1109/TASE.2024.3370216}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1801--1816}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The stack loading problem with load-bearing limit}, volume = {22}, year = {2025}, }
• (2025). Fuzzy-sliding-mode strategy for multi-channel nonlinear delay system with disturbance. TASE, 22, 1791--1800. (www)

  @article{ , abstract = {In this paper, the robust control of multi-channel nonlinear delay system is discussed. Firstly, a four-order dynamical model of wind turbine nonlinear system suffers from time-varying delay and match disturbances is established. Then, a channel-dependent fuzzy sliding-mode function is designed and the sliding dynamic is obtained. Meanwhile, the stability analysis of sliding dynamic is proved and the channel-dependent sliding parameters are solved. Furthermore, a fuzzy sliding mode controller is proposed to ensure the sliding surface is reachable. Additionally, an adaptive fuzzy sliding mode controller is presented to deal with the situation that the upper bound of the disturbance is unknown. Finally, a numerical simulation and a semi physical simulation are implemented to demonstrate the effectiveness of the algorithm. Note to Practitioners—This paper was motivated by the issue of the robust control of multi-channel wind turbine nonlinear system. Different from traditional controlled system, the wind turbine nonlinear system often suffers from the delay and disturbances. This paper designs a new fuzzy-sliding-mode strategy to address the above issues. Firstly, we established a four-order nonlinear dynamic of wind turbine system under the time-varying delay and match disturbances. In order to obtain the sliding dynamic, a channel-dependent fuzzy sliding-mode function is presented. For ensuring the sliding surface is reachable, a fuzzy sliding mode controller is proposed. In order to handle the upper bound of the disturbance is unknown, an adaptive fuzzy sliding mode controller is presented. Both simulation and experiment indicate that proposed approaches are feasible and effective.}, archive = {J_TASE}, author = {Meng Li and Yong Chen and Meng Zhang and Tao Ding}, doi = {10.1109/TASE.2024.3370173}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1791--1800}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fuzzy-sliding-mode strategy for multi-channel nonlinear delay system with disturbance}, volume = {22}, year = {2025}, }
• (2025). Rotation angle control strategy for the hip joint of an exoskeleton robot assisted by paraplegic patients considering time-varying inertia. TASE, 22, 1780--1790. (www)

  @article{ , abstract = {The design and development of new exoskeleton robots can help patients with lower limb paralysis realize autonomous walking. During the motion of an exoskeleton robot carrying patients, the inertia of the hip joint will have time-varying characteristics, which will cause fluctuations in the rotation angle and affect the walking stability of the exoskeleton robot. In this paper, we present the rotation angle control strategy for an exoskeleton robot assisted by paraplegic patients with the BP neural network tuning control strategy. First, based on the skeleton structure of human lower limbs, an exoskeleton robot with 12 degrees of freedom is designed and manufactured to help patients autonomously walk. Next, the dynamic model of the exoskeleton robot hip joint is established, which takes into account nonlinear factors such as transfer flexibility, friction torque, and time-varying load inertia. Then, BP neural networks are used to adjust the parameters of the position loop PID controllers in the hip joint, and the tracking error is reduced by adjusting the controller parameters in real time. Finally, walking experiments of the physical prototype of the exoskeleton robot show that the exoskeleton robot designed in this paper can help patients with lower limb paralysis walk autonomously, and the proposed control strategy can reduce the rotation angle tracking error of the hip joint. Note to Practitioners—This paper addresses the importance of dynamic modeling and control for the hip joint in the motion accuracy of exoskeleton robots. The mechanical structure of the proposed exoskeleton robot can realize the autonomous walking of patients with lower limb paralysis. The proposed dynamic modeling method is suitable for the servo system time-varying model of split-limb robots. In addition, the proposed BP neural network control strategy can improve the position control accuracy of time-varying systems, which is suitable for the real-time control of robots. Numerical simulation and physical experiments demonstrate the effectiveness of the proposed control strategy.}, archive = {J_TASE}, author = {Meng Yin and Dongyang Shang and Wujing Cao and Yue Ma and Jinke Li and Dingkui Tian and Xinyu Wu}, doi = {10.1109/TASE.2024.3370168}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1780--1790}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Rotation angle control strategy for the hip joint of an exoskeleton robot assisted by paraplegic patients considering time-varying inertia}, volume = {22}, year = {2025}, }
• (2025). Domain adaptive detection of MAVs: A benchmark and noise suppression network. TASE, 22, 1764--1779. (www)

  @article{ , abstract = {Visual detection of Micro Air Vehicles (MAVs) has attracted increasing attention in recent years due to its important application in various tasks. The existing methods for MAV detection assume that the training set and testing set have the same distribution. As a result, when deployed in new domains, the detectors would have a significant performance degradation due to domain discrepancy. In this paper, we study the problem of cross-domain MAV detection. The contributions of this paper are threefold. 1) We propose a Multi-MAV-Multi-Domain (M3D) dataset consisting of both simulation and realistic images. Compared to other existing datasets, the proposed one is more comprehensive in the sense that it covers rich scenes, diverse MAV types, and various viewing angles. A new benchmark for cross-domain MAV detection is proposed based on the proposed dataset. 2) We propose a Noise Suppression Network (NSN) based on the framework of pseudo-labeling and a large-to-small training procedure. To reduce the challenging pseudo-label noises, two novel modules are designed in this network. The first is a prior-based curriculum learning module for allocating adaptive thresholds for pseudo labels with different difficulties. The second is a masked copy-paste augmentation module for pasting truly-labeled MAVs on unlabeled target images and thus decreasing pseudo-label noises. 3) Extensive experimental results verify the superior performance of the proposed method compared to the state-of-the-art ones. In particular, it achieves mAP of 46.9%(+5.8%), 50.5%(+3.7%), and 61.5%(+11.3%) on the tasks of simulation-to-real adaptation, cross-scene adaptation, and cross-camera adaptation, respectively. Note to Practitioners—To study the cross-domain MAV detection problem, this paper establishes a novel benchmark that consists of three domain adaptation tasks: simulation-to-real adaptation, cross-scene adaptation, and cross-camera adaptation, respectively. The benchmark is based on a novel MAV dataset called Multi-MAV-Multi-Domain (M3D), which is available at: https://github.com/WestlakeAerialRobotics/M3D. To reduce the noises caused by pseudo labels, a noise suppression network is proposed to overcome the error accumulation. Extensive experiments are conducted to prove the effectiveness of the proposed approach.}, archive = {J_TASE}, author = {Yin Zhang and Jinhong Deng and Peidong Liu and Wen Li and Shiyu Zhao}, doi = {10.1109/TASE.2024.3370147}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1764--1779}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Domain adaptive detection of MAVs: A benchmark and noise suppression network}, volume = {22}, year = {2025}, }
• (2025). A dynamic movement primitives-based tool use skill learning and transfer framework for robot manipulation. TASE, 22, 1748--1763. (www)

  @article{ , abstract = {This paper presents a framework for learning and transferring robot tool-use skills based on Dynamic Movement Primitives (DMPs) for robot fine manipulation. DMPs and their enhanced methods are employed to acquire a specific tool-use skill applicable to tools with similar sizes, shapes, and uses. However, the acquired skills may not be transferable to other scenarios and tools with variations. The new framework introduces two new types of skills based on DMPs: Object Operating (O2) skill and Tool Flipping (TF) skill. The O2 skill enables robots to handle tools for manipulating objects to achieve desired effects. The learning process for the O2 skill considers limitations imposed by tools and the environment during human demonstrations. Distinguishing between whether constraints can be modelled or not, we propose both a model-based and a constraint-based method to separate a constraint-irrelevant (CI) skill and the constrained conditions. The CI skill is generalized using a novel method called constrained -DMP lite, enabling adaptation to new tasks with special tools. The TF skill addresses situations where tools must generate an action to alter contacting positions on both objects and tools while avoiding conflicts during movement. Finally, the TF and O2 skills are generalized to be applied in creating a continuous action chain. We conduct several experiments to compare and analyze the advantages and disadvantages of the proposed methods with other approaches in terms of generalizability and calculation complexity. Note to Practitioners—Strengthening robot tool-use ability has been a hot research topic in recent years because these tools can extend the reachability and enhance the flexibility of robots. The previous research on DMPs has been utilized for learning tool-use skills. However, the learned skills few considered the tools’ special use regulations, therefore the skill of using a tool is hard to transfer to another tool-use case. This paper explores tool-use skill learning and transfer between different tools by developing a framework based on the DMPs for this problem. The framework consists of two kinds of skills: O2 skill and TF skill with different purposes as well as a series of newly developed algorithms, such as constrained -DMP lite, a model-based and a constraint-based CI skill learning methods. These methods can separate the constraints from human demonstrations of using tools to achieve a CI skill and generalize the CI skill according to the constraints generated from a new tool-use manipulation task. We verify the effectiveness of the proposed framework through some typical tool-use experiments, including pushing objects, cutting and obstacle avoidance in actuality. The development of this framework can be used in industrial and house working scenarios.}, archive = {J_TASE}, author = {Zhenyu Lu and Ning Wang and Chenguang Yang}, doi = {10.1109/TASE.2024.3370139}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1748--1763}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A dynamic movement primitives-based tool use skill learning and transfer framework for robot manipulation}, volume = {22}, year = {2025}, }
• (2025). Virtual verification and validation to enhance sustainability of manufacturing systems. TASE, 22, 1738--1747. (www)

  @article{ , abstract = {In this paper, the concept of Sustainable Manufacturing (SM) is discussed, and the focus is put on the elimination of wastes in a product’s life cycle. Recent development on enabling technologies to advance SM are surveyed to identify promising research areas. We have found virtual Verification and Validation (V&V) should be promoted to enhance the sustainability especially in Small and Medium sized Enterprises (SMEs). V&V are typically non-value-added and the effects on V&V should be minimized; while due to lack of expertise, most of SMEs rely heavily on prototyping and physical experiments to evaluate their products. We propose virtual V&V to replace physical experiments to the maximum extent. To show the significance of the proposed concept, a case study of virtual V&V system is developed to reduce the needs of physical prototyping and testing in product development. This benefits greatly to (1) cost savings by replacing numerous physical prototyping and testing by virtual V&V, (2) lead-time reduction for new product to enter emerging markets, and (3) global optimization of product design by analyzing and exploring a large number of design options. The improvement at these aspects contributes to system sustainability significantly. The concept of using virtual V&V to reduce non-value-added prototyping and testing is applicable to manufacturing businesses in most of SMEs who design and test their own products. Note to Practitioners—This work is highly motivated by a number of the authors’ industry projects with regional SMEs who put heavy investments in testing to prove products’ quality to clients before the orders for products can be awarded; manufacturing businesses related to testing are non-value-added that should be minimized from the perspective of sustainability. Virtual V&V is proposed as a vital solution to overcome their dilemma, and the proposed solution has its theoretical and practical significance to reduce wastes, shorten lead-times, and optimize products based on parametric study for a wide scope of design alternatives. This helps to increase the lifespan of SMEs ultimately.}, archive = {J_TASE}, author = {Zhuming Bi and Aki Mikkola and Andrew W. H. Ip and K. L. Yung and Chaomin Luo}, doi = {10.1109/TASE.2024.3370053}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1738--1747}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Virtual verification and validation to enhance sustainability of manufacturing systems}, volume = {22}, year = {2025}, }
• (2025). Generalized disturbance estimation based continuous integral terminal sliding mode control for magnetic levitation systems. TASE, 22, 1725--1737. (www)

  @article{ , abstract = {For the purpose of better inhibiting the chattering phenomenon and elevating the control performance of the magnetic levitation system under multiple disturbances, a continuous integral terminal sliding mode controller (CITSMC) based on generalized disturbance estimation (GDE) is put forward in this paper. Firstly, stemming from the mechanism modeling approach, an approximate nonlinear model of the magnetic levitation ball system is established, and then the integral terminal sliding mode controller (ITSMC) is designed to guarantee finite-time convergence. Due to the system’s vulnerability to multiple external disturbances, the switching gain needs to be selected at a larger value, which will lead to serious chattering problem. As a result of these concerns, a generalized disturbance estimator is proposed, which can effectively estimate and compensate for the multiple disturbances and reduce chattering by selecting a smaller switching gain value. To further eliminate the negative influence of chattering, the super-twisting reaching law is introduced in the design process of ITSMC. The proposed composite controller has the advantages of being chattering-free and having robust anti-interference capabilities. Computer simulations and experiments are employed to double-check the effectiveness and ascendancy of the proposed strategy. Note to Practitioners—Magnetic levitation systems are subject to complicated external multiple disturbances and internal uncertainties as typical open-loop unstable systems with nonlinearity. As a consequence, developing a powerful controller for magnetic levitation systems is a challenging proposition. Firstly, aiming at the asymptotic convergence problem of traditional sliding mode, an integral terminal sliding mode controller is devised to realize finite-time convergence and avoid singularity. Secondly, a generalized disturbance estimator has been proposed for estimating and compensating for multiple disturbances. Finally, for the chattering problem, a continuous integral terminal sliding mode controller combined with super-twisting is designed to effectively eliminate the adverse effects of chattering. Simulation and experimental results verify the effectiveness and superiority of the proposed method.}, archive = {J_TASE}, author = {Junxiao Wang and Qinchen Jiang and Huiming Wang}, doi = {10.1109/TASE.2024.3369905}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1725--1737}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Generalized disturbance estimation based continuous integral terminal sliding mode control for magnetic levitation systems}, volume = {22}, year = {2025}, }
• (2025). Mutual supervision of MFL heterogeneous signals for insufficient sample defect detection on pipeline safety operation. TASE, 22, 1714--1724. (www)

  @article{ , abstract = {Magnetic flux leakage (MFL) testing is an effective non-destructive testing (NDT) method for pipeline safety operation, and defect detection is one of the core issues in MFL signal processing. Currently, MFL defect detection is tough due to insufficient defect samples. To obtain more serviceable information, heterogeneous signals are collected in MFL process, while taking full advantage of heterogeneous signals is still a hard problem. In this paper, an end-to-end MFL defect detection architecture with insufficient training samples called heterogeneous signal fusion method (IHSF) is proposed. Firstly, heterogeneous signals from the same pipeline are collected by axial and radial sensors. Secondly, the features from fine-tuned and non-fine-tuned pre-trained models are fused, which increases the generality and adaptability of the features. Moreover, fine-tuning of a few parameters reduces the number of parameters during model training, which is more suitable for the insufficient sample training. Thirdly, a mutual supervision training strategy based on the features fusion of general and adaptive features is proposed to update the parameters of the proposed network, which establishes the latent relationship of heterogeneous signals. Finally, experiments on MFL defect detection are conducted, and IHSF is compared to the state-of-the-art methods. The results validate that the proposed method is effective. Note to Practitioners—The motivation of this paper is a hot signal processing issue on pipeline safety operation called defect detection with insufficient samples. Traditionally, the defect is usually detected by training a large number of defect samples, while defect samples are usually insufficient in practical MFL measurement. Moreover, conventional deep networks only use a single signal, which ignores information from multiple signals. Regarding the problems above, a heterogeneous signal mutual supervision method is proposed to replace the traditional fixed label mechanism. The mutual supervision method fully extracts the information in multiple signals by establishing hidden connections between defective heterogeneous signals. Furthermore, the fusion of the general features and adaptive features, which are extracted by the pre-trained model with non-fine-tuning and fine-tuning respectively, not only reduces the parameters in the network to be suitable for the insufficient sample problem, but also enriches the extracted features. The case on MFL signals demonstrates that the proposed method is more effective than the comparison methods.}, archive = {J_TASE}, author = {Lin Jiang and Huaguang Zhang and Jinhai Liu and Xiangkai Shen and Hang Xu}, doi = {10.1109/TASE.2024.3369659}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1714--1724}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Mutual supervision of MFL heterogeneous signals for insufficient sample defect detection on pipeline safety operation}, volume = {22}, year = {2025}, }
• (2025). A multistage update rule framework for iterative learning control systems. TASE, 22, 1697--1713. (www)

  @article{ , abstract = {The proportional type update rule (PTUR) is the most widely used iterative learning control (ILC) scheme. Recently, a fractional-power type update rule (FTUR) was proposed to accelerate PTUR. However, PTUR and FTUR converge slowly for small and large tracking errors, respectively. In this study, a multistage update rule (MSUR) is designed to accelerate PTUR and FTUR along the whole iteration axis. Under the proposed switching mechanism, PTUR and FTUR are adopted for large and small errors for fast convergence, and then PTUR is applied for zero-error tracking. The convergence of MSUR is proved by the analysis of a nonlinear recursion with perturbation. Moreover, for system information that is unknown, an extended MSUR is presented, and its zero-error convergence is proved. In addition, we discuss the influence of the parameters in MSUR on the convergence rate and propose a set of parameter selection rules to maximize the convergence rate of MSUR. Meanwhile, variable-gain and variable-power MSURs are designed to further accelerate the MSUR that only has a single gain and fractional power. Numerical simulations and experimental test verify the theoretical results. Note to Practitioners—Many engineering systems, including high-speed trains, earth-orbiting satellites, and robotic arms, complete a given tracking task over a finite time interval repeatedly. Fast convergence rate and high tracking precision are key technical requirements in these applications. Iterative learning control (ILC) has been shown as an effective control method for these tasks. However, the widely-used proportional-type update rule (PTUR) and fractional-power type update rule (FTUR) in ILC cannot meet the abovementioned requirements well. PTUR converges fast for large tracking errors but slow for small tracking errors; in contrast, FTUR delivers fast convergence rate for small tracking errors but converges slow for large errors without achieving zero-error. Combining advantages of both PTUR and FTUR rules, a novel multistage update rule (MSUR) is proposed to adaptively switch between them to deliver fast convergence and zero tracking errors. The selection of parameters and their effect on control performance are detailed for engineering applications.}, archive = {J_TASE}, author = {Zihan Li and Dong Shen and Xinghuo Yu}, doi = {10.1109/TASE.2024.3369651}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1697--1713}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A multistage update rule framework for iterative learning control systems}, volume = {22}, year = {2025}, }
• (2025). Multi-agent reinforcement learning with decentralized distribution correction. TASE, 22, 1684--1696. (www)

  @article{ , abstract = {This work considers decentralized multi-agent reinforcement learning (MARL), where the global states and rewards are assumed to be fully observable, while the local behavior policy is preserved locally for resisting adversarial attack. In order to cooperatively accumulate more rewards, the agents exchange messages among a time-varying communication network to reach consensus. For these cooperative tasks, we propose a decentralized actor-critic algorithm, where the agents make individual decisions, but the joint behavior policy is optimized towards more cumulative rewards. We provide the theoretical analysis towards the convergence under the tabular setting and then expand it to nonlinear function approximations. Furthermore, by incorporating decentralized distribution correction, the agents are trained in an off-policy manner for higher sample efficiency. Finally, we conduct experiments to evaluate the algorithms, where the proposed algorithm performs competitively in both stability and asymptotic performance. Note to Practitioners—Fully decentralized MARL algorithms are widely applied in multi-agent systems for generating cooperative behaviors, e.g., multiple unmanned aerial vehicles (UAV) cooperatively performing search and rescue tasks, multiple vehicles efficiently passing a crowded intersection, and multiple robots cooperatively handling cargo or obstacles. Focusing on these potential applications, this work is motivated to improve the sample efficiency of recent decentralized MARL algorithms by incorporating off-policy training approaches. In this work, we reweight historical trajectories via a decentralized average consensus step and develop corresponding policy-optimization procedures, with which previous trajectories could be used to stabilize later iterations. Since the training materials are augmented by historical samples, the sample efficiency is significantly improved, and the training process is stabilized. With the fully decentralized training approach, the proposed algorithms are expected to be applied in large-scale systems, e.g., vehicle teams and UAV groups, for effective real-time control.}, archive = {J_TASE}, author = {Kuo Li and Qing-Shan Jia}, doi = {10.1109/TASE.2024.3369592}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1684--1696}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multi-agent reinforcement learning with decentralized distribution correction}, volume = {22}, year = {2025}, }
• (2025). A two-stage individual feedback NSGA-III for dynamic many-objective flexible job shop scheduling problem. TASE, 22, 1673--1683. (www)

  @article{ , abstract = {Dynamic events, such as machine fault and rush order insertion, are fairly common in the job shop scheduling, which may lead to significant delay in order delivery and low production efficiency. Under such circumstance, it is urgent to consider more perspectives in the scheduling, such as delay time and equipment load rate. In this article, a dynamic many-objective flexible job shop scheduling problem (DMaFJSP) is founded to simultaneously optimize the completion time, delay time, total equipment load and energy consumption. Canonical many-objective optimization algorithms are seeing difficulties in maintaining population diversity and enduring poor adaptability in dynamic scheduling problems. The paper proposes a two-stage individual feedback non-dominated sorting genetic algorithm-III (TSIF-NSGA-III), where a new population diversity strategy and an individual feedback strategy are added to expand the global search faculty and stronger dynamic adaptability. Numerical study in many-objective problem and dynamic many-objective problem are conducted. The final results illustrate that the proposed algorithm can with effect dispose of the DMaFJSP. Note to Practitioners—This paper was motivated by the flexible job shop scheduling problem (FJSP) in practical dynamic situations. In the actual production procedure, however, FJSP is a more challenging issue. Not only operation sequencing and machine allocation matters, but also uncertain factors in the environment, such as machine fault, rush order insertion, etc., are important. In addition, the majority of current researchers formulate the FJSP simply focusing on maximum completion time. However, low carbon and high efficient manufacturing calls for more objectives. In this paper, two dynamic incidents, machine stoppage and rush order insertion, are considered. In addition, the model of DMaFJSP is established with many objectives such as total energy consumption, completion time, equipment load and delay time. To resolve foregoing problems, this article proposes a TSIF-NSGA-III algorithm, which adopts a diversity generation strategy and an individual feedback strategy to strengthen the search ability and dynamic adaptability of this algorithm. Preliminary simulation outcomes illuminate that this algorithm has certain advantages. In addition, the algorithm can also be applied to other multi-objective workshop scheduling problems, such as mixed flow workshop, distributed workshop, etc.}, archive = {J_TASE}, author = {Yi Feng and Yating Lin and Zhile Yang and Yunlang Xu and Di Li and Xiaoou Li and Dongsheng Yang}, doi = {10.1109/TASE.2024.3369019}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1673--1683}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A two-stage individual feedback NSGA-III for dynamic many-objective flexible job shop scheduling problem}, volume = {22}, year = {2025}, }
• (2025). An autonomous guidewire delivery method with distal position-based impedance control. TASE, 22, 1658--1672. (www)

  @article{ , abstract = {Leader-follower robotic guidewire delivery system has been a research focus in terms of vascular interventional surgery (VIS) for a few years. However, due to the inconvenient operations and lack of intuition in leader robot manipulation, such systems are not widely used. To get rid of cumbersome operations and promote the safety of guidewire delivery in a VIS, the first autonomous guidewire delivery scheme for VIS which incorporates both human skill and robotic autonomous manipulation is proposed in this article. The delivery criterion of a surgeon remotely manipulating the guidewire is formulated by an impedance model. Then, an autonomous delivery controller based on this model steers the guidewire distal position, not the proximal position which is commonly used in relevant research, towards the desired location. We reduce the effect of uncertainty in flexible guidewire deformation by a real-time estimator. The proposed method exhibits distinct superiorities in both human skill and robotic fast response. Experimental results confirmed the advantages of the proposed autonomous delivery scheme in reducing cumbersome operations and smoothening contact force in a VIS. Note to Practitioners—Many researchers have developed a variety of leader-follower robotic systems for the VIS to reduce harm to surgeons from radiation, because of the occupational use of X-rays in interventional surgeries. However, the insertion manipulation, as a common movement in a VIS, needs a large workspace which is in contradiction to the workspaces of commonly used leader robots. For the sake of less cumbersome movements and smooth operational force, an autonomous guidewire delivery scheme is proposed. This scheme is composed of two phases. In the first phase, the surgeon manipulates the leader robot to remotely complete just one insertion movement. From this movement, manual manipulation skill is extracted by an impedance model, which reveals the relationship between the guidewire tip position and the contact force. Then, in the second phase, an autonomous delivery controller based on the impedance model is devised and adopted to complete the insertion task. By taking the guidewire tip position, not the follower robot position, as a fundamental component of the impedance model, this scheme conforms to sophisticated manual manipulation skill. Such an autonomous scheme takes advantage of both human skills and robotic stability and can be applied in autopilot, assembling, etc.}, archive = {J_TASE}, author = {Zheng Liu and Shuang Wang and Yongfeng Cao and Zhenggang Cao and Li Lin and Le Xie}, doi = {10.1109/TASE.2024.3368764}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1658--1672}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An autonomous guidewire delivery method with distal position-based impedance control}, volume = {22}, year = {2025}, }
• (2025). Goal-oriented visual semantic navigation using semantic knowledge graph and transformer. TASE, 22, 1647--1657. (www)

  @article{ , abstract = {When determining navigation actions, it is important to design effective visual and semantic representations of the observation scenes and robust navigation strategies. The paper proposes a goal-oriented visual semantic navigation method using semantic knowledge graph and transformer. Two kinds of knowledge graphs representing the location relationship between objects are constructed, namely current knowledge graph and prior knowledge graph. The pre-constructed prior knowledge graph is periodically updated by the current knowledge graph obtained in real time, and embedded into the semantic feature vector through graph convolutional network (GCN). The semantic features and extracted scene features are jointly embedded and stored, they are jointly fed into the transformer module to explore the spatio-temporal dependencies between objects in the environment. The navigation strategy is obtained from the Asynchronous Advantage Actor-Critic (A3C) model composed of Long-Short Term Memory (LSTM) and Multi-Layer Perception (MLP). Experiments show that the knowledge graph can significantly improve the navigation performance. More importantly, our experimental results show that our method can improve the generalization ability of navigation across novel scenes and novel objects. Video can be available at https://youtu.be/ZMjNvoK2rbY. Note to Practitioners—The motivation of this work is to develop an efficient visual semantic navigation method. Conventional navigation algorithms lack semantic information and learning ability, and can not adapt to the complex unknown environments. When semantic information is included in navigation, the location relationship between objects can be obtained as a prior knowledge, which can be combined with reinforcement learning to achieve autonomous navigation of agents. In this article, a knowledge graph representing the location relationships between objects has been constructed and regularly updated in real-time. The proposed visual semantic navigation method further improves the generalization ability of navigation. This navigation method can be applied to mobile robots and deployed in many scenarios such as home, restaurant, hospitals, and even factories.}, archive = {J_TASE}, author = {Zhongli Wang and Guohui Tian}, doi = {10.1109/TASE.2024.3368696}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1647--1657}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Goal-oriented visual semantic navigation using semantic knowledge graph and transformer}, volume = {22}, year = {2025}, }
• (2025). Scheduling unrelated parallel batch processing machines under time-of-use electricity prices. TASE, 22, 1631--1646. (www)

  @article{ , abstract = {The increasing energy consumption of manufacturing companies has placed a lot of pressure on the power grid. To alleviate the pressure and flatten the demand peaks over a day, time-of-use electricity pricing scheme has been widely adopted around the world, which provides an opportunity for manufacturers to reduce their energy costs by avoiding production during peak hours. In this study, we investigate the unrelated parallel batch processing machine scheduling under time-of-use electricity prices. The objective is to minimize the total electricity cost while ensuring that the makespan does not exceed a predetermined deadline. To address this problem, we propose a novel mixed integer linear programming formulation along with several valid cuts. Moreover, the Dantzig-Wolfe decomposition technique is utilized to reformulate this problem, and a branch and price algorithm is developed to deal with large-scale instances. Extensive numerical experiments demonstrate the superiority of our new formulation compared to existing formulations. Furthermore, the inclusion of valid cuts can significantly improve the efficiency of both the existing and proposed formulations. And our branch and price algorithm can efficiently solve instances with 200 jobs with an average gap of less than 3% in 30 minutes. Note to Practitioners—As an efficient production method, parallel batch processing is widely used in many manufacturing scenarios, such as the burn-in operation in semiconductor manufacturing, the hardening operation in aircraft industry, and the glass ceramization operation. The workpieces are usually processed in a high temperature and high-pressure environment for hours in these scenarios, and the energy consumption cost is relatively high. In recent years, time-of-use pricing scheme has been widely adopted around the world to rebalance the energy demand, which provides manufacturers with great opportunities to save their energy consumption costs. In this paper, we investigate how to optimize the electricity cost under time-of-use prices in a parallel batch processing machine environment. An efficient time-interval-indexed mixed integer linear programming formulation and a branch and price algorithm are proposed to solve this problem. The proposed algorithm can provide high-quality solutions for real-world-scale instances within a reasonable time limit. We also analyze the performances of different methods under various situations to help the practitioners choose the most suitable method.}, archive = {J_TASE}, author = {Zheng Tian and Li Zheng}, doi = {10.1109/TASE.2024.3368617}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1631--1646}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling unrelated parallel batch processing machines under time-of-use electricity prices}, volume = {22}, year = {2025}, }
• (2025). Event-triggered optimal control for a class of continuous-time switched nonlinear systems. TASE, 22, 1620--1630. (www)

  @article{ , abstract = {This paper studies the optimal switching and control co-design for a class of continuous-time switched nonlinear systems. An event-triggered adaptive dynamic programming (ADP) algorithm is developed to obtain the optimal hybrid control policy. At the event-triggered instant, the switching controller determines which subsystem to activate and the input controller updates the system input. Compared with the time-triggered methods, the computation and communication are reduced. A critic neural network (NN) is applied to approximate the solution of the switched Hamilton-Jacobi-Bellman (HJB) equation. The proposed critic NN is tuned according to the HJB error in real time. A stability analysis of the closed-loop system is given by Lyapunov method. Moreover, the design of the event-triggered mechanism can also exclude Zeno behavior of the switching signal without fixing a minimum dwell time. Finally, the effectiveness of the developed algorithm is evaluated by a numerical simulation. Note to Practitioners—This paper is motivated by the great potential switched nonlinear systems have demonstrated in improving system dynamic performances. However, the existing results mainly focus on optimal control for discrete-time switched nonlinear systems. Moreover, fairly few researchers have investigated the optimal switching and control co-design for switched nonlinear systems. Therefore, this paper develops a novel event-triggered adaptive dynamic programming algorithm that can learn the optimal hybrid control policy online. The design of the event triggering mechanism can exclude Zeno behavior of the switching signal. It is noteworthy that the proposed control method is of great significance for many practical systems, such as automotive engine systems and single-link robot arm systems.}, archive = {J_TASE}, author = {Zhengrong Xiang and Pingchuan Li and Wencheng Zou}, doi = {10.1109/TASE.2024.3368438}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1620--1630}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered optimal control for a class of continuous-time switched nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). Event-triggered finite-time tracking control for nonlinear systems via immersion and invariance techniques. TASE, 22, 1606--1619. (www)

  @article{ , abstract = {This paper researches the event-triggered forward immersion and invariant (I&I) tracking control problem for a class of strict feedback nonlinear systems. A forward I&I-based control method is developed for the tracking problem with an dynamic event-triggered mechanism. Since I&I-based method does not require the introduction of Lyapunov functions in the controller design, the design complexity is greatly reduced. Since the I&I-based method can be used to decouple the design by constructing two manifolds separately, it avoids the need of the traditional backstepping method to combine the Lyapunov function coupled design control law of radial basis function neural networks (RBFNNs). I&I adaptive technique is introduced to improve the weight update in RBFNNs. It can improve the learning performance and convergence speed of neural networks under the event-triggered mechanism. Furthermore, finite-time technique is employed to improve the error convergence time of the event-triggered forward I&I control method. For stability analysis, an event-triggered control system is denoted as a nonlinear impulsive dynamical system, and a Lyapunov theorem is then used to represent the stability of the closed-loop system without Zeno behavior. Finally, the validity of the theoretical results is illustrated by simulation examples and experiments. Note to Practitioners—The motivation of this paper is to present an event-triggered forward I&I tracking control method in finite time for a class strict feedback nonlinear system. The use of I&I technique can reduce the complexity of control method design. To reduce the computational resources, the event-triggered mechanism is introduced in the forward I&I technique. The I&I technique is introduced to construct two manifolds separately and decouple the design of the control law and the RBFNNs weight update law. Moreover, the I&I adaptive technique with the event-triggered mechanism is employed to improve the approximation effect of RBFNNs. Finally, the finite-time technique is introduced to reduce the convergence time of the tracking error under the event-triggered mechanism. This proposed method can be simply and efficiently applied in industrial applications.}, archive = {J_TASE}, author = {Jianchao He and Zijun Cheng and Tianshui Chang and Qidong Liu and Yang Yang}, doi = {10.1109/TASE.2024.3368415}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1606--1619}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered finite-time tracking control for nonlinear systems via immersion and invariance techniques}, volume = {22}, year = {2025}, }
• (2025). AMI-net: Adaptive mask inpainting network for industrial anomaly detection and localization. TASE, 22, 1591--1605. (www)

  @article{ , abstract = {Unsupervised visual anomaly detection is crucial for enhancing industrial production quality and efficiency. Among unsupervised methods, reconstruction approaches are popular due to their simplicity and effectiveness. The key aspect of reconstruction methods lies in the restoration of anomalous regions, which current methods have not satisfactorily achieved. To tackle this issue, we introduce a novel Adaptive Mask Inpainting Network (AMI-Net) from the perspective of adaptive mask-inpainting. In contrast to traditional reconstruction methods that treat non-semantic image pixels as targets, our method uses a pre-trained network to extract multi-scale semantic features as reconstruction targets. Given the multiscale nature of industrial defects, we incorporate a training strategy involving random positional and quantitative masking. Moreover, we propose an innovative adaptive mask generator capable of generating adaptive masks that effectively mask anomalous regions while preserving normal regions. In this manner, the model can leverage the visible normal global contextual information to restore the masked anomalous regions, thereby effectively suppressing the reconstruction of defects. Extensive experimental results on the MVTec AD and BTAD industrial datasets validate the effectiveness of the proposed method. Additionally, AMI-Net exhibits exceptional real-time performance, striking a favorable balance between detection accuracy and speed, rendering it highly suitable for industrial applications. Note to Practitioners—AMI-Net restores defective images to normal ones and subsequently detects defects by leveraging the differences between them. This method only needs to collect about a few hundred defect-free samples for training, without the need for additional defect samples. It is noteworthy that AMI-Net is applicable not only to the detection of simple texture surface defects, such as carpet, leather, and tile, but also to the detection of surface defects in objects with posture diversity, such as cable, transistor, and screw. The trained model not only exhibits high detection accuracy but also demonstrates superior real-time performance, showcasing significant potential in practical industrial settings.}, archive = {J_TASE}, author = {Wei Luo and Haiming Yao and Wenyong Yu and Zhengyong Li}, doi = {10.1109/TASE.2024.3368142}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1591--1605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {AMI-net: Adaptive mask inpainting network for industrial anomaly detection and localization}, volume = {22}, year = {2025}, }
• (2025). Feasibility in multistage robust dispatch with renewables: A recursive characterization and scalable approximation. TASE, 22, 1579--1590. (www)

  @article{ , abstract = {Finding a feasible solution is the primary concern in power system dispatch. This paper studies the feasibility condition of power system dispatch under a multistage robust optimization framework considering the non-anticipativity of dispatch policy, which is difficult to be expressed via explicit constraints. The multistage robust feasible regions (MRFRs) are defined as the sets in the state space containing all points that can maintain the robust feasibility in the next period against renewable and demand uncertainties; we give a polyhedral projection condition to characterize exact MRFRs in a recursive manner, which can be regarded as an analog of Bellman’s optimality condition. However, because the multistage dispatch problem of a bulk power system has a high-dimensional state space, the computation of exact MRFRs suffers from the curse of dimensionality. We propose an inner approximation method that identifies the maximal polyhedra that are embraced by the unknown exact MRFRs; we devise a computationally efficient algorithm to retrieve the hyperplane representation of the inner approximator. Finally, we discuss how MRFRs can be used in combination with existing approaches, such as dynamic programming and rolling horizon optimization. Numerical simulations on a modified IEEE 118-bus system verify the effectiveness and advantages of the proposed method. Note to Practitioners—This paper is motivated by the problem of maintaining the feasibility in power system multistage dispatch under renewable generation uncertainty. The proposed method regards the multistage dispatch as a sequential decision-making process and recursively defines the MRFR using polyhedral projection technique, which contains all the state points in each period that can guarantee the robust feasibility in the next period. To release the curses of dimensionality, the large-scale MRFR is approximated by an inner hyper-rectangle in the state space, which can be decomposed into independent intervals of state variables, such as the generation power range of coal-fired unit and the state-of-charge range of battery storage unit. These decoupled intervals are convenient for practical use and desired in the real-world power system. At the current stage, the calculation of MRFR requires a linear model; in the future research, integers will be addressed so that more facilities such as non-ideal energy storage units and fast-startup generators can be involved.}, archive = {J_TASE}, author = {Zhongjie Guo and Jiayu Bai and Wei Wei and Shengwei Mei and Weihao Hu}, doi = {10.1109/TASE.2024.3368088}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1579--1590}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Feasibility in multistage robust dispatch with renewables: A recursive characterization and scalable approximation}, volume = {22}, year = {2025}, }
• (2025). Adaptive fuzzy control for underactuated robot systems with inaccurate actuated states and unavailable unactuated states. TASE, 22, 1566--1578. (www)

  @article{ , abstract = {Underactuated robot systems, due to their unique flexibility and economy, are widely used in modern industry and intelligent manufacturing. However, their underactuated nature and complex nonlinearity make the control problem challenging. In addition, most existing control methods for underactuated systems take the measurable actuated and unactuated states as an implicit premise and do not consider the unknown measurement sensitivity of sensors. Unfortunately, due to space/costs or manufacturing technology limitations in practical underactuated robot systems, it is difficult to measure unactuated states and achieve the ideal working mode of sensors. To this end, an adaptive fuzzy control scheme is proposed for a class of uncertain underactuated robot systems without unactuated state sensors, which stabilizes the system even with actuated state measurement errors. Specifically, the dynamics of underactuated systems are reconstructed into a nontriangular normal form, and the backstepping design is completed by using the boundedness of fuzzy basis functions. At the same time, the explosion of complexity is avoided by using dynamic surface control technology. Moreover, the proposed update law availably compensates parameter/structure uncertainties for underactuated robot systems. The asymptotic stability of the closed-loop system is proved by incorporating Lyapunov candidates with sensitivity information. Finally, the proposed control scheme is applied to a tower crane system, whose effectiveness is verified by hardware experiments. Note to Practitioners—This paper is motivated by the common sensor problems in uncertain underactuated robot systems control. Some states in underactuated systems are called unactuated states because they have no direct corresponding control inputs. Such states bring significant challenges to controllers design of underactuated systems. Furthermore, most existing control methods require all states to be measurable and ignore the measurement uncertainty of sensors. However, in many practical situations, it is difficult to install unactuated state sensors, and there are inevitable measurement errors in actuated state sensors. These factors make the feedback controllers developed for underactuated systems difficult to practically apply. To address these issues, this paper presents a control scheme that does not require unactuated state measurements and model information, which can still stabilize underactuated systems to the origin even in the presence of actuated state measurement errors. The stability is rigorously proven theoretically and the experimental results obtained on a self-built tower crane platform demonstrate the feasibility and effectiveness of the proposed control scheme. In future efforts, we intend to apply the proposed control scheme to practical industrial underactuated robot systems.}, archive = {J_TASE}, author = {Meng Zhai and Shuzhen Diao and Tong Yang and Qingxiang Wu and Yongchun Fang and Ning Sun}, doi = {10.1109/TASE.2024.3367921}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1566--1578}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive fuzzy control for underactuated robot systems with inaccurate actuated states and unavailable unactuated states}, volume = {22}, year = {2025}, }
• (2025). An evolutionary partition based method for solving scheduling problems with hard Q-times. TASE, 22, 1554--1565. (www)

  @article{ , abstract = {This paper proposes a new evolutionary partition based method to tackle complex combinatorial optimization problems, such as a class of scheduling problems with hard Q-time constraints. The proposed method creatively integrates prune mechanism and evolutionary theory into the Nested Partition scheme. Theoretical results reveals that this method converges with probability one to the optimal solution for any given combinatorial optimization problem with finite feasible solutions. The study also presents a hybrid variant of this method, incorporating a partheno-genetic algorithm, thereby demonstrating the flexibility and robustness of the proposed method. Through numerical experiments, the robust capacity and efficiency of the proposed methods compared with the Branch and Bound method, the Nested Partition method, and Genetic Algorithms are highlighted. Note to Practitioners—This paper is motivated by a complex large-scale scheduling problem with hard Q-Time constraints prevalent in various industries. Existing methods either fail to meet these hard constraints or satisfy them at the expense of a large makespan. This paper presents a novel evolutionary partition based method with solid mathematical proofs ensuring its capability of problem solving. Numerical results show that our method outperforms existing methods by providing superior quality solutions in a shorter computational time while ensuring that the hard Q-times are met. The effectiveness, efficiency, and flexibility make this method an attractive approach for dealing with practical scheduling problems.}, archive = {J_TASE}, author = {Chaoran Wang and Leyuan Shi}, doi = {10.1109/TASE.2024.3367914}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1554--1565}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An evolutionary partition based method for solving scheduling problems with hard Q-times}, volume = {22}, year = {2025}, }
• (2025). Synchronizability and protocol design of discrete-time second-order networks over finite fields. TASE, 22, 1543--1553. (www)

  @article{ , abstract = {In this paper, the distributed cooperative control problems of discrete-time second-order networks over finite fields are investigated, under which a new perspective is provided for discrete-time second-order networks of agents with limited capacities of memory and communication. Firstly, the discrete-time second-order multi-agent model with a given protocol is established over the finite field. Then the leaderless finite-field synchronization network is studied, some necessary and sufficient conditions along with the design principles of the synchronous protocol are given to guarantee the network synchronization. In addition, the finite-field consensus problems including leaderless and leader-following cases are studied in sequence, both of which prove to be unachievable under the distributed control protocol. Finally, an application on camera networks and numerical examples are provided to support the obtained results. Note to Practitioners—The multi-agent system is a collection of multiple agents with computing and mobility capabilities, and its distributed cooperative control problem aims to achieve the global objective by neighbor-based protocols, mainly including consensus, synchronization, formation control, etc. Because of its strong fault tolerance, robustness and scalability, it is widely used in military, industrial and agricultural production, medicine, transportation and so on. These three factors that mainly affect the group behavior of agents are the dynamic evolution, the network topology (that is, the connection between agents), and the communication mechanism. However, in practical applications, due to limited computation, memory, and communication capabilities, the state of the network is constrainted, which directly affects the dynamic evolution of the system. In this paper, a new kind of second-order synchronization dynamics, which can be used in certain applications such as multi-agent systems with periodic boundary conditions on position and velocity, is proposed. The states and communication weights of agents can only attain values from a finite alphabet equipped with a prime number of integers, called finite fields. Such networks have a finite convergence time, and are therefore friendly to multi-agent systems with limited capacities.}, archive = {J_TASE}, author = {Jin Wang and Jun-e Feng and Yongyuan Yu}, doi = {10.1109/TASE.2024.3367886}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1543--1553}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Synchronizability and protocol design of discrete-time second-order networks over finite fields}, volume = {22}, year = {2025}, }
• (2025). A serial game distributed algorithm to ϵ-minimum vertex cover of networks in finite time. TASE, 22, 1534--1542. (www)

  @article{ , abstract = {Vertex cover problem is a typical nondeterministic polynomial combinatorial optimization problem with wide applications. In order to seek a near-optimal solution in finite time, this article achieves the goal from the perspective of serial potential game optimization. Specifically, we present the vertex cover problem as a potential game, where the corresponding potential function minimizers are equivalent to the minimum vertex cover state. Then, we propose a novel polynomial-based serial game distributed algorithm, and prove that the algorithm can guarantee that strategies of all vertices converge to a $\epsilon $ -minimum vertex cover ( $\epsilon $ -MVC) state in finite time. Compared with the existing representative optimization algorithms on networks and standard benchmarks, numerical simulations demonstrate our proposed algorithm can effectively balance solution efficiency and computation time. We hope this work provides insight for decision-makers designing a reasonable algorithm when solving nondeterministic polynomial optimization problems, so as to guarantee both solution quality and computation time. Note to Practitioners—Vertex cover problem has a wide range of practical applications. Many distributed optimization algorithms for the vertex cover problem are available in the existing literatures. However, those algorithms mainly pursue high-quality solutions in infinite time. Thus, one of the main challenges is to design a distributed algorithm that guarantees both solution quality and computation time. This article design a polynomial-based serial game algorithm, and prove that it can converge to the $\epsilon $ -MVC state in finite time. This article can serve as a supplement to the existing optimization algorithms for solving the vertex cover problem.}, archive = {J_TASE}, author = {Jie Chen and Rongpei Zhou and Weihua Gui}, doi = {10.1109/TASE.2024.3367722}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1534--1542}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A serial game distributed algorithm to ϵ-minimum vertex cover of networks in finite time}, volume = {22}, year = {2025}, }
• (2025). An end-to-end deep reinforcement learning based modular task allocation framework for autonomous mobile systems. TASE, 22, 1519--1533. (www)

  @article{ , abstract = {Intelligent decision-making systems that can solve task allocation problems are critical for multi-robot systems to conduct industrial applications in a collaborative and automated way, such as warehouse inspection using mobile robots, hydrographic surveying using unmanned surface vehicles, etc. This paper, therefore, aims to address the task allocation problem for multi-agent autonomous mobile systems to autonomously and intelligently allocate multiple tasks to a fleet of robots. Such a problem is normally regarded as an independent decision-making process decoupled from the following task planning for the member robots. To avoid the sub-optimal allocation caused by the decoupling, an end-to-end task allocation framework is proposed to tackle this combinatorial optimisation problem while taking the succeeding task planning into account during the optimisation process. The problem is formulated as a special variant of the multi-depot multiple travelling salesmen problem (mTSP). The proposed end-to-end task allocation framework employs deep reinforcement learning methods to replace the handcrafted heuristics used in previous works. The proposed framework features a modular design of the reinforcement learning agent which can be customised for various applications. Moreover, a real-robot implementation setup based on the Robot Operating System 2 is presented to fulfil the simulation-to-reality gap. A warehouse inspection mission is executed to validate the training outcome of the proposed framework. The framework has been cross-validated via both simulated and real-robot tests with various parameter settings, where adaptability and performance are well demonstrated.Note to Practitioners—This paper is motivated by the problem of dispatching a fleet of autonomous mobile robots to tackle a mission that can be resolved into multiple waypoint-following tasks. An end-to-end modular framework is proposed, making task allocation decisions based on the given waypoint information. By using the reinforcement learning technique, the deep neural network could learn sophisticated policies for allocating tasks. The policies are trained in a specific pattern which ensures their joint optimisation for a solver that outputs the near optimal task execution sequences in an efficient way. This leads to a multiple travelling salesmen problem (mTSP) solution. Pre-trained policies are tested in several industrial scenarios reflecting the applications of search and rescue, maritime surveying, and warehouse automation, among others. A hardware implementation configuration based on the Robot Operating System 2 is also presented to support the practical deployment the framework.}, archive = {J_TASE}, author = {Song Ma and Jingqing Ruan and Yali Du and Richard Bucknall and Yuanchang Liu}, doi = {10.1109/TASE.2024.3367237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1519--1533}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An end-to-end deep reinforcement learning based modular task allocation framework for autonomous mobile systems}, volume = {22}, year = {2025}, }
• (2025). Robust cooperative load frequency control for enhancing wind energy integration in multi-area power systems. TASE, 22, 1508--1518. (www)

  @article{ , abstract = {The wind energy, as a kind of renewable energy resources, has the potential to replace traditional fossil fuels. However, its intermittent power output can incur frequency instability due to the instantaneous unbalance between power generation and load demand. To smooth the penetration of wind energy, this paper presents a robust cooperative load frequency control (LFC) strategy for multi-area power systems, which is a hierarchical control approach. For the low-level wind turbine control, this paper adopts model predictive control (MPC) method to achieve the rated wind power tracking. In the meantime, an improved event-triggered scheme (ETS) considering multiple historic released signals is employed to relieve the computational burden of MPC. For the high-level cooperative LFC, this paper incorporates the robust performance index in the control synthesis to suppress the impact of intermittent wind power on frequency stability. In addition, to address the underlying shift of the steady-state operating point caused by the intermittent wind power supply, this paper improves the commonly used small-signal LFC model by adding an uncertain matrix, which reasonably explains the possible change of system parameters and extends the applicability of the traditional LFC model. Simulations are done on a four-area power system, and the results verify the efficacy of the presented event-triggered scheme and the robust cooperative LFC approach.Note to Practitioners—To promote the penetration of wind energy into power systems, this work explores a robust cooperative LFC approach under multi-agent structure to ensure the stability of the system, aiming at extending the applicability of existing approaches. The proposed approach is hierarchical. At the rated wind power tracking level, the MPC is employed to handle constraints associated with actuating devices, such as heterogeneous convertors. Simultaneously, an improved ETS considering multiple historic triggered signals is integrated in the MPC to reduce the computational burden. At the power system level, the robust performance index is incorporated in the control design to smooth the impacts of intermittent wind power on frequency stability. Additionally, the study accounts for the potential shift of the steady-state operating point and improves the traditional small-signal LFC model by adding an uncertain matrix, which can better explain the variation of system parameters and is more applicable in practical power system engineering. Simulation results demonstrate that the proposed robust cooperative LFC approach can effectively maintain the system frequency within the admissible range under the high penetration of wind energy, whereas the traditional PI controller falls short in this regard.}, archive = {J_TASE}, author = {Zhijian Hu and Kun Zhang and Rong Su and Ruiping Wang}, doi = {10.1109/TASE.2024.3367030}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1508--1518}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust cooperative load frequency control for enhancing wind energy integration in multi-area power systems}, volume = {22}, year = {2025}, }
• (2025). Exact-optimal consensus of uncertain nonlinear multi-agent systems based on fuzzy approximation. TASE, 22, 1497--1507. (www)

  @article{ , abstract = {This paper is concerned with the distributed optimal consensus problem of multi-agent systems with uncertain nonlinear dynamics. The essential work of this paper is the development of an exact-optimal consensus control methodology via fuzzy adaptive control technique. Specially, an optimal signal generator for each agent is established to cooperatively estimate the accurately global optimal solution. Subsequently, a novel fuzzy adaptive controller is designed with the aid of command filtered backstepping and adaptive compensation technique, making it powerful enough to the fuzzy approximate error and filtering error. Thereby, it provides a concise optimal consensus control algorithm for strict-feedback nonlinear systems, which avoids the requirement that the gradient of the local cost function is high-order differentiable. Based on Lyapunov stability theory, it is proven that the outputs of all agents synchronize to an optimal agreement with asymptotic convergence; i.e., the exact-optimal consensus is able to be guaranteed under the proposed control approach. Finally, simulation studies and comparisons are presented to show the effectiveness of the proposed control approach. Note to Practitioners—This paper addresses the distributed optimal consensus problem of nonlinear multi-agent systems with uncertain dynamics, which is more challengeable. It is the first effort to study the exact-optimal consensus control problem via function approximation technique. Through constructing multiple optimal signal generators to cooperatively estimate the accurately global optimal solution, the distributed optimal consensus control is converted into an asymptotic tracking control. Then, a concise optimal consensus control is proposed by command filtered backstepping technique, which is easy to implement in practice. The developed control protocol can be applied to unmaned swarm systems, and it has contributions to academic and practical applications.}, archive = {J_TASE}, author = {Wei Wang and Yongming Li and Shaocheng Tong}, doi = {10.1109/TASE.2024.3366999}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1497--1507}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Exact-optimal consensus of uncertain nonlinear multi-agent systems based on fuzzy approximation}, volume = {22}, year = {2025}, }
• (2025). Learning neural force manifolds for Sim2Real robotic symmetrical paper folding. TASE, 22, 1483--1496. (www)

  @article{ , abstract = {Robotic manipulation of slender objects is challenging, especially when the induced deformations are large and nonlinear. Traditionally, learning-based control approaches, such as imitation learning, have been used to address deformable material manipulation. These approaches lack generality and often suffer critical failure from a simple switch of material, geometric, and/or environmental (e.g., friction) properties. This article tackles a fundamental but difficult deformable manipulation task: forming a predefined fold in paper with only a single manipulator. A sim2real framework combining physically-accurate simulation and machine learning is used to train a deep neural network capable of predicting the external forces induced on the manipulated paper given a grasp position. We frame the problem using scaling analysis, resulting in a control framework robust against material and geometric changes. Path planning is then carried out over the generated “neural force manifold” to produce robot manipulation trajectories optimized to prevent sliding, with offline trajectory generation finishing $15\times $ faster than previous physics-based folding methods. The inference speed of the trained model enables the incorporation of real-time visual feedback to achieve closed-loop model-predictive control. Real-world experiments demonstrate that our framework can greatly improve robotic manipulation performance compared to state-of-the-art folding strategies, even when manipulating paper objects of various materials and shapes. Note to Practitioners—This article is motivated by the need for efficient robotic folding strategies for stiff materials such as paper. Previous robot folding strategies have focused primarily on soft materials (e.g., cloth) possessing minimal bending resistance or relied on multiple complex manipulators and sensors, significantly increasing computational and monetary costs. In contrast, we formulate a robust, sim2real, physics-based method capable of folding papers of varying stiffness with a single manipulator. The proposed folding scheme is limited to papers of homogeneous material and folding along symmetric centerlines. Future work will involve formulating efficient methods for folding along arbitrary geometries and preexisting creases.}, archive = {J_TASE}, author = {Andrew Choi and Dezhong Tong and Demetri Terzopoulos and Jungseock Joo and Mohammad Khalid Jawed}, doi = {10.1109/TASE.2024.3366909}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1483--1496}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning neural force manifolds for Sim2Real robotic symmetrical paper folding}, volume = {22}, year = {2025}, }
• (2025). VSG-based adaptive optimal frequency regulation for AC microgrids with nonlinear dynamics. TASE, 22, 1471--1482. (www)

  @article{ , abstract = {In this paper, an optimal frequency control approach is proposed for VSG-based AC Microgrids (MGs) to improve the frequency regulation performance by considering the nonlinear dynamics. First, a nonlinear dynamics system of the MG are analytically modelled by using the VSG controller. Following, the optimal controller is designed based on the Hamilton-Jacobi-Bellman (HJB) equation and is obtained by using the adaptive dynamic programming (ADP) method without linearizing. Then, a single neural networks (NNs) construction is used to approximate the optimal controller and cost function, simultaneously. Unlike the conventional VSG approach with constant inertia, the novel method can determine the optimal inertia in an online and adaptive way, as well as preserve a balance between the frequency regulation performance and the control cost. Finally, simulation results based on MATLAB/Simulink verify that the proposed control method improves the frequency nadir and the rate of change of the frequency (RoCoF) while DC-side energy is also drastically preserved by comparing the existing three studies. Note to Practitioners—a) Previous methods always assume that the generator can instantly generate and absorb infinite power, however, capacitance is finite in practice. This paper proposes an adaptive VSG-based method to provide a bounded threshold of inertia, which facilitates practical applications. b) The AC MG system is modeled as a nonlinear system to cope with the strong disturbance, which is more suitable for practical system. c) The controller is approximated by NN, which is designed by three physical variables, i.e., angular deviation, frequency deviation and rate of change of the frequency (RoCoF), which can be measured from the practical system. d) Small signal state space model is used to analyze the stability of the controller, which further proves the stability of the proposed method and enhances its application prospect in engineering practice.}, archive = {J_TASE}, author = {Chong Liu and Zhousheng Chu and Zhongxing Duan and Yi Zhang}, doi = {10.1109/TASE.2024.3366700}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1471--1482}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {VSG-based adaptive optimal frequency regulation for AC microgrids with nonlinear dynamics}, volume = {22}, year = {2025}, }
• (2025). Micro transfer learning mechanism for cross-domain equipment RUL prediction. TASE, 22, 1460--1470. (www)

  @article{ , abstract = {Transfer learning generally addresses to reduce the distribution distance between source-domain and target-domain. However, it is unreasonable to use a distribution to represent the life-cycle signals as they are always time-varying, and the improper assumption affects the efficacy of transfer remaining useful life (RUL) prediction. To fill this gap, this research proposes a micro transfer learning mechanism for multiple differentiated distributions, and a transfer RUL prediction model is constructed. First, a multi-cellular long short-term memory (MCLSTM) neural network is applied to obtain multiple differentiated distributions of the monitoring data at some point. Then the domain adversarial mechanism is used to achieve the knowledge transfer of multiple differentiated distributions at the cell level. Furthermore, an active screen mechanism is designed for weighting the domain discrimination losses of multiple differentiated distributions. Through the transfer RUL prediction experiments on aero-engines and actual wind turbine gearboxes, the superiority of this model over the advanced transfer prediction models is verified. Note to Practitioners—The work is motivated by the accuracy reduction problem caused by the time-varying characteristics of life-cycle data in the cross domain equipment RUL prediction scenario, where a fixed single distribution is difficult to cover the full life-cycle data. This article proposes a micro transfer learning mechanism containing multiple differentiated distributions, and a novel transfer RUL prediction model based on the mechanism is constructed for solving the problem caused by the time-varying characteristics of life-cycle data. There are four steps for implementing this method in practice: 1) collecting the full-life cycle signals of historical equipment; 2) modeling the degradation curves of equipment by MCLSTM; 3) solving the cross domain RUL prediction by narrowing the distributions of degradation curves by the micro transfer learning mechanism; and 4) making prognostics for new equipment. The novelty is that the proposed mechanism can self-adaptively align multiple differentiated subspaces of the source domain and the target domain, that is, it can adaptively extract the domain invariant features over time. As a result, the proposed method has two main advantages: 1) capable of characterizing the degradation processes of different equipment; and 2) superior prognostic results on cross domain RUL prediction.}, archive = {J_TASE}, author = {Sheng Xiang and Penghua Li and Jun Luo and Yi Qin}, doi = {10.1109/TASE.2024.3366288}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1460--1470}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Micro transfer learning mechanism for cross-domain equipment RUL prediction}, volume = {22}, year = {2025}, }
• (2025). Intermittent iterative learning control for robot manipulators under packet dropouts. TASE, 22, 1451--1459. (www)

  @article{ , abstract = {This paper is concerned with the intermittent iterative learning control for robot manipulators with packet dropouts. A composite controller is presented by a proportional-derivative feedback and an iterative learning feedforward to deal with the dropouts. A modified reference trajectory is adopted to treat input intermittence. With the help of composite energy function, rigorous theoretical derivations are provided to ensure the convergence of the estimation error of the iterative estimator and the boundedness of the tracking error. The validity of the proposed method is demonstrated by simulation studies on a single-link manipulator and a two-degrees-of-freedom planar manipulator, respectively. Note to Practitioners—Robot manipulators have drawn significant attentions due to their wide industrial applications. Due to their repetitive feature, iterative learning control is an effective strategy to achieve high tracking accuracy for them. However, this strategy faces new challenges incurred from network environment, such as packet dropouts and how to save energy. To patch this gap, an interpolation method is exploited for an iterative estimator to handle packet dropouts and a modified reference trajectory is employed to tackle with input intermittence. Simulation studies with practical background is supplied to show the validity of the proposed method. Therefore, this paper lays a theoretical basis for the tracking control of robot manipulators in practical applications.}, archive = {J_TASE}, author = {Mouquan Shen and Xingzheng Wu and Song Zhu and Tingwen Huang and Huaicheng Yan}, doi = {10.1109/TASE.2024.3365813}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1451--1459}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intermittent iterative learning control for robot manipulators under packet dropouts}, volume = {22}, year = {2025}, }
• (2025). PyRoboCOP: Python-based robotic control and optimization package for manipulation and collision avoidance. TASE, 22, 1435--1450. (www)

  @article{ , abstract = {Contacts are central to most manipulation tasks as they provide additional dexterity to robots to perform challenging tasks. However, frictional contacts leads to complex complementarity constraints. Planning in the presence of contacts requires robust handling of these constraints to find feasible solutions. This paper presents PY ROBO COP which is a lightweight Python-based package for control and optimization of robotic systems described by nonlinear Differential Algebraic Equations (DAEs). In particular, the proposed optimization package can handle systems with contacts that are described by complementarity constraints. We also present a general framework for specifying obstacle avoidance constraints using complementarity constraints. The package performs direct transcription of the DAEs into a set of nonlinear equations by performing orthogonal collocation on finite elements. The resulting optimization problem belongs to the class of Mathematical Programs with Complementarity Constraints (MPCCs). MPCCs fail to satisfy commonly assumed constraint qualifications and require special handling of the complementarity constraints in order for NonLinear Program (NLP) solvers to solve them effectively. PY ROBO COP provides automatic reformulation of the complementarity constraints that enables NLP solvers to perform optimization of robotic systems. The package is interfaced with ADOL-C for obtaining sparse derivatives by automatic differentiation and IPOPT for performing optimization. We provide extensive numerical examples for various different robotic systems with collision avoidance as well as contact constraints represented using complementarity constraints. We provide comparisons with other open source optimization packages like CasADi and Pyomo. The code is open sourced and available at https://github.com/merlresearch/PyRoboCOP. Note to Practitioners—PY ROBO COP is intended to be an easy-to-use software package written in Python which can be used for optimization, estimation and control for a large class of robotic systems. Including, in particular, contact-rich applications to deal with complex scenarios that arise when making and breaking contacts during a task. Typical problems that can be solved with our work are trajectory and control sequence optimization, parameter estimation. To make the proposed software package easier for practitioners, the paper provides access to the package and a large number of example problems. Furthermore, the package also provides a guide describing the details of all the methods a user might have to implement for their own system. Compared to some of the other packages, PY ROBO COP works with NumPy object arrays which is the native computing package in Python. We believe that this will make it much easier to learn and use compared to some of the other optimal control packages.}, archive = {J_TASE}, author = {Arvind U. Raghunathan and Devesh K. Jha and Diego Romeres}, doi = {10.1109/TASE.2024.3365637}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1435--1450}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {PyRoboCOP: Python-based robotic control and optimization package for manipulation and collision avoidance}, volume = {22}, year = {2025}, }
• (2025). In-motion coarse alignment for SINS/USBL based on USBL relative position. TASE, 22, 1425--1434. (www)

  @article{ , abstract = {The accuracy of initial attitude calculated from initial alignment has an important influence on the performance of SINS/USBL integrated navigation system. Currently, two main problems need to be overcome for SINS to complete the in-motion coarse alignment assisted by USBL system. 1) As SINS cannot obtain accurate initial attitude during the initial alignment phase, so USBL system can only provide the relative position in acoustic frame, but not the precise absolute position in navigation frame. 2) The outliers contained in the USBL raw data will seriously affect the performance of coarse alignment. In this paper, an in-motion coarse alignment method for SINS/USBL using the relative position of USBL is creatively proposed. Firstly, after constructing the apparent displacement vector from the relative position provided by USBL system and the output of IMU, an in-motion vector observation model is constructed. Secondly, the proposed method can suppress the influence of outliers in USBL raw data based on the advantages of displacement vector. Finally, the experimental results of the simulation and field tests show that the proposed method can not only complete the in-motion coarse alignment with favorable performance, but also effectively suppress the influence of the outliers in USBL data on the coarse alignment.Note to Practitioners—This paper was motivated by the problem of rapid in-motion coarse alignment for the SINS/USBL integrated navigation system in an underwater environment. Due to the failure of GNSS signals in underwater environments, the existing in-motion coarse alignment algorithms assisted by GNSS position/velocity are no longer applicable. In this paper, an in-motion coarse alignment method assisted by USBL relative position is proposed for SINS in underwater environments. In this article, we compare and analyze the differences between the principle of USBL positioning and GNSS positioning. Based on the characteristics of USBL positioning principle, the displacement vectors are extracted from the relative position of USBL for coarse alignment in this paper. The designed method can effectively suppress the impact of noise and outliers in USBL data on coarse alignment performance. The results of the simulation and field tests indicate that the method investigated in this paper is feasible in practical engineering.}, archive = {J_TASE}, author = {Yongyun Zhu and Tao Zhang and Bingbo Cui and Xinhua Wei and Bonan Jin}, doi = {10.1109/TASE.2024.3398018}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1425--1434}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {In-motion coarse alignment for SINS/USBL based on USBL relative position}, volume = {22}, year = {2025}, }
• (2025). A unified framework of in-situ calibration and synchronous identification for industrial robots using composite sensing. TASE, 22, 1405--1424. (www)

  @article{ , abstract = {Previous current-based dynamic identification methods suffer from coupled motor-side uncertainties, imprecise joint drive gains, and current noises. The combined identification methods integrate the bedplate wrench and motor current to derive the decoupled link-side and motor-side dynamic parameters. However, the in-situ calibration demand for the bedplate sensor and drive gain complicates the identification process. Further, asynchronous calibration and identification accumulate and amplify errors. Therefore, this article proposes a unified framework of synchronous calibration and identification for industrial robots. First, the bedplate sensor calibration and dynamic identification are formulated into a unified optimization model and solved. Next, the drive gains and joint dynamic parameters are extracted under the same optimization structure. Then, a unified framework is founded to seamlessly integrate the in-situ calibration, synchronous identification, physical feasibility constraints, composite sensing fusion, and nonlinear friction estimation. Finally, the excitation trajectory for combined identification is delicately designed and executed. The high identification precision is evaluated through cross-validation experiments by the prediction performance of bedplate wrench and joint torque by up to 97.581% and 93.318%, respectively. The suggested framework outperforms other advanced methods and decreases parameter standard deviations. Note to Practitioners—Accurate robot dynamics are crucial for high-performance model-based control techniques. This research aims to address two issues that interfere with combined dynamic identification results: the coupled motor-side uncertainties induced by current signals and the accumulated errors caused by asynchronous sensing calibration and robot dynamic identification. Thus, this article proposes a unified framework of in-situ calibration and synchronous identification for industrial robots using composite sensing, which can bring many benefits to practitioners. Firstly, the proposed framework extracts the robot dynamic and sensing device parameters within one excitation trajectory execution, which significantly refines the results and simplifies the experimental process by the integrated algorithm structure. Secondly, the suggested optimization paradigm of synchronous calibration and identification can be implemented for different types of industrial robots in various scenarios easily. Thirdly, the calibrated device parameters and measurement covariance matrices contribute to the further applications of robot force perception and control.}, archive = {J_TASE}, author = {Yan Lu and Zhikai Shen and Hongbo Hu and Chungang Zhuang and Han Ding}, doi = {10.1109/TASE.2024.3378383}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1405--1424}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A unified framework of in-situ calibration and synchronous identification for industrial robots using composite sensing}, volume = {22}, year = {2025}, }
• (2025). Evolutionary multitasking memetic algorithm for distributed hybrid flow-shop scheduling problem with deterioration effect. TASE, 22, 1390--1404. (www)

  @article{ , abstract = {In the production enterprises, the distributed hybrid flow-shop scheduling problems widely exist in the actual production controlling and decision, especially in the production of steel and aluminum. Considering the uncertain processing time in the actual production environment, the constraint of deteriorated variable processing time is added in some problems. In this paper, the distributed hybrid flow-shop scheduling problem with deterioration effect (DHFSP-DE) is investigated. The framework of evolutionary multitasking memetic algorithm (MTMA) is designed to address the proposed model. In the proposed method, evolutionary transfer learning is utilized to communicate between two independent DHFSP-DE solvers. The implicit knowledge of the scheduling scheme can be transferred into other solvers to guide the evolution of the population. The memetic algorithm combines a strategy of local intensification with a population-based paradigm. These strategies can capture the implicit knowledge of DHFSP-DE. This paper makes the first attempt to work on the framework of evolutionary multitasking learning for DHFSP-DE problems. The experimental results on the different instances show the effectiveness and efficiency of the proposed MTMA algorithm. Note to Practitioners—The distributed hybrid flow-shop scheduling problem with deterioration effect (DHFSP-DE) is modeled based on the production process of aluminum. The DHFSP-DE is also widely used in the production process of various industries. As the processing time of the job is varied with time, the determined scheduling problem is transformed into a scheduling problem with deterioration effect. This transformation makes the problem even more complicated. The problems are more complex than static problems because they require greater computational dimensionality for evolutionary computation, resulting in the use of computational resources. An evolutionary multitasking memetic algorithm is designed to solve DHFSP-DEs cooperatively. The solutions of the solver can be transferred to other solvers through the mapping of DHFSP-DEs. The transferred solution can affect the evolution process. The efficiency and effectiveness of the proposed MTMA are verified by the comparison experiments. In addition, the MTMA framework can be applied to other scheduling problems.}, archive = {J_TASE}, author = {Huan Liu and Fuqing Zhao and Ling Wang and Tianpeng Xu and Chenxing Dong}, doi = {10.1109/TASE.2024.3365518}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1390--1404}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Evolutionary multitasking memetic algorithm for distributed hybrid flow-shop scheduling problem with deterioration effect}, volume = {22}, year = {2025}, }
• (2025). Protocol-based synchronization of semi-markovian jump neural networks with DoS attacks and application to quadruple-tank process. TASE, 22, 1377--1389. (www)

  @article{ , abstract = {This paper deals with the synchronization for semi-Markovian jump neural networks (S-MJNNs) with DoS attacks and adaptive event-triggered protocol. Aiming at reducing some conservatism of sojourn-time-exponential distribution of Markovian process, the semi-Markovian process is adopted to describe the sudden changes in structures and parameters. As a class of networked systems, the securities of dynamical systems are vulnerable to the DoS attacks that are supposed to obey the frequency constraint. First, on the basis of fully considering the characteristics of aperiodic DoS attacks, a feedback controller is constructed to realize the synchronization among the master-slave systems, in which an adaptive event-triggered protocol is adopted to adjust the amount of triggered data and save network resources more effectively than traditional event-triggered protocol. Then, sufficient conditions to ensure the exponential synchronization and solvable controller gain of the underlying S-MJNNs are given by means of stochastic Lyapunov stability and integral inequality. Finally, the quadruple-tank process model is shown to verify the proposed method. Note to Practitioners—As one of the research hotspots of neural networks, synchronization plays a crucial role in pattern recognition, associative memory, and information science. With the development of computer science and network technology, the information transmission is carried out through the network in the synchronized control of neural networks, which is bound to be affected by the non-ideal network environment. Therefore, the research on synchronization control should not only consider the realization of control method, but also consider the influence of network delay, cyber attacks, and other non-ideal network environment on system performance from the perspective of communication network. In this paper, the synchronization is studied for S-MJNNs with DoS attacks and adaptive event-triggered protocol. On the basis of fully considering the characteristics of aperiodic DoS attacks, the synchronization among the master-slave systems is realized by means of multiple Lyapunov functions and integral inequality. This study provides a new method for the practitioners to study the synchronization control strategy of neural networks affected by the cyber attacks.}, archive = {J_TASE}, author = {Wenhai Qi and Ning Zhang and Ju H. Park and Zheng-Guang Wu and Huaicheng Yan}, doi = {10.1109/TASE.2024.3365503}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1377--1389}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Protocol-based synchronization of semi-markovian jump neural networks with DoS attacks and application to quadruple-tank process}, volume = {22}, year = {2025}, }
• (2025). Voltage-adaptive strategy for transient stability enhancement of power systems with 100% renewable energy. TASE, 22, 1364--1376. (www)

  @article{ , abstract = {This paper proposes a novel voltage-adaptive strategy (VAS) considering current limits of renewable energy resources (RESs), to enhance the transient stability of the power system with 100% renewable energy (PSRE). First, taking the current limits into account, a new transient stability mechanism is revealed by deriving the fault critical clearing time (CCT) of a PSRE with two RESs. Next, leveraging the Lie derivative and the Lyapunov theory, a novel adaptive control method is proposed, which is more in line with the output saturation characteristic of RESs. Then, VAS is formulated based on the proposed adaptive control method for enhancing the transient stability of PSRE. Finally, the proposed VAS is verified on a modified IEEE 11-bus system with 100% renewable energy generation. Note to Practitioners—Transient control can be considered one of the challenges in power system field for constructing the PSRE, which has significant implications for reducing carbon emissions. This work presents a VAS of RESs taking the current limits into account for improving the transient characteristics of the PSRE. Practitioners should be able to apply the VAS to RESs represented by wind farms and photovoltaic power stations. The VAS implements the transient control through the Lie derivative and the proposed adaptive control law. From a practical point of view, it should be highly emphasized that the proposed VAS can achieve transient control of the PSRE without increasing any investment and bringing negative impacts.}, archive = {J_TASE}, author = {Guoteng Wang and Ying Huang and Chongyu Wang and Mohammad Shahidehpour and Quanrui Hao}, doi = {10.1109/TASE.2024.3364709}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1364--1376}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Voltage-adaptive strategy for transient stability enhancement of power systems with 100% renewable energy}, volume = {22}, year = {2025}, }
• (2025). Globally stealthy attacks against distributed state estimation in smart grid. TASE, 22, 1353--1363. (www)

  @article{ , abstract = {With the continuous expansion of grid nodes, traditional centralized methods exhibit certain limitations in the amount of communication data and computational cost for state estimation. In the past few decades, distributed state estimation has been fully developed in multi-area smart grid (SG). According to different areas or electrical equipment, the SG is divided into several nodes, and each node can estimate the entire state of the SG through local sensor measurement information and neighbor state estimation information. Although malicious attackers can use the different information between local sensor data and neighbor state estimation communication data to attack grid nodes, it will inevitably cause the attack detectors of other nodes to trigger alarms due to the consistency of grid state estimation. In this paper, we propose a globally stealthy attack strategy that biases the state estimation of the attacked node without triggering alarms on all nodes’ detectors. We provide sufficient and necessary conditions to implement a globally stealthy attack on grid nodes and extend it to multi-node attacks in order to reduce the number of communication links and increase the degree of damage. We also give the corresponding attack algorithm to realize these attacks. Finally, our proposed globally stealthy attack methods are simulated and verified in IEEE 39-bus power system. Note to Practitioners—In this paper, we propose two attack strategies against distributed state estimation in smart grids and present the corresponding attack methods. Since the grid is a multi-area power system, attacks between different nodes are more likely than attacks within nodes. Most of the previous literature focused on the ability of an attack to remain stealthy within the attacked node. However, due to the presence of the consistency term of the distributed state estimation, the state of the attacked node will be propagated to other nodes and detected by other nodes. Therefore, in this paper, we consider an approach that can maintain stealthy at all nodes. The effectiveness of the proposed attack strategy against distributed state estimation in smart grid is demonstrated in IEEE 39-bus power system. It is shown that a globally stealthy attack leads to a bias in the state estimation of the attacked node and does not trigger an alarm at any node. In summary, the proposed attack strategy is useful for the research community to design targeted detection strategies and can be conveniently applied to real-world security management systems in smart grids.}, archive = {J_TASE}, author = {Shaojie Xu and Dan Ye and Guangdi Li and Dongsheng Yang}, doi = {10.1109/TASE.2024.3364231}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1353--1363}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Globally stealthy attacks against distributed state estimation in smart grid}, volume = {22}, year = {2025}, }
• (2025). Jointly appointment scheduling in a two-phase service system with two types of patients considering multiple servers and stochastic service time. TASE, 22, 1339--1352. (www)

  @article{ , abstract = {This article focuses on the appointment scheduling problem with stochastic service times in a two-phase healthcare service system. In this system, the two phases’ services refer to medical test and physician consulting, respectively. And two types of patients with different service process exist on a specific day. One type of patient is the new-coming patient who only needs to be served by the physician (if a test is needed, it will be booked for another day due to resource limitation). The other type of patient is the “revisiting” patient who enters the system to get complex test followed by previous suggestions of physician, and he/she will visit the physician with the test results on the same day. All patients get their services through appointments. And the current policy of managing those appointments applied in hospitals is to schedule the appointments in two phases separately. However, as this policy does not take the interaction of two phases’ service into consideration, lots of conflicts emerge in the operation process. To avoid the conflict of patients from different types and improve the system’s efficiency, we innovatively proposed a jointly scheduling policy, which aims to make a joint scheduling decision for all the patients in two phases. To achieve this, we first formulate this problem as a stochastic program and conduct the sample average approximation approach to reformulate it as a deterministic problem. To solve it efficiently, a hybrid VNS_LSHAPE algorithm which combines the advantage of VNS (Variable Neighborhood Search) and L shape algorithm is subtly developed based on the properties of the problem. Finally, numerical experiments show that our proposed jointly scheduling policy has an overwhelming performance on the system’s total cost compared with separately scheduling policy. Both types of patients’ satisfaction and the utility of physicians are improved with this policy. Note to Practitioners—In many departments of hospitals, medical test and physician diagnose are two necessary service items provided to the patients. Such a department can be regarded as a two-phase service system. Generally, two types of patients usually exist in this system simultaneously on a work day. One type refers to the new-coming patients who come for an initial consulting of the physician. And the other type refers to the patients who revisit this system to take medical test and physician’s further diagnosis sequentially. Since making appointments can relieve systems’ congestion, many hospitals begin to provide service through appointing. How to schedule the appointments reasonably for those two types of patients in a two-phase service system proves to be a crucial issue. In this article, a jointly scheduling policy which determines the appointment times for all the patients in two phases is proposed. To obtain a high-quality schedule in reasonable computational time according to this policy, a hybrid VNS_LSHAPE algorithm is also developed. Numerical experiments based on the cardiovascular department of Tongji Hospital in Shanghai is conducted in this article. It is proved that the proposed jointly scheduling policy can reduce the waiting for both types of patients and improve the system’s efficiency at the same time.}, archive = {J_TASE}, author = {Na Li and Huangyu Chen and Zhi Pei and Tao Wang}, doi = {10.1109/TASE.2024.3363691}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1339--1352}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Jointly appointment scheduling in a two-phase service system with two types of patients considering multiple servers and stochastic service time}, volume = {22}, year = {2025}, }
• (2025). Fine-tuned convex approximations of probabilistic reachable sets under data-driven uncertainties. TASE, 22, 1319--1338. (www)

  @article{ , abstract = {This paper proposes a mechanism to fine-tune convex approximations of probabilistic reachable sets (PRS) of uncertain dynamic systems. We consider the case of unbounded uncertainties, for which it may be impossible to find a bounded reachable set of the system. Instead, we turn to find a PRS that bounds system states with high confidence. Our data-driven approach builds on a kernel density estimator (KDE) accelerated by a fast Fourier transform (FFT), which is customized to model the uncertainties and obtain the PRS efficiently. However, the non-convex shape of the PRS can make it impractical for subsequent optimal designs. Motivated by this, we formulate a mixed integer nonlinear programming (MINLP) problem whose solution result is an optimal $n$ sided convex polygon that approximates the PRS. Leveraging this formulation, we propose a heuristic algorithm to find this convex set efficiently while ensuring accuracy. The algorithm is tested on comprehensive case studies that demonstrate its near-optimality, accuracy, efficiency, and robustness. The benefits of this work pave the way for promising applications to safety-critical, real-time motion planning of uncertain dynamic systems. Note to Practitioners—This study is motivated by the realization of safety-critical real-time motion planning for a dynamic system under uncertainties. A popular method used to guarantee the safe operation of uncertain dynamic systems is reachability analysis. However, this method may not work well in the face of the following challenges: unbounded uncertainties, unknown distributions, generality, convexity, and efficiency. To address these issues, we first present a data-driven approach to model arbitrary unknown uncertainties and obtain a set encompassing the system states with high confidence; Then we propose an algorithm to efficiently find a tight convex polygon approximation for the set. This clearly benefits real motion planning. When considering collision avoidance in motion planning, a tight convex approximation allows a larger feasible search area which may provide a better-planned trajectory. Also, the efficiency of the algorithm ensures that the motion planning can be realized in real-time.}, archive = {J_TASE}, author = {Pengcheng Wu and Sonia Martinez and Jun Chen}, doi = {10.1109/TASE.2024.3363624}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1319--1338}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fine-tuned convex approximations of probabilistic reachable sets under data-driven uncertainties}, volume = {22}, year = {2025}, }
• (2025). Direct edge-to-edge attention-based multiple representation latent feature transfer learning. TASE, 22, 1305--1318. (www)

  @article{ , abstract = {Deploying a large number of smart cameras and training their models is a very time-consuming and labor-intensive process. Although there have been studies that utilized direct edge-to-edge (e2e) feature transfer learning without requiring help from centralized servers, they do not make good use of advanced latent features and even require constant adjustment of certain hyperparameters. This not only takes a lot of time for initial model training but also does not meet one of the key characteristics of the Internet of Thing, i.e., reduced human intervention. Therefore, we propose a Lightweight Multi-representation Attention-based Residual (LiMAR) module to reduce the computational load on an edge device via its lightweight architecture, which can extract multi-representation and diverse features to make the best use of the latent features for transfer learning in a direct e2e manner. In addition, we also propose a Lightweight Joint-distribution Autonomous Domain Adaptation (LiJADA) module to further reduce the load of model training and reduce the time required for camera deployment and the necessity of human intervention. The experimental results show that incorporating these two modules with one-to-many transfer can improve the total accuracy by 6.30% compared to a baseline study, yet with 75% less transmission data. In addition, the two modules working with many-to-one transfer learning can reduce the transmission cost by 83.33% and improve the total accuracy by 5.10%. Note to Practitioners—This study was motivated by the practical problem of enabling an unmanned application to efficiently learn its initial models. This problem is important for practitioners in unmanned areas who aim to reduce the effort required for system deployment by utilizing direct edge-device to edge-device (or edge-to-edge) model transfer learning. Existing approaches mainly rely on basic feature-level transfer learning, and they often require constant adjustment of hyperparameters to achieve optimal model performance. To tackle these challenges, we propose a novel approach that involves transferring multi-representation features with automatic hyper-parameter adjustment. This approach can improve knowledge reusability across different environments, and it is also important for real-life scenarios. Preliminary evaluations suggest that the approach is feasible and can achieve knowledge reuse while reducing data transmission costs. However, since we still use raw input images for matching the source and target cameras rather than using features for matching, this may limit non-source-free applications. To address this limitation, incorporating model-based transfer in the next-phase implementation can mitigate the drawback.}, archive = {J_TASE}, author = {Yung-Chen Tsai and Ching-Hu Lu}, doi = {10.1109/TASE.2024.3363618}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1305--1318}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Direct edge-to-edge attention-based multiple representation latent feature transfer learning}, volume = {22}, year = {2025}, }
• (2025). Fast adaptation trajectory prediction method based on online multisource transfer learning. TASE, 22, 1289--1304. (www)

  @article{ , abstract = {Conventional deep learning-based trajectory prediction methods always adopt offline training based on trajectory data collected in known scenes. Despite its high prediction accuracy, it is unable to process trajectory data acquired in real-time, making it non-trivial to adapt to unknown scenes. To mitigate the above problem, an online multi-source transfer learning-based pedestrian trajectory predictor, dubbed OMTL-PTP, is proposed to achieve fast adaptation of trajectory prediction. OMTL-PTP resorts to online transfer learning to transfer trajectory knowledge from multiple source domains to the target domain, enabling the model to learn from the new scene and continuously improve its trajectory prediction ability. Concretely, we propose several base learners with external memory modules to preserve source domain trajectory knowledge for online knowledge transfer. A multi-hop attention mechanism is introduced in each learner to handle the future uncertainty of generated trajectories. To fully utilize the knowledge from multiple source domains, OMTL-PTP leverages ensemble learning to transfer knowledge from multiple base learners in the source domains to the online learner and fine-tunes the online learner in the target domain. Specifically, all base learners are combined to update the online learner, improving its ability to process future arriving samples and adapt to unknown scenes quickly. Qualitative and quantitative evaluations on ETH/UCY indicate the effectiveness of OMTL-PTP in online learning, which is beneficial for deploying trajectory prediction methods on intelligent edge devices. The code will be released at https://github.com/zjrcczu/OMTL-PTP after acceptance. Note to Practitioners—This paper is motivated by the challenge of online sustained trajectory prediction for unmanned autonomous agents, but it also applies to other trajectory prediction tasks, such as intelligent monitoring. Existing approaches always collect trajectory data from different scenes for training, making the model generalize to other scenarios. However, they may suffer from performance degradation since they cannot learn trajectory knowledge from unknown scenes. This paper suggests a new approach by transferring trajectory knowledge from known scenes to unknown scenes and gradually learning from unknown scenes, inspired by online transfer learning. In this paper, we propose a trajectory predictor based on a memory network and introduce the multi-hop attention mechanism to mitigate future uncertainty of trajectory prediction. We then show how the external memory can preserve trajectory knowledge, which facilitates transferring knowledge from source domains to the target domain. Afterward, we train an online trajectory predictor based on online multi-source transfer learning, improving the generalization and adaptability of trajectory prediction models in unknown scenes and facilitating deploying trajectory prediction models in edge devices. This method also applies to other neural network-based regression tasks that require online sustained learning. In future research, we will improve the trajectory prediction performance while maintaining the online learning ability.}, archive = {J_TASE}, author = {Biao Yang and Junrui Zhu and Zhitao Yu and Fucheng Fan and Xiaofeng Liu and Rongrong Ni}, doi = {10.1109/TASE.2024.3362980}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1289--1304}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fast adaptation trajectory prediction method based on online multisource transfer learning}, volume = {22}, year = {2025}, }
• (2025). A sensor-enhanced hierarchical torque vector control for distributed mobile robot on slippery road. TASE, 22, 1276--1288. (www)

  @article{ , abstract = {Hierarchical torque vector control has proven effective in enhancing stability and handling in vehicles; however, its application in mobile robots faces challenges due to dynamic model variations and hardware differences. In response, this paper introduces a novel framework that leverages rich sensor information to optimize torque vector control in mobile robots. In this paper, we establish a comprehensive total torque parameter-varying error (TTPE) model at the upper level, effectively decoupling the yaw moment generated by the steering angle from the additional yaw moment. To achieve accurate and stable control, we employ a TTPE model predictive controller (TTPE-MPC) that generates precise desired total torque. Furthermore, we introduce a slip loss objective function based on kinematics in the middle layer, in addition to the conventional torque error and tire load objective functions. We enhance the optimization of tire load by incorporating information on the six-dimensional leg force, which significantly improves maneuverability and tracking precision during the torque distribution stage. To further enhance precision and real-time performance of torque distribution, a disturbance observer is designed for each individual wheel to estimate the tire’s longitudinal force at the lower layer. The control law of the motor input current is then designed based on this estimation, resulting in improved control accuracy. The proposed framework is validated through experiments conducted on a six-wheeled robot. The results demonstrate that the stability of the expected value of the total moment improves, with tracking errors of velocity and angular velocity reduced by approximately 43%. Moreover, the slip loss between wheels is reduced by approximately 46% compared to conventional methods. Note to Practitioners—This article aims to improve robotic steering maneuverability and provide safer, stable solutions for autonomous vehicles with independent propulsion. Current methods for robot operation often lack optimal torque management, primarily focusing on direct speed control while neglecting valuable sensor data. We propose a model-based, hierarchical control strategy to efficiently allocate torque to the robot’s propulsion motors. Our approach establishes a model that distinguishes between angular difference and motor torque, employing a Total Torque Parameter-Varying Error Model Predictive Controller (TTPE-MPC) for cumulative torque calculation. Optimization metrics, with a focus on sensor data, guide torque allocation, taking anti-slip considerations into account. Furthermore, a speed-based control observer ensures fair torque distribution, theoretically converging expected, estimated, and true values. Initial robot experiments validate reduced speed and angular velocity errors, supporting the effectiveness of our approach. Future work will explore incorporating suspension system aerodynamics into torque distribution, further enhancing stability.}, archive = {J_TASE}, author = {Wang Liang and Junfeng Xue and Wang Shoukun and Tianwei Niu and Wang Junzheng}, doi = {10.1109/TASE.2024.3362979}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1276--1288}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A sensor-enhanced hierarchical torque vector control for distributed mobile robot on slippery road}, volume = {22}, year = {2025}, }
• (2025). Automatic robotic cranium-milling: A motion control study of in vitro animal experiments. TASE, 22, 1262--1275. (www)

  @article{ , abstract = {Autonomous robotic surgery offers enhanced effectiveness, precision, and reliability, regardless of the surgeons’ expertise. Prior neurosurgery robot studies involved surgeons manually assisting the robot in skull-milling tasks by holding the milling cutter shank, constraining the robot’s autonomy. A model-free adaptive nonlinear force control algorithm is designed to accomplish automatic cranial-milling tasks. Furthermore, a skull-milling breakthrough detection algorithm by monitoring the change of feed force is proposed to determine the completion of the milling task autonomously. A robotic system is developed for automatic cranium-milling and 72 in vitro skull-milling experiments indicate that when using the proposed control algorithm, the maximum root mean square error percentage of the vertical force is 0.99%, while the control error percentages of other mainstream methods are all above 5.5%. Moreover, the success rate of breakthrough detection is 98.61% and the robot autonomously performs the skull milling task with minimal human intervention during the whole experiment. The results demonstrate that the proposed method provides the potential to improve the intelligence of neurosurgery. Note to Practitioners—The purpose is to propose a model-free adaptive nonlinear force control method for automatic skull-milling tasks. In previous studies, the involvement of surgeons manually holding the milling cutter shank to assist neurosurgery robots in skull-milling tasks has been observed. However, the autonomy of the robot is restricted and its potential for precise control tasks is failed to leverage. Therefore, a model-free adaptive nonlinear force control algorithm is proposed and a robotic system is built to enable the robot to autonomously perform cranial-milling tasks in this work. This application aims to enhance the autonomy of robot-assisted neurosurgery, making it a potential solution for remote surgery and addressing the shortage of medical resources in rural areas.}, archive = {J_TASE}, author = {Gui-Bin Bian and Chen Qian and Zhen Li and Qiang Ye and Peicong Ge and Jizong Zhao}, doi = {10.1109/TASE.2024.3362932}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1262--1275}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Automatic robotic cranium-milling: A motion control study of in vitro animal experiments}, volume = {22}, year = {2025}, }
• (2025). Ontology versioning for managing inconsistencies in engineering models arising from model changes in the design of intralogistics systems. TASE, 22, 1249--1261. (www)

  @article{ , abstract = {The interdisciplinary design of intralogistics systems (ILS) involves engineers from various disciplines, resulting in the generation of discipline-specific model files with overlapping information. For instance, a conveyor system can be represented from various perspectives, such as 3D-CAD models that capture its geometric information and discrete-event simulation models that depict the system’s dynamic material flow performance. The growing demands for flexible reconfigurability and adaptability in intralogistics systems necessitate frequent updates to engineering models. However, these updates often result in potential model inconsistencies due to insufficient stakeholder communication. Detecting the impact of model changes and related inconsistencies is challenging in practice due to data heterogeneity and complex inter-model relations. To address these challenges, we propose an ontology-versioning approach that automates the identification of inconsistencies resulting from model changes. Our approach facilitates the integration of heterogeneous model data, enables database versioning, detects inconsistencies caused by model updates, and provides traceability for identified issues. The concept is evaluated utilizing models from a prototypical implementation on a lab-sized demonstrator. Note to Practitioners—In the industry, the current development of intralogistics systems often lacks automated synchronization of overlapping model information and consistent model interfaces, frequently leading to contradictions among the models. This has been identified as a significant source of errors in the design of both industrial and academic intralogistics systems, as revealed by a study involving intralogistics experts from different technical disciplines. Effectively managing model inconsistencies is crucial for project success, particularly when frequent model changes occur. A promising approach to tackle this issue is to systematically link model data from different disciplines, through which model inconsistencies caused by inadequate communication among engineers can be identified and prevented. However, in many cases, changes in different model versions and their resulting inconsistencies are not adequately considered. To address this issue, we propose a concept based on ontology versioning that allows for the generation, comparison, and analysis of different versions of an ontological model database. This concept automatically identifies model changes, assesses their impacts on other models, and provides information to assist engineers in problem-solving. The effectiveness of our approach is assessed through an evaluation of three representative change scenarios, simplified from real-world use cases. In future research, we plan to extend the approach to general production systems and incorporate industrial-scale models from the broad range of disciplines involved in the design process.}, archive = {J_TASE}, author = {Fan Ji and Birgit Vogel-Heuser and Rafael Schypula and Maximilian Wünnenberg and Michael Goedicke and Johannes Fottner}, doi = {10.1109/TASE.2024.3362599}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1249--1261}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Ontology versioning for managing inconsistencies in engineering models arising from model changes in the design of intralogistics systems}, volume = {22}, year = {2025}, }
• (2025). Distributed MPC for cooperative tracking periodic references of heterogeneous systems. TASE, 22, 1233--1248. (www)

  @article{ , abstract = {This paper investigates a distributed model predictive control (DMPC) for linear heterogeneous systems tracking arbitrary periodic references. The control objective consists of two parts: (i) driving the output of each subsystem consensus; (ii) steering the outputs as close as possible to an exogenous periodic reference. The artificial state reference and control input are considered as decision variables to track unreachable references. The optimal control problem (OCP) is then solved in a distributed manner using Alternating Direction Multiplier Method (ADMM). The proposed method does not need ADMM convergence at each time step, which greatly reduces the computation time. Under several mild assumptions, the feasibility of the OCP and the closed-loop asymptotic stability with respect to an optimal reachable cooperative trajectory are presented. The performance of the approach is demonstrated with some simulation results. Note to Practitioners—The paper is motivated by the problem of cooperative tracking unreachable references for heterogeneous systems. The generation of the reference signal often ignores the dynamics feature of systems, leading to such reference may not be fully tracked (unreachable reference). However, existing methods either lack optimality or cannot achieve cooperative tracking of unreachable references. Therefore, this study develops a novel DMPC approach to make up for the above lack. In addition, the proposed method greatly reduces the computation time while ensuring optimality, and has a wider initial feasibility. The proposed controller can be extended to cooperative track unreachable constant signal. The proposed method can be used for highly collaborative tasks, such as formation missions, collaborative transportation and spacecraft collaboration. In future research, we will address the problem of cooperative tracking unreachable references for nonlinear heterogeneous systems.}, archive = {J_TASE}, author = {Yunshan Deng and Yuanqing Xia and Zhongqi Sun and Li Dai and Bing Cui}, doi = {10.1109/TASE.2024.3362470}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1233--1248}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed MPC for cooperative tracking periodic references of heterogeneous systems}, volume = {22}, year = {2025}, }
• (2025). Auxiliary knowledge-based fine-tuning mechanism for industrial time-lag parameter prediction. TASE, 22, 1220--1232. (www)

  @article{ , abstract = {Compared with the optimization technology of retraining models, fine-tuning pretrained neural networks have been widely used in industrial process monitoring because of their high precision and low training cost. However, data privacy protection in industry makes it impossible to improve the upstream pertaining model to avoid negative transfer when the target domain and the source domain are too different or the pertaining model is maliciously attacked. To solve the above problem, we proposed a fine-tuning method assisted by target domain knowledge for industrial time-lag parameter prediction. This method uses an Auto-Encoder to learn the representation of black-box and time-lag knowledge in the target domain. Then, the black-box and time-lag knowledge are used as auxiliary information to update the high-level weights of the pretrained network. At the same time, we proposed an auxiliary learning method that can dynamically update weights without introducing new parameters and provide training methods for different neural network optimizers. The experimental results on the heating furnace temperature prediction and wind condition prediction of wind farms demonstrate that the prediction performance can be effectively improved, and the defective inheritance of the pretrained model can be effectively reduced to avoid negative transfer. Note to Practitioners—Due to the complexity of industrial processes, the detection of key parameters is time-consuming and it is difficult to obtain labeled samples. With the widespread use of large models in computer vision and natural language processing, fine-tuning pretrained models can alleviate these challenges. However, the problem that comes with it is safety. The industrial prediction model cannot be built when the publicly available pretrained model is attacked. Moreover, the pretrained model could not be modified because the pretrained data could not be obtained. Therefore, this paper used the time-lag prior knowledge of industrial data to modify the pretrained model in the fine-tuning stage, and helps improve the pretrained accuracy and model robustness by mining the knowledge of existing data. Two groups of experiments show that the auxiliary knowledge-based fine-tuning mechanism can significantly improve the prediction accuracy and alleviate the defect inheritance of the pretrained model, which is valuable for the application of large models in the field of industrial parameter prediction.}, archive = {J_TASE}, author = {Naiju Zhai and Xiaofeng Zhou and Shuai Li and Haibo Shi}, doi = {10.1109/TASE.2024.3362322}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1220--1232}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Auxiliary knowledge-based fine-tuning mechanism for industrial time-lag parameter prediction}, volume = {22}, year = {2025}, }
• (2025). Dynamic execution of engineering processes in cyber-physical systems of systems toolchains. TASE, 22, 1208--1219. (www)

  @article{ , abstract = {Engineering tools support the process of creating, operating, maintaining, and evolving systems throughout their lifecycle. Toolchains are sequences of tools that build on each others’ output during this procedure. The complete chain of tools itself may not even be recognized by the humans who utilize them, people may just recognize the right tool being used at the right place in time. Modern engineering processes, however, do not value such ad-hoc choice of tooling, because of their uncontrolled nature. Building upon the Extended Automation Engineering Model defined by the IEC 81346 standard, this paper proposes to automate the toolchain building and execution process for Cyber-Physical System of Systems (CPSoS), utilizing key principles of the Eclipse Arrowhead framework. The proposed toolchain automation solution addresses issues such as tool interoperability, interaction, automation, and dynamic choreography. The feasibility of this set of integrated concepts is validated through an Arrowhead-based toolchain choreography demonstration. Note to Practitioners—The paper discusses approaches to the automated execution of various industry-related processes. As the processes are becoming more complex and involve numerous systems which have to be orchestrated, a simple and preprogrammed workflow is not enough anymore. Therefore, building on top of the principles of the Eclipse Arrowhead framework, an adequate model of toolchains, allowing for their automated execution, is proposed. Different approaches to supervision of toolchain execution are discussed showing the benefits of reaching higher automation levels. Further, four adoption levels are introduced, which are a measure of the toolchain automation progress. Finally, a simplified demonstrator is shown and steps to elevate it to higher adoption levels are highlighted. To ensure that the approach is industry-oriented, several examples of how the proposed methodology can be used in the industrial context are discussed.}, archive = {J_TASE}, author = {Federico Montori and Marek S. Tatara and Pál Varga}, doi = {10.1109/TASE.2024.3362132}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1208--1219}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic execution of engineering processes in cyber-physical systems of systems toolchains}, volume = {22}, year = {2025}, }
• (2025). Computer-assisted automatic preoperative path planning method for pelvic fracture reduction surgery based on enlarged RRT* algorithm. TASE, 22, 1196--1207. (www)

  @article{ , abstract = {Pelvic fracture reduction surgery (PFRS) has always been one of the most challenging procedures in trauma orthopedics. Excellent preoperative planning is crucial for surgery, especially with the increasingly mature robot-assisted surgical systems. However, current preoperative reduction planning heavily relies on surgeons’ experience. This paper proposes an automatic preoperative planning framework for PFRS. Firstly, an enlarged RRT $^{\ast} $ (ERRT $\ast$ ) algorithm is proposed to search feasible paths, which adopts a synchronized exploration and asynchronous adjustment strategy in 6D space to change the position and orientation of fragments during the reduction process. Secondly, a collision detection method based on surface point cloud is proposed to improve the safety of the reduction path by taking into account the actual volume of fragments. Finally, a post-processing method combining path shortening (PS) algorithm and cubic spline interpolation is proposed to optimize and smooth the reduction path. The clinical case simulation results show that the ERRT $^{\ast} $ algorithm can find a feasible reduction path within a few seconds ( $ < 10\text{s}$ ), and the length of the path is reduced by an average of 11.11% with the PS algorithm. Furthermore, repeated experimental results demonstrate that the method has good consistency. The proposed preoperative planning method can serve as a powerful tool to provide references for surgeons and also provide a quantitative basis for robot-assisted PFRS. Note to Practitioners—This work addresses the challenge of automating preoperative planning for pelvic fracture reduction surgery, specifically in determining the target reduction pose and operative path for repositioning the fragment. The motion of the fragment is decoupled into translation along a point and rotation through the coordinate system of that point to quantify the difference between the initial pose and target pose and determine the planning requirements. An enlarged RRT $^{\ast} $ algorithm based on six-dimensional generalized coordinates is proposed, combined with an efficient collision avoidance algorithm, to quickly find a safe and feasible reduction operation path. The proposed planning method not only provides guidance to physicians but also establishes a basis for robot-assisted fracture reduction surgery.}, archive = {J_TASE}, author = {Shaolin Lu and Shibo Li and Lihai Zhang and Xiaozhi Qi and Yuanyuan Yang and Bing Li and Ying Hu}, doi = {10.1109/TASE.2024.3361990}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1196--1207}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Computer-assisted automatic preoperative path planning method for pelvic fracture reduction surgery based on enlarged RRT* algorithm}, volume = {22}, year = {2025}, }
• (2025). Probabilistic forecasting framework oriented to distribution networks and microgrids. TASE, 22, 1183--1195. (www)

  @article{ , abstract = {In electrical distribution networks an adequate management is key for supporting the deployment of renewable generation sources and microgrids while extracting their maximum potential. Among the existing optimization approaches, stochastic and probabilistic methods are experiencing a growth in their use. However, one of the problems when applying these approaches is the complexity of creating and evaluating the quality of the required stochastic forecasts compared to deterministic forecasts. To mitigate this difficulty, this paper proposes a probabilistic forecasting framework that integrates model creation, their evaluation, and the selection of the best model for predicting. Additionally, two novel methods are proposed for creating scenario sets, and a new metric is defined for evaluating and selecting which model to use. The proposed framework is applied in a case study over a dataset of ten secondary distribution substations from a real distribution network located in Manzanilla (Spain), showing the effect of the selection criteria over the forecasting quality. Note to Practitioners—This article was motivated by the challenge of probabilistic forecasting inclusion in automatic management systems applied to power distribution networks and microgrids. Modern stochastic management optimization methods are fed with probabilistic forecasts, which offer richer information than classic deterministic forecasting. Therefore, the management systems should be able to automatically train a certain number of forecasting models (e.g., machine learning models), evaluate and compare them, and apply the best ones for obtaining the forecasts to feed the management optimization system. Considering the variety of models, techniques, probabilistic forecast types, and evaluation metrics, it can be unclear how to perform this process. For these reasons, this article proposes a probabilistic forecasting framework that integrates methods for the construction of diverse types of predictions (quantiles, intervals, and scenario sets), their evaluation, and the selection of the best model for performing each required prediction for feeding optimization systems. This framework could help to facilitate the implantation of modern stochastic optimization management systems for distribution networks and microgrids, as it simplifies the forecasting process.}, archive = {J_TASE}, author = {Antonio Parejo and Sebastián García and Enrique Personal and Juan Ignacio Guerrero and Alejandro Carrasco and Carlos León}, doi = {10.1109/TASE.2024.3361651}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1183--1195}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Probabilistic forecasting framework oriented to distribution networks and microgrids}, volume = {22}, year = {2025}, }
• (2025). Fractional-order systems optimal control via actor-critic reinforcement learning and its validation for chaotic MFET. TASE, 22, 1173--1182. (www)

  @article{ , abstract = {Since the existence of fractional order dynamics, it is difficult to obtain an optimality equation to solve for fractional-order optimal control. In this paper, a fractional Hamilton-Jacobi-Bellman (HJB) equation based on error derivative is proposed, and a corresponding online learning algorithm is designed. The scheme can handle the optimal tracking problem for the $0< \alpha \leq 1$ order nonlinear systems. Since the traditional quadratic cost function is unbounded at infinite time and the optimal control derived from the discounted cost function fails to stabilize the system asymptotically, a cost function based on the error derivative is proposed, which can avoid these problems, and the system is not restricted to be zero equilibrium. Then, the fractional HJB equation is derived by constructing an auxiliary signal without directly using the chain rule of differentiation. The optimality, stability and convergence of its solution are proved, and actor-critic neural networks (NNs) are established to perform the RL algorithm. Finally, the algorithm is applied to a chaotic magnetic-field electromechanical transducer (MFET) system to verify the effectiveness and advantages. Note to Practitioners—In engineering, fractional-order dynamics properties are used in many practical systems such as chaotic MFET systems, chaotic arch micro-electro-mechanical systems and fractional order resonant controllers, etc. For these fractional-order systems, traditional integer-order control methods cannot be applied. Moreover, optimal performance with minimal cost/resources can be achieved through optimal control. Therefore, a intelligent control method based on reinforcement learning is proposed in this paper to realize the optimal control for fractional-order systems. The proposed method is applicable to most fractional-order systems and can achieve stable control while optimizing the objective performance and reducing energy consumption. Finally, the proposed algorithm is successfully applied to the chaotic MFET systems.}, archive = {J_TASE}, author = {Dongdong Li and Jiuxiang Dong}, doi = {10.1109/TASE.2024.3361213}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1173--1182}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Fractional-order systems optimal control via actor-critic reinforcement learning and its validation for chaotic MFET}, volume = {22}, year = {2025}, }
• (2025). Attack tolerant fault detection for CPSs: An unknown input interval observer approach. TASE, 22, 1163--1172. (www)

  @article{ , abstract = {The interval observer, recognized as a potent tool for fault detection (FD), diverges from the adoption of fixed thresholds to enhance the timeliness and precision of detection. This study endeavors to design an FD mechanism for fuzzy cyber-physical systems (CPSs) subjected to adversarial influences based on improved attack tolerant interval observer. Specifically, the proposed frequency-information-based unknown input interval observer (UIIO) not only enables a more precise estimation of the interval range under the influence of attack signals but also isolates unknown decoupled inputs. Furthermore, it facilitates specific performance design for signals within a particular frequency range. Following the stability analysis of the system, linear solvable conditions are presented to ensure the robustness, fault sensitivity of the augmented system, as well as non-negativity of the augment system matrix. Finally, the proposed attack-tolerant FD mechanism is validated through simulation examples, including the influence of the prior attack information to detection interval width and the detection performance for small-amplitude faults or faults with specific frequency characteristics. Note to Practitioners—CPSs constitute the core of the next-generation manufacturing industry, applicable in areas like unmanned autonomous systems, smart grids, and intelligent healthcare, capitalizing on wide-ranging cyber space utilization. Nonetheless, this breadth also exposes them to elevated risks, including the threat of actuator attacks. The novel fault detection scheme developed in this paper is capable of operating in real-time when the system is confronted with various threats such as disturbances, faults, and actuator attacks. This approach obviates the requirement for knowledge about the nature of actuator attack signals, while enabling the generation of a system state estimation interval with a tunable error range. Leveraging frequency-dependent techniques, the developed scheme, even in the presence of potential actuator attack and disturbance, is particularly well-suited for the detection of fault signals with specific frequency domain characteristics.}, archive = {J_TASE}, author = {Qidong Liu and Yue Long and Tieshan Li and C. L. Philip Chen}, doi = {10.1109/TASE.2024.3360967}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1163--1172}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attack tolerant fault detection for CPSs: An unknown input interval observer approach}, volume = {22}, year = {2025}, }
• (2025). Outlier-resistant distributed filtering over sensor networks under dynamic event-triggered schemes and DoS attacks. TASE, 22, 1152--1162. (www)

  @article{ , abstract = {This paper focuses on the problem of outlier-resistant distributed filtering under dynamic event-triggered scheme (DETS) and cyber-attacks over sensor networks. A novel distributed DETS is proposed, which allows us to minimize the data-releasing rate while simultaneously preventing abrupt data from being sent over the network. Consider the situation of measurement outliers in presence, a saturation function is used in distributed filter design to confine the contaminated signals so as to improve filtering performance. The occurrence of stochastic DoS attacks which may corrupt the measurement in the sensor networks are also considered. Sufficient conditions for the expected filter are derived while guaranteeing the asymptotically stability of the filtering error system with a guaranteed $H_{\infty }$ performance index $\gamma $ . A numerical simulation is given to illustrate the effectiveness of the theoretical analysis and design method. Note to Practitioners—The issue of filtering over sensor networks is subject to network impections induced by the limited communication resources and cyber-attacks. Since these imperfections may lead to performance degration or instability of the system, it is critical to take some measures to reduce unnecessary signal communication among sensor nodes and eliminate the adverse effects of cyber-attacks. Besides, the measurement outliers are frequently encountered due to sensor aging or abnormal disturbances, which may impair the accuracy of the estimation resuts. In this paper, the outlier-resistant distributed filtering problem is investigated in sensor networks subject to dynamic event-triggered schemes and DoS attacks. A co-design method of the DETS and the distributed filter is propsed which can gurantee the desired system performance.}, archive = {J_TASE}, author = {Lijuan Zha and Rongfei Liao and Jinliang Liu and Xiangpeng Xie and Engang Tian and Jinde Cao}, doi = {10.1109/TASE.2024.3360718}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1152--1162}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Outlier-resistant distributed filtering over sensor networks under dynamic event-triggered schemes and DoS attacks}, volume = {22}, year = {2025}, }
• (2025). Event-triggered set-membership filtering for active power distribution systems under fading channels: A zonotope-based approach. TASE, 22, 1139--1151. (www)

  @article{ , abstract = {This paper is concerned with the set-membership filtering problem for active power distribution systems that are influenced by unknown but bounded noises. Both the phenomena of fading channels and limited communication capacity are taken into account. In consideration of the integration of photovoltaic generation systems, an active power distribution system model is formulated which encompasses the conventional power distribution networks and distributed power sources. Network data are transmitted to a remote filter through fading channels, where a component-based dynamic event-triggered mechanism is introduced, by which the transmission frequency is reduced while sustaining the filtering performance, thereby mitigating the transmission load of the communication network. The purpose of this paper is to design a dynamic event-triggered filter such that, when faced with unknown but bounded noises, a set of zonotopes is devised to confine the system states. By minimizing the $F$ -radius of these zonotopes, the time-varying filter gain is determined recursively at each time step. Additionally, simulation experiments on the IEEE 34 distribution test system are carried out, through which the efficiency of the proposed filtering methodology is validated. Note to Practitioners—The rapid increase in energy demand, coupled with the fast progression of energy technologies, has resulted in the widespread merging of traditional distribution networks with various other distributed power systems. Consequently, these have transformed into active power distribution systems (APDSs), undergoing notable alterations in their distribution network configurations. Challenges like power back-flow and overvoltage, which might seriously compromise the stability, operation, and control of the APDS, can be introduced by this transformation. For addressing these challenges, having precise information about the system state is deemed crucial for the facilitation of real-time monitoring, improved control, and reliable protection of the APDS, and therefore the state estimation problem has been attracting an increasing research interest for ensuring the system’s safe operation and economical dispatch. In this paper, the problem of zonotopic set-membership filtering is examined for the APDS with unknown-but-bounded noises in fading channels. Initially, a component-based dynamic event-triggered mechanism is used in the APDS, which eases the load on communication networks with limited resources, and the fading coefficients are represented as uncertain variables within a specified range. Subsequently, by harnessing mathematical induction and set theory, the state estimation algorithm is introduced, ensuring the real system states are strictly encompassed within zonotopic sets and system states are estimated accurately. In conclusion, simulation experiments on the IEEE 34 distribution test system demonstrate the effectiveness of the introduced filtering algorithm.}, archive = {J_TASE}, author = {Guhui Li and Zidong Wang and Xingzhen Bai and Zhongyi Zhao and Hongli Dong}, doi = {10.1109/TASE.2024.3360600}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1139--1151}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-triggered set-membership filtering for active power distribution systems under fading channels: A zonotope-based approach}, volume = {22}, year = {2025}, }
• (2025). Uncertainty-aware autonomous robot exploration using confidence-rich localization and mapping. TASE, 22, 1124--1138. (www)

  @article{ , abstract = {Information-based autonomous robot exploration methods, aiming to maximize the exploration rewards, e.g., mutual information (MI), get more prevalent in field robotics applications. However, most MI-based exploration methods assume known poses or use inaccurate pose uncertainty approximation, which may lead to deviation or even failure when exploring prior unknown environments. In this paper, we explicitly consider full-state (pose & map) uncertainty for balancing exploration and localizability, i.e., avoiding the robot guiding itself to complex scenes with high exploration rewards but hard to localize. We first propose a Rao-Blackwellized particle filter-based localization and mapping framework (RBPF-CLAM) for a dense environmental map with continuous occupancy distribution. Then we develop a new closed-form particle weighting method to improve the localization accuracy and robustness. We further use these weighted particles to approximate the unknown pose uncertainty and combine it with our previous confidence-rich mutual information (CRMI) metric to evaluate the expected information utility of the robot’s new control actions. This new information metric is called uncertain CRMI (UCRMI). Dataset experiments show our RBPF-CLAM improves about 44.7% average root mean square error than the state-of-the-art RBPF localization method, and real-world experimental results show that our UCRMI reduces the pose uncertainty about 32.85% more than CRMI and 25.36% time cost than UGPVR in the exploration of unknown and unstructured scenes given sparse measurements, which shows better performance than other state-of-the-art information metrics. Note to Practitioners—This work was motivated by the problem of ‘planning for state estimation’ for a range-sensing robot, i.e., the robot can choose a better future place to facilitate its localization more accurately and explore new areas rationally to gather more information. Existing methods mainly assume the robot’s poses during the exploration can be estimated by an independent localization approach or simply propagated via a predefined probabilistic distribution. However, localization failure would lead to higher planning deviation for the planner that does not consider the pose uncertainty, and manually set parametric distribution is more prone to overestimate the pose uncertainty. This paper proposes an RBPF-based localization and mapping scheme and an improved particle weight update method in a confidence-rich map, then uses the weighted particles to approximate trajectory entropy and combines it with CRMI to evaluate the expected information gain of a candidate action/node. Our newly defined information function ‘UCRMI’ can prevent the robot from exploring too aggressively without considering its localizability in prior unknown and unstructured environments. These scenes may lack robust features to conduct feature-based SLAM or lack accurate external localization information such as GPS. This method can be applied in underwater, planetary, and subterranean robot exploration tasks, even using low-resolution sensors. Future work mainly involves adapting UCRMI to applications in large-scale scenes using small autonomous platforms.}, archive = {J_TASE}, author = {Yang Xu and Ronghao Zheng and Senlin Zhang and Meiqin Liu and Junzhi Yu}, doi = {10.1109/TASE.2024.3360442}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1124--1138}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Uncertainty-aware autonomous robot exploration using confidence-rich localization and mapping}, volume = {22}, year = {2025}, }
• (2025). Interval estimation for delayed reaction-diffusion systems. TASE, 22, 1112--1123. (www)

  @article{ , abstract = {This paper considers the interval estimation for delayed linear reaction-diffusion systems. To obtain the estimates of the states for delayed reaction-diffusion systems, a novel interval estimation scheme is proposed based on a spatial finite difference method and decoupling technology. First, the delayed reaction-diffusion system is discretized by the finite difference method, resulting in approximated delay differential equations. The bounds of the discretization errors are presented via a rigorous analysis. Then, we involve decoupling techniques into interval observer design for approximated delay differential equations, and its estimates are obtained. Moreover, combining the observations with the discretization errors, we provide the interval estimation of the delayed reaction-diffusion systems. Furthermore, a controller is designed using the observations to achieve the $H_{\infty }$ performance. Finally, the proposed methods are validated by two numerical simulations. Note to Practitioners—This work is motivated by industrial problems instance of the thermal diffusion in CPU chips. Due to the significant cost of real-time measurement of chip’s temperature and the influence of environment, it is more feasible to obtain the thresholds of operating temperature based on known information and available measurements. To obtain the state thresholds, this paper focuses on designing interval observer and analysis of discretization error to guarantee the accuracy while reducing the impact of uncertainties. The methods in this paper is simple and fast to execute in practical engineering, and the discretization error does not need to be obtained through a priori numerical experiment. More importantly, the estimates include the worst-case during system operation, which is relatively practical in real applications. The proposed results aim to provide a helpful reference for controller synthesis and fault detection of spatio-temporal system with time delays, thereby promoting corresponding application research.}, archive = {J_TASE}, author = {Yu Gao and Kai-Ning Wu and Song Zhu}, doi = {10.1109/TASE.2024.3360247}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1112--1123}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Interval estimation for delayed reaction-diffusion systems}, volume = {22}, year = {2025}, }
• (2025). Asynchronous switching control for fuzzy markov jump systems with periodically varying delay and its application to electronic circuits. TASE, 22, 1100--1111. (www)

  @article{ , abstract = {This article focuses on addressing the issue of asynchronous ${H_{\infty} }$ control for Takagi-Sugeno (T-S) fuzzy Markov jump systems with generally incomplete transition probabilities (TPs). The delay is assumed to vary periodically, resulting in one monotonically increasing interval and one monotonically decreasing interval during each period. Meanwhile, a new Lyapunov-Krasovskii functional (LKF) is devised, which depends on membership functions (MFs) and two looped functions formulated for the monotonic intervals. Since the modes and TPs of the original system are assumed to be unavailable, an asynchronous switching fuzzy controller on the basis of hidden Markov model is proposed to stabilize the fuzzy Markov jump systems (FMJSs) with generally incomplete TPs. Consequently, a stability criterion with improved practicality and reduced conservatism is derived, ensuring the stochastic stability and ${H_{\infty} }$ performance of the closed-loop system. Finally, this technique is employed to the tunnel diode circuit system, and a comparison example is given, which verifies the practicality and superiority of the method. Note to Practitioners—As a category of stochastic hybrid nonlinear systems driven by continuous time and discrete events, FMJSs have significant applications in practical engineering such as aircraft control systems and large-scale manufacturing systems. However, the real-time acquisition of system mode information and TPs is difficult due to technological constraints and limited resources. Moreover, the presence of periodically varying delay is prevalent in many industrial processes, resulting in degradation of dynamic system performance and instability. Therefore, it is necessary to study the FMJSs with generally incomplete TPs and periodically varying delays in asynchronous framework. Accordingly, unlike previous work, the designed LKF relies on system modes, MFs, and two looped functions for monotonic intervals. This new LKF exhibits a high degree of flexibility, fully leveraging the information of MFs and periodically varying delays, which can significantly reduce conservatism.}, archive = {J_TASE}, author = {Yinghong Zhao and Likui Wang and Xiangpeng Xie and Hak-Keung Lam and Junhua Gu}, doi = {10.1109/TASE.2024.3360036}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1100--1111}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Asynchronous switching control for fuzzy markov jump systems with periodically varying delay and its application to electronic circuits}, volume = {22}, year = {2025}, }
• (2025). GSNN: A neuromorphic computing model for the flexible path planning in various constraint environments. TASE, 22, 1081--1099. (www)

  @article{ , abstract = {Spiking Neural Networks (SNNs) represent a new generation of artificial neural networks that draw inspiration from biological systems. However, due to the intricate dynamics they exhibit and the discontinuity inherent in spike signals, SNNs often encounter performance limitations when addressing optimization problems. In this paper, we introduce the Graph-connected Spiking Neural Network model (GSNN), an extension of the SNN framework. The GSNN model holds the potential for integration with various existing path planning methods, rendering it applicable to a wide array of common path planning tasks. We specifically present two fundamental models within the GSNN framework. The first model employs GSNN to extract heuristic information from constrained pixel maps. This extracted data is then amalgamated with a novel sampling method, resulting in enhanced planning efficiency when compared to conventional techniques. The second model leverages GSNN to map a weighted graph, effectively utilizing plasticity methods to ascertain the shortest path within the graph. Moreover, this model facilitates path planning under diverse constraint environments, encompassing dynamic considerations, cost-awareness, and the collision dimensions of moving objects. Recognizing that the size of pixel maps or the number of nodes within weighted graphs might constrain GSNN’s capabilities, we propose a partitioning strategy to address this limitation. Empirical results unequivocally demonstrate the superiority of both GSNN models in resolving static path planning problems. Furthermore, the second GSNN model demonstrates rational performance across various constrained scenarios. Note to Practitioners—The primary motivation behind this study is to explore the utilization of neural morphic computation methods in addressing path planning challenges. To accomplish this, we introduce an encompassing model named GSNN. Given the limited coverage of neural morphic computation within this field, we present a comprehensive overview of its potential applications, with the intention of providing a valuable reference for both researchers and practitioners. Moreover, GSNN has the capability to seamlessly integrate with existing advanced optimization methods, thereby leading to enhanced performance and the capacity to tackle even more intricate problems.}, archive = {J_TASE}, author = {Haojie Ruan and Yinghui Chang and Weikang Wu and Zenan Huang and Yabin Deng and Leilei Li and Hongyan Luo and Shan He and Donghui Guo}, doi = {10.1109/TASE.2024.3359641}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1081--1099}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {GSNN: A neuromorphic computing model for the flexible path planning in various constraint environments}, volume = {22}, year = {2025}, }
• (2025). Design and development of a teleoperated telepresence robot system with high-fidelity haptic feedback assistance. TASE, 22, 1069--1080. (www)

  @article{ , abstract = {This paper proposes a new teleoperated telepresence robot system with high-fidelity haptic feedback for interaction work, such as pre-operation of explosive ordnance disposal (EOD). The system includes a master device and a slave collaborative robot. The master device enables the human operator to feel the operation procedure when performing the environmental interaction task. The system’s kinematics and dynamics models are derived, laying the foundation for the control scheme design. To meet the requirements of EOD missions, a hybrid motion mapping method including position-velocity (PV) and position-position (PP) mapping modes is presented to achieve a balance between working efficiency and manipulation accuracy. A hybrid haptic force rendering method is introduced to facilitate the human’s feeling of the interacting force and control the slave robot in PP and PV mapping modes, respectively. Experimental results reveal that the developed system exhibits good position and velocity tracking performance with root-mean-square (RMS) errors of 0.004 m and 0.007 m/s, respectively. The haptic tracking is realized with an RMS force error of 0.14 N. Moreover, the developed system can improve the EOD working efficiency by 10.7% with accurate operation while ensuring the fidelity of the haptic feedback to feel the contact process. The reported telepresence robot system provides a promising solution to delicate remote interaction operations. Note to Practitioners—Teleoperated robotic systems are commonly utilized to perform hazardous tasks such as EOD tasks. However, the existing EOD robots controlled by the joystick have limited dexterity. Moreover, visual feedback alone cannot provide enough telepresence for an operator, which leads to low efficiency. This paper proposes a new teleoperated robot system with high-fidelity haptic feedback assistance intended for EOD tasks. The system integrates the motion mapping method that combines PP and PV modes to improve the working efficiency while maintaining manipulation accuracy. The haptic sensations feedback can switch the rendering mode according to different control modes. The haptic feedback mechanism enables the operators to feel reliable contact under PP mapping mode and ensures safety and stability under PV mapping mode. Performance testing and application experiment verified the feasibility of the reported robotic system.}, archive = {J_TASE}, author = {Kai Feng and Qingsong Xu and Seng Fat Wong and Bin Zi}, doi = {10.1109/TASE.2024.3359243}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1069--1080}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and development of a teleoperated telepresence robot system with high-fidelity haptic feedback assistance}, volume = {22}, year = {2025}, }
• (2025). Application of reinforcement learning-based adaptive PID controller for automatic generation control of multi-area power system. TASE, 22, 1057--1068. (www)

  @article{ , abstract = {This paper presents an application of an actor-critic reinforcement learning (RL) algorithm-based adaptive proportional-integral-derivative (PID) controller for automatic generation control of multi-area power systems. The proposed approach has several advantages over other deep RL algorithm-driven PID controllers, such as simplicity in structure, elimination of pre-learning requirements, and prior tuning of PID parameters. Online adaption of PID parameters is achieved through actor-critic policy. The proposed method implements a single radial basis function (RBF) based neural network for actor and critic networks. Three different case studies are demonstrated with proper illustration and analysis of the result to present the effectiveness and robustness of the proposed control strategy against various uncertainties. The outcomes of the proposed controller are compared with the conventional PID controller tuned by the Particle Swarm Optimization (PSO) algorithm. The results seem competent enough to maintain the frequency within an acceptable limit under various uncertainties. Note to Practitioners—This paper describes the application of an online adaptive PID controller for automatic generation control of power systems. The controller is designed using a model-free reinforcement learning algorithm, which enables it to control the system without requiring prior knowledge of the system dynamics. Additionally, the controller does not need any global optimization algorithm for tuning the parameters ( $K_{P}$ , $K_{I}$ , $K_{D}$ ) beforehand. This controller can be implemented for both linear and non-linear systems.}, archive = {J_TASE}, author = {Rasananda Muduli and Debashisha Jena and Tukaram Moger}, doi = {10.1109/TASE.2024.3359219}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1057--1068}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Application of reinforcement learning-based adaptive PID controller for automatic generation control of multi-area power system}, volume = {22}, year = {2025}, }
• (2025). Integrated model-based control allocation strategies oriented to predictive maintenance of saturated actuators. TASE, 22, 1045--1056. (www)

  @article{ , abstract = {Predictive Maintenance approaches are gaining popularity in the new Industry 4.0 paradigm as they offer superior benefits in terms of time and money savings when it is required to assess the current working capabilities of operating equipment to carefully schedule maintenance operations. This work deals with a control allocation strategy inspired by Model Predictive Control ideas and able to address the loss of effectiveness of actuating equipment arising from their continuous usage. The scheme here presented comprises two modules: a prognostic unit for monitoring the reliability conditions of the actuators and a re-configurable control allocation block that operates according to the deterioration degree of the present actuators. The benefits of the proposed approach are testified by the numerical simulations carried out on both an unstable system and a tanks network. In particular, it can be observed that the proposed method is capable of suggesting a time-window for maintenance interventions that prevents either stability or feasibility issues. Note to Practitioners—Nowadays, in the viewpoint of financial and technical issues in the industries, the predictive maintenance (PM) plays the main and important subject. In the variety of industries, such as oil, gas, petrochemical, power plant, or transmission and distribution infrastructures, which are using the actuators and pipelines, predictive maintenance (PM) is a critical topic to prevent the unwanted shutdown and unrequired maintenance costs. However, most PM research and solutions have paid attention mainly to the estimation of the Remaining Useful Life (RUL) of critical devices (i.e., pipelines) only. On the contrary, in this work we focus on the possibility of influencing the RUL by properly acting on the actuators’ effort. The proposed method consists in a dual-mode fault-tolerant control allocation technique developed for discrete-time systems subject to input saturation. The proposed predictive model-based control solution can be considered as a simple, structural and practical approach to minimize the maintenance cost and reliability by decreasing the number of maintenance interventions and optimizing the time for their repairing, by preventing or decreasing the probability of occurrence of unwanted failures events. The main limitation in this approach is that it assumes the RUL of the critical assets under investigation known. As a future research, this work will be extended to include a related RUL estimator hinging on online data.}, archive = {J_TASE}, author = {Mehdi Forouzanfar and Gianfranco Gagliardi and Francesco Tedesco and Alessandro Casavola}, doi = {10.1109/TASE.2024.3358912}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1045--1056}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated model-based control allocation strategies oriented to predictive maintenance of saturated actuators}, volume = {22}, year = {2025}, }
• (2025). DL-DRL: A double-level deep reinforcement learning approach for large-scale task scheduling of multi-UAV. TASE, 22, 1028--1044. (www)

  @article{ , abstract = {Exploiting unmanned aerial vehicles (UAVs) to execute tasks is gaining growing popularity recently. To address the underlying task scheduling problem, conventional exact and heuristic algorithms encounter challenges such as rapidly increasing computation time and heavy reliance on domain knowledge, particularly when dealing with large-scale problems. The deep reinforcement learning (DRL) based methods that learn useful patterns from massive data demonstrate notable advantages. However, their decision space will become prohibitively huge as the problem scales up, thus deteriorating the computation efficiency. To alleviate this issue, we propose a double-level deep reinforcement learning (DL-DRL) approach based on a divide and conquer framework (DCF), where we decompose the task scheduling of multi-UAV into task allocation and route planning. Particularly, we design an encoder-decoder structured policy network in our upper-level DRL model to allocate the tasks to different UAVs, and we exploit another attention-based policy network in our lower-level DRL model to construct the route for each UAV, with the objective to maximize the total value of executed tasks given the maximum flight distance of the UAV. To effectively train the two models, we design an interactive training strategy (ITS), which includes pre-training, intensive training and alternate training. Experimental results show that our DL-DRL performs favorably against the learning-based and conventional baselines including the OR-Tools, in terms of solution quality and computation efficiency. We also verify the generalization performance of our approach by applying it to larger sizes of up to 1500 tasks and to different flight distances of UAVs. Moreover, we also show via an ablation study that our ITS can help achieve a balance between the performance and training efficiency. Our code is publicly available at https://faculty.csu.edu.cn/guohuawu/zh_CN/zdylm/193832/list/index.htm. Note to Practitioners—Unmanned aerial vehicles (UAVs) are of great practical usage, as they have many real world applications. When a group of UAVs are employed to execute large-scale tasks, a core question is how to scheduling the UAVs, so that they could complete the tasks efficiently. However, it is a computationally hard problem due to the exponentially increasing search space. To solve this problem, we propose a double-level deep reinforcement learning (DL-DRL) approach within a divide-and-conquer framework (DCF), where the upper-level DRL model is responsible for the task allocation, and the lower-level DRL model is responsible for the UAV route planning. To better train the two DRL models who have interplay with each other, we propose a simple yet efficient training strategy, termed interactive training strategy (ITS), which includes pre-training, intensive training and alternate training. The experimental results based on instances of various scales show that our DL-DRL approach outperformed learning-based and conventional baselines, and the designed ITS could strike a good balance between performance and training efficiency. In light of those verified advantages, we believe that our DL-DRL approach has favorable potential to solve the practical task scheduling problem of multi-UAV in real world.}, archive = {J_TASE}, author = {Xiao Mao and Guohua Wu and Mingfeng Fan and Zhiguang Cao and Witold Pedrycz}, doi = {10.1109/TASE.2024.3358894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1028--1044}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DL-DRL: A double-level deep reinforcement learning approach for large-scale task scheduling of multi-UAV}, volume = {22}, year = {2025}, }
• (2025). A multiscale attention mechanism super-resolution confocal microscopy for wafer defect detection. TASE, 22, 1016--1027. (www)

  @article{ , abstract = {Confocal microscopy is an essential component of wafer defect detection systems. Wafers are raw materials used in the manufacture of semiconductor chips. The semiconductor chip manufacturing process undergoes frequent updates, which cause an increase in the number and types of defects. This leads to lengthy scanning times for large wafers, and warrants the need to enhance the throughput of optical microscopy inspections. To address this issue, we propose the use of the multi-scale residual dilated convolution attention mechanism network (MRDCAN) super-resolution reconstruction algorithm to reproduce high-resolution images from low-magnification objective lens acquired images. The algorithm introduces the attention mechanism to enhance the information richness of wafer images, introduces the multi-scale expansion convolution to expand the convolutional sensor field to eliminate artefacts to enrich the detailed information of wafer image contours, and meets the image quality requirements through the loss calculation method based on the combination of mean-square error (MSE) and structural similarity (SSIM) image evaluation indices. It is shown that the reconstruction of low-resolution wafer images using this algorithm breaks the optical diffraction limit and achieves the purpose of improving the wafer image resolution. Compared with state-of-the-art models, the proposed algorithm can achieve the best performance with an SSIM index of 94.26 percent for the reconstructed super-resolution wafer images. Our algorithm provides fresh insights into the current challenges of confocal microscopy in the field of wafer defect detection. Note to Practitioners—Shrinking semiconductor wafer sizes and increasingly complex inspection steps lead to reduced throughput of optical microscope inspection systems. Current convolutional neural network (CNN) networks cannot solve the problem of super-resolution of complex wafer images well. This seriously affects their application in practical detection. Compared with other algorithms, the super-resolution reconstruction algorithm proposed in this paper has a short training time and a multi-scale structure that effectively prevents the loss function curve from oscillating. And the reconstructed wafer image achieves obvious advantages in terms of visual effect and evaluation indices, with strong robustness to Gaussian noise. In addition, the final discussion shows that high-resolution images can be reproduced through the combination of low-magnification objective lens and deep learning super-resolution algorithm, which can simplify the steps of wafer defect detection and increase the efficiency of the whole wafer defect detection by more than 100%. This study demonstrates the potential of super-resolution confocal microscopy for wafer defect detection.}, archive = {J_TASE}, author = {Xuefeng Sun and Baoyuan Zhang and Yushan Wang and Jialuo Mai and Yuhang Wang and Jiubin Tan and Weibo Wang}, doi = {10.1109/TASE.2024.3358693}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1016--1027}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A multiscale attention mechanism super-resolution confocal microscopy for wafer defect detection}, volume = {22}, year = {2025}, }
• (2025). Improving the uniformity of the residual magnetic field in the MSR using independent coils compensation method. TASE, 22, 1006--1015. (www)

  @article{ , abstract = {The Magnetoencephalography (MEG) based on atomic magnetometer has high demand for the uniformity in the measurement area. Due to the door and holes of the magnetically shielding room (MSR), the distribution of residual magnetic field is non-linear, introducing much trouble into the magnetic field compensation. This study proposes a method for improving the uniformity of the residual magnetic field in the compact MSR by injecting different currents into the two coils in each pair of the outside compensation coils. The model and theoretical analysis method have been established, the feasibility has been verified both by the simulation and the experimental test. Results show that the maximum of the three components of the residual magnetic field in the direction of the three axes can be reduced by 18 times, 10.6 times and 8.1 times, respectively, when compared with the results without compensation, and 2.4 times, 2.3 times and 2.2 times, respectively, when compared with the results compensated with the same current. This study can provide certain help for achieving ultra-weak magnetic environment with high uniformity for the MEG measurement. Note to Practitioners—In this article, we focus on the problem of compensating the residual magnetic field in the compact MSR with high efficiency and low cost, which remains the key factor that limits the promotion of the MEG. Method based on the compensation coils in series outside the MSR only deals with the linear or symmetric components, but has nothing to do with the non-linear and asymmetric one caused by the door. We propose a new method for compensating the residual magnetic field based on the coupling between the active and passive shielding, high compensation efficiency and uniformity can be achieved. In future work, we will aim at applying this method to create an ultra-weak magnetic field environment with high uniformity for the MEG measurement.}, archive = {J_TASE}, author = {Zhouqiang Yang and Yanbin Li and Xikai Liu and Yupeng Li and Peiling Cui and Shiqiang Zheng and Xinxiu Zhou}, doi = {10.1109/TASE.2024.3358542}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1006--1015}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Improving the uniformity of the residual magnetic field in the MSR using independent coils compensation method}, volume = {22}, year = {2025}, }
• (2025). Frequency domain optimization design of the dual-loop controller for piezoelectric tube scanners with compound dynamics. TASE, 22, 996--1005. (www)

  @article{ , abstract = {The dual-loop controller (DLC) with both inner damping and outer tracking controllers has demonstrated exceptional performance in high-speed control of piezo-actuated nano-positioners. However, the accurate low-order model of the plant is imperative for designing the damping controller, according to the conventional DLC design principle. This limits its application in controlling piezoelectric tube scanners (PTSs) with compound dynamics. To handle this problem, this study introduces a frequency domain method for designing the DLC based on the frequency response data of the PTS. This method mitigates issues related to modeling errors. Specifically, the Nyquist diagram is employed to provide the stability boundary for parameters determination. A constraint optimization problem is formulated to achieve a high bandwidth with a flat amplitude frequency response. And the differential evolution algorithm is then adopted to find an optimal solution. Experimental validation on a PTS comfirms the effectiveness of this frequency domain design method. The results indicate that the control bandwidth of the optimized DLC achieves 848 Hz for a PTS with the first resonant frequency of 702 Hz. The superiorities of the DLC designed by the proposed optimization method are also validated via comparative tracking experiments involving step and triangular trajectories. Note to Practitioners—Frequency response data (FRD)-based methods open a new perspective to the controller design and optimization. Unlike traditional methods relying on transfer function models, the FRD of the system is directly identified and employed for the controller design and optimization, eliminating the need for extensive modeling and system identification efforts, thereby reducing errors. This paper proposes an FRD-based method for the simultaneous optimization of the DLC to achieve a high control bandwidth with a flat amplitude frequency response. This method relieves the requisite of accurate low-order model of the plant in conventional DLC designs. A closed-loop control bandwidth that exceeds the first resonant frequency is obtained for a PTS with complex high-order dynamic models. The proposed method enriches the scope of the DLC for the high bandwidth control, extending its applications to the systems with compound dynamics, which is significant to the high-speed control tasks of nano-positioners, such as the atomic force microscope imaging.}, archive = {J_TASE}, author = {Yixuan Meng and Minyu Pan and Linlin Li and Xiangyuan Wang and Zhiwei Zhu and Limin Zhu}, doi = {10.1109/TASE.2024.3357908}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {996--1005}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Frequency domain optimization design of the dual-loop controller for piezoelectric tube scanners with compound dynamics}, volume = {22}, year = {2025}, }
• (2025). Data-efficient learning control of continuum robots in constrained environments. TASE, 22, 984--995. (www)

  @article{ , abstract = {This research investigates learning-based control of continuum robots in constrained environments without relying on analytical models. We propose a data-efficient stochastic control strategy incorporating online model updates to achieve precise manipulation even when arbitrary robot deformations occur due to environmental interactions. A localized Gaussian process regression approach accounting for state stochasticity is first presented to approximate the forward kinematics. The learned model enables uncertainty-aware stochastic predictions via the proposed scaled unscented transform (SUT)-based method for efficient exploration. Leveraging new data, online model updates are performed in a highly sample-efficient manner. Furthermore, a probabilistic model predictive control approach integrating the learned models and chance constraints based on Chebyshev’s inequality is developed for searching an optimal control sequence. Simulations and experiments are performed to demonstrate the effectiveness of the proposed approach for controlling continuum robots in constrained environments using limited observational data. Note to Practitioners—The motivation of this research is to solve the problem of controlling continuum robots in constraint environment. The flexibility of continuum robots significantly affects the manipulation accuracy, and the interaction between the continuum robot and environmental constraints can also lead to unpredictable behavior. Learning control methods that rely only on sensory data, provide a feasible solution to the aforementioned problem. However, current methods lack sample efficiency and the capability to handle unknown environmental constraints. This research proposes a learning control method which can control a flexible continuum robot in constrained environments with high data-efficiency and robustness even when the robot shape undergoes sudden deformations due to contact with obstacles.}, archive = {J_TASE}, author = {Hangjie Mo and Ruofeng Wei and Xiaowen Kong and Yujia Zhai and Yunhui Liu and Dong Sun}, doi = {10.1109/TASE.2024.3357816}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {984--995}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-efficient learning control of continuum robots in constrained environments}, volume = {22}, year = {2025}, }
• (2025). Robust distributed average tracking with disturbance observer control. TASE, 22, 970--983. (www)

  @article{ , abstract = {This paper is concerned with the study of robust distributed average tracking (DAT) algorithms for networked control systems in the presence of external disturbances or false data injection attacks (FDIAs). To eliminate the impacts of external disturbances and FDIAs, the technologies of disturbance-observer-based control (DOBC) and active disturbance rejection control (ADRC) are introduced into the context of DAT problems. First, for a class of external disturbances with known dynamics, we propose an anti-disturbance DAT (AD-DAT) algorithm, where a stand-alone disturbance observer based on the idea of DOBC is employed to estimate the disturbance and then to compensate it in the design of control inputs. The proposed AD-DAT algorithm can track the average of multiple time-varying reference signals with zero steady-state error and the accurate tracking is robust with respect to initialization constraints. Furthermore, for another class of FDIAs with unknown dynamics, we design an anti-attack DAT (AA-DAT) algorithm where the control input is based on the estimates of states instead of original states, and construct an extended state observer including a state observer and an FDIAs observer based on the idea of ADRC. The extended state observer plays a key role in estimating and eliminating the impact of FDIAs without compromising the accurate tracking performance. In addition, sufficient conditions are derived for the proposed two algorithms from a theoretical point of view to guarantee accurate average tracking. Finally, some numerical examples are given to illustrate the validity and effectiveness of the proposed algorithms. Note to Practitioners—This paper is motivated by the problem of robust distributed average tracking (DAT) for the time-varying centroid of the formation of a group of autonomous vehicles. The problem arises in the scenario where two groups of unmanned ground vehicles (UGVs) and unmanned aerial vehicles (UAVs) perform a combined surveillance-reconnaissance mission (where the UAVs aim to provide aerial coverage and early warning against threats for the UGVs, as shown in Fig. 1) in an uncertain environment where the external disturbances might exist or the FDIAs might be launched by adversaries. Obviously, the tracking accuracy of the target signal will be compromised in the presence of external disturbances or FDIAs. However, most existing works for disturbance rejection mainly focused on the static average consensus rather than the dynamic one even though a few works mentioned the DAT problem with only considering the case of external disturbances. Based on this, we propose an AD-DAT algorithm and an AA-DAT algorithm for the DAT problem based on the ideas of DOBC and ADRC, respectively. Numerical examples show that the proposed algorithms are able to estimate and eliminate the impacts of the external disturbances and the FDIAs, which implies that the algorithms can be implemented in practical scenarios. In future research, we will extend the results to more general scenarios in the presence of external disturbances and FDIAs.}, archive = {J_TASE}, author = {Lan Gao and Hao Lu and Jianliang Wang and Xiang Yu Yu and Peng Jiang and Fei Chen and Huaqing Li}, doi = {10.1109/TASE.2024.3357527}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {970--983}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust distributed average tracking with disturbance observer control}, volume = {22}, year = {2025}, }
• (2025). A physics-constrained TD3 algorithm for simultaneous virtual inertia and damping control of grid-connected variable speed DFIG wind turbines. TASE, 22, 958--969. (www)

  @article{ , abstract = {This paper proposed a physics-constrained twin delayed deep deterministic policy gradient (TD3) algorithm for simultaneous virtual inertia and damping control of a grid-connected variable speed doubly-fed induction generator (DFIG) wind turbine using a combined deep reinforcement learning (DRL) and quadratic programming as a novel solution to suppress frequency fluctuations caused by the control mechanism which decouples the active power from the system frequency, thus hiding the rotating kinetic energy of the wind generator. The optimization stage modifies the action of the DRL agent, thus preventing the agent from taking certain unsafe actions. We tested the effectiveness of the proposed scheme under various scenarios through simulations on an IEEE 9-bus test system in MATLAB/Simulink. Compared with other virtual inertia controls, the results show that the proposed scheme achieved improved dynamic performance with the lowest system frequency deviation and fastest frequency recovery under wind and load variations and severe grid faults. A further test on the IEEE 39-bus system shows that the grid size does not affect the performance of our proposed technique. Note to Practitioners—Integrating the wind turbine systems into the utility grid results in power quality problems such as frequency fluctuation, voltage dip, power loss, and severe power outages. The unpredictability and uncontrollability of the wind pose a serious problem in integrating wind energy conversion systems. This problem becomes worse with the increasing number of connected wind turbines. Therefore, new control strategies are required to mitigate this issue. The droop-based virtual and damping control method traditionally provides frequency support in grid-tied wind turbine systems. However, the fixed droop gain is a significant drawback of this method. In this paper, we proposed a novel physics-constrained deep-reinforcement learning-based virtual inertial and damping control. The proposed control agent is constrained from unsafe actions and rewarded for maintaining the grid frequency within operational limits. Simulation results with the IEEE 9 bus system validated the feasibility and effectiveness of our proposed approach. A comparison of our method with the conventional control scheme, adaptive droop-based virtual inertia control, etc., carried out under various operational scenarios verified the enhanced performance of our proposed strategy.}, archive = {J_TASE}, author = {Osarodion Emmanuel Egbomwan and Hicham Chaoui and Shichao Liu}, doi = {10.1109/TASE.2024.3357204}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {958--969}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A physics-constrained TD3 algorithm for simultaneous virtual inertia and damping control of grid-connected variable speed DFIG wind turbines}, volume = {22}, year = {2025}, }
• (2025). Adaptive T-S fuzzy control for an unknown structure system with a self-adjusting control accuracy. TASE, 22, 944--957. (www)

  @article{ , abstract = {Adaptive Takagi-Sugeno (T-S) fuzzy control was a real-time control method without off-line input and output data, but the approximate error between this T-S fuzzy model and the actual system model was rarely considered in the existing methods. This problem can lead to the degradation of controller accuracy in practical engineering. In order to solve this problem, the mathematical expression of the upper bound for the approximate error between adaptive T-S fuzzy logic system and the actual system was derived for the first time under any number of rules. This achievement was the key to the design of robust fuzzy controller based on Lyapunov synthesis method under finite number of rules, and this controller can achieve predictability of accuracy. Compared with existing T-S fuzzy control methods, the proposed method can realize the self-adjusting accuracy control for the unknown-structure system without the off-line input and output data. The unknown non-affine control system simulation and 3-DOF robotic arm experiment were carried out to verify the effectiveness of proposed method. Note to Practitioners—Due to the wide application of T-S fuzzy model in practice control system, such as the robot control, the unmanned vehicle control and so on, the actual control accuracy of T-S fuzzy controller was concerned by many scholars. However, the actual precision degradation of the controller was a difficult problem in on-line T-S fuzzy control. In order to solve this problem, a novel adaptive T-S fuzzy control method was proposed in this paper. In practical application, T-S fuzzy model often had a large approximate error with the actual model because of structure uncertainty, and this approximate error was the key factor affecting the actual accuracy of T-S fuzzy controller. Hence, the mathematical expression of the upper bound for the approximate error between adaptive T-S fuzzy logic system and the actual system was derived for the first time under any number of rules. Based on this mathematical expression, we can design a self-adjusting accuracy robust fuzzy controller based on Lyapunov synthesis method. 3-DOF robotic arm experiment verified that the proposed method can achieve the predefined control error with or without the feedforward of robot dynamics. 3-DOF robotic arm comparison experiments verified that the advantages of proposed method.}, archive = {J_TASE}, author = {Wen Yan and Tao Zhao and Ben Niu and Xin Wang}, doi = {10.1109/TASE.2024.3356752}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {944--957}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive T-S fuzzy control for an unknown structure system with a self-adjusting control accuracy}, volume = {22}, year = {2025}, }
• (2025). Development of a novel integrated control strategy for automotive electric power steering systems. TASE, 22, 926--943. (www)

  @article{ , abstract = {This paper introduces a new approach for controlling electric steering systems. This new control algorithm combines backstepping (BS) and PID (Proportional Integral Derivative) techniques. The output of the PID algorithm is the input of the BS controller. The parameters of the PID algorithm are dynamically adjusted by a fuzzy algorithm with two inputs called FPIDBS. An automotive dynamics model is established in this work to determine the variation of road reaction torque, which considers the influence of other disturbances. The calculations and simulations are performed in the Simulink-MATLAB interface to evaluate the effectiveness of the control algorithm. According to research findings, the value of the controlled object (steering motor angle) always closely tracks the desired signal with negligible error. Under some conditions, the error between the output and reference signals is approximately zero (after rounding) when applying the FPIDBS algorithm. In addition, the power consumption of the FPIDBS controller is lower than that of the conventional BS controller. In conclusion, the stability and adaptation of the system are always guaranteed in all investigated conditions. Note to Practitioners—This paper is motivated by solving problems relating to other disturbances in automotive electric power steering systems. The use of an integrated algorithm is a suitable solution. Based on this idea, we propose an algorithm that combines PID, BS, and fuzzy techniques. The controlled object will be virtualized through the PID controller before becoming an input to the BS controller. The parameters of the PID controller are adjusted flexibly based on a fuzzy technique. The input of the membership functions is the change of driver torque, velocity, and system error, while the fuzzy rule is determined appropriately, which provides high efficiency. The results show that the output values always track the desired signal. In addition, the effects of other disturbances are eliminated almost entirely.}, archive = {J_TASE}, author = {Tuan Anh Nguyen}, doi = {10.1109/TASE.2024.3356509}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {926--943}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Development of a novel integrated control strategy for automotive electric power steering systems}, volume = {22}, year = {2025}, }
• (2025). Bumpless transfer MPC for hidden markovian jump nonlinear systems with saturated inputs. TASE, 22, 916--925. (www)

  @article{ , abstract = {This paper addresses the problem of the model predictive control (MPC) for a class of T–S fuzzy Markovian jump systems with bounded inputs and states. Two scenarios are considered, which are often overlooked in existing literature but common in practice: (i) asynchronous switching between the observed mode and the actual mode, and (ii) the excessive bump of system states. By virtue of the emission probability, the underlying systems are modeled as hidden Markovian jump systems to describe the asynchrony phenomenon. The time-varying emission probabilities are transformed into the ones in a more tractable form. Moreover, the bumpless transfer MPC (BT-MPC) is introduced to tackle the state bump phenomenon. An asynchronous fuzzy controller is obtained by the receding-horizon optimization, while ensuring recursive feasibility and stochastic stability. Besides, by a saturated state feedback scheme, the control law exploits the full control range with less conservatism compared with the case of directly dealing with control constraints. Two simulation examples are provided to demonstrate the effectiveness and advantages of the proposed control approach. Note to Practitioners—As a special hybrid system, Markovian jump systems can effectively describe the randomness of multi-modal systems. However, both the asynchrony phenomenon and hard constraints in the actual process reduce the range of existing control strategies. In addition, bumpless transfer performance is significant in many practical engineering systems. In this paper, a novel methodology is proposed to design bumpless transfer control for the fuzzy hidden Markovian jump system with time-varying emission probabilities. The proposed approach can handle both multiple control objectives and system constraints under the framework of BT-MPC. The proposed theoretical result is of a convex optimization characterization, which can be implemented algorithmically by means of the advanced and proven convex optimization toolbox. Finally, both a numerical example and a practical example of a robot arm are presented to demonstrate the effectiveness and great application prospect of the proposed approach.}, archive = {J_TASE}, author = {Rui Weng and Bo Li and Yifei Dong and Bo Cai and Tong Wu and Chengzhe Han and Lixian Zhang}, doi = {10.1109/TASE.2024.3356062}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {916--925}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Bumpless transfer MPC for hidden markovian jump nonlinear systems with saturated inputs}, volume = {22}, year = {2025}, }
• (2025). A masked multi-view bidirectional network for class extremely imbalanced object detection under magnetic flux leakage signals. TASE, 22, 905--915. (www)

  @article{ , abstract = {Magnetic flux leakage (MFL) detector can analyze the state of pipelines and is widely used in pipeline inspection. Object detection of class extremely imbalanced MFL signals is a challenging and demanding task. It is difficult for existing deep learning-based detection methods to effectively detect a small number of component classes. To address the above issues, a novel masked multi-view bidirectional network (MMB-Net) is proposed in this paper. First, based on the problem of insufficient component samples, a mask-based data augmentation module is designed, where we develop a novel wavelet convolution blocking module and Gauss-like distribution mask strategy so as to expand component sample sets. Second, based on the problem that poor feature extraction of partial MFL signals leads to low accuracy of component detection, a learnable multi-view attention (LMA) module is designed to expand the perceptual field of the network and fully mine the features of MFL signals. Finally, a bidirectional attention module based on multiple classification layers (BAMC) is proposed to learn different decision boundaries, which maintains the accuracy of the defects while improving the accuracy of the components. Experimental results illustrate that the proposed network can effectively build a detection model under MFL signals and outperforms the state-of-the-art methods on mean precision (9.5%). Note to Practitioners—This article is motivated by the problem of extreme class imbalance in the MFL signals collected from pipelines, which seriously reduces the detection accuracy of defects and components in the MFL signals. In terms of the MFL signal intrinsic characteristics, a masked multi-view bidirectional network (MMB-Net) based on data dynamic adjustment of model parameters and bidirectional learning decision boundaries are constructed. Our method can extend the component samples and learn different decision boundaries to improve the detection accuracy of the component classes. The experimental results show that our proposed method is more advantageous than the state-of-the-art methods, and therefore, our proposed method has strong practical value.}, archive = {J_TASE}, author = {He Zhao and Jinhai Liu and Huaguang Zhang and Qiannan Wang and Xiangkai Shen and Xin Zhang}, doi = {10.1109/TASE.2024.3355984}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {905--915}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A masked multi-view bidirectional network for class extremely imbalanced object detection under magnetic flux leakage signals}, volume = {22}, year = {2025}, }
• (2025). Three-filters-to-normal+: Revisiting discontinuity discrimination in depth-to-normal translation. TASE, 22, 895--904. (www)

  @article{ , abstract = {This article introduces three-filters-to-normal+ (3F2N+), an extension of our previous work three-filters-to-normal (3F2N), with a specific focus on incorporating discontinuity discrimination capability into surface normal estimators (SNEs). 3F2N+ achieves this capability by utilizing a novel discontinuity discrimination module (DDM), which combines depth curvature minimization and correlation coefficient maximization through conditional random fields (CRFs). To evaluate the robustness of SNEs on noisy data, we create a large-scale synthetic surface normal (SSN) dataset containing 20 scenarios (ten indoor scenarios and ten outdoor scenarios with and without random Gaussian noise added to depth images). Extensive experiments demonstrate that 3F2N+ achieves greater performance than all other geometry-based surface normal estimators, with average angular errors of 7.85°, 8.95°, 9.25°, and 11.98° on the clean-indoor, clean-outdoor, noisy-indoor, and noisy-outdoor datasets, respectively. We conduct three additional experiments to demonstrate the effectiveness of incorporating our proposed 3F2N+ into downstream robot perception tasks, including freespace detection, 6D object pose estimation, and point cloud completion. Our source code and datasets are publicly available at https://mias.group/3F2Nplus. Note to Practitioners—The primary motivation behind this work arises from the need to develop a high-performing surface normal estimator for practical robotics and computer vision applications. While geometry-based surface normal estimators have been widely used in these domains, the existing solutions focus merely on discontinuity discrimination. To tackle this problem, this article introduces a plug-and-play module that leverages both depth curvature and correlation coefficient to quantify discontinuity levels, thereby optimizing surface normal estimation, particularly near or on discontinuous regions. Moreover, this article also introduces a large-scale public dataset with random noise added to depth images, providing a more realistic and robust platform for algorithm evaluation within this research community. Extensive experimental results demonstrate that our method outperforms other state-of-the-art algorithms.}, archive = {J_TASE}, author = {Jingwei Yang and Bohuan Xue and Yi Feng and Deming Wang and Rui Fan and Qijun Chen}, doi = {10.1109/TASE.2024.3355941}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {895--904}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Three-filters-to-normal+: Revisiting discontinuity discrimination in depth-to-normal translation}, volume = {22}, year = {2025}, }
• (2025). Adaptive displacement constraint control with predefined performance for active magnetic bearings. TASE, 22, 883--894. (www)

  @article{ , abstract = {This manuscript presents an adaptive funnel controller applying to the active magnetic bearing for position constraint control with prescriptive tracking performance. To better depict the actual active magnetic bearing dynamics, a nonlinear active magnetic bearing model with switched parameters instead of the existing fixed parameters is constructed considering the inherent properties of uncertainties and nonstationarities of active magnetic bearings. Then, by designing the funnel control scheme and adaptive laws based on Lyapunov stability theory and backstepping technique, the rotor displacement is constrained to not exceed the prescribed funnel boundary in view of safety concerns. Moreover, the funnel boundary integrated into the proposed controller is in line with the rotor displacement characteristics of active magnetic bearing systems under different speeds during real operation process. The boundness of the position tracking error is confirmed via Lyapunov synthesis. Eventually, the effectiveness of the proposed controller is verified by simulated and experimental examples. Note to Practitioners—The motivation of this manuscript is to investigate a control strategy with potential application for the displacement control of active magnetic bearings. In most of the existing displacement control schemes for active magnetic bearings, the constraint and limitation of displacement is not taken into account. However, considering the high-speed rotation and active controllability of active magnetic bearings, it is necessary to preset the performance indicators of rotor displacement including convergence speed, overshoot, and constraint limit, which can not only avoid serious safety accidents, but also give full play to the active control advantages of active magnetic bearings in different actual situations. Therefore, this manuscript suggests an adaptive funnel control strategy for prescriptive performance. Moreover, this manuscript solves the modeling and control problems for active magnetic bearings subject to nonlinearity, uncertainty, and external disturbances, which are difficult to be tackled in practical applications. The stability and convergence of the proposed controller are analyzed mathematically, and the practical experimental results reveal that the proposed controller has the potential and possibility to be applied in practice.}, archive = {J_TASE}, author = {Xiaoting Gao and Enchang Cui and Dongsheng Yang and Zilong Tan and Jiayue Sun}, doi = {10.1109/TASE.2024.3355271}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {883--894}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Adaptive displacement constraint control with predefined performance for active magnetic bearings}, volume = {22}, year = {2025}, }
• (2025). Risk-constrained MPC with application in safety-critical control of turbofan engines. TASE, 22, 872--882. (www)

  @article{ , abstract = {Several research have been adopted to design robust or stochastic model predictive control (MPC) strategies to tackle the model uncertainty. However, the worst-case events do not occur frequently, and the statistical information is neglected, which may render robust MPC strategies conservative. In addition, the stochastic MPC controller design, which computes the probability of constraint satisfaction or performance improvement using expected indicators, cannot rule out the incident of less probability to reduce the possibly catastrophic consequences. For instance, severe variations in wind speed can affect the airflow into the engine, impacting the compression and combustion processes, which may cause a surge or choke phenomenon. In this paper, we take risk-aware strategies into account for MPC design to tackle the tails of a probability distribution that can contribute to moderating the risks. Utilizing the proposed risk measurement method, we establish a risk-constrained MPC framework to seek tradeoffs between improving the control performance of interest and restricting the risk of catastrophic incidents, and we reformulate the proposed risk-constrained MPC in a favorable form that is computationally tractable. Moreover, to ensure the safety requirement of the turbofan engine operating processes, we explicitly analyze the closed-loop properties, e.g., recursive feasibility and stability, of the proposed risk-constrained MPC method. Lastly, we demonstrate the efficiency of the proposed risk-constrained MPC algorithm on the safety-critical control of turbofan engines using real-world data to verify the availability of the proposed algorithm. Note to Practitioners—Model predictive control (MPC) is fundamental for several applications in the process industry, transportation, and robotics. However, due to the existence of the model uncertainty, there may occur severe constraint violations or even instability for certain safety-critical systems, which regard normally given safety constraints as a priority in the deployment of control algorithm. For instance, in the safety-critical control of turbofan engine, the latent force of the dynamical system can be affected by some external surrounding factors, such as inlet wind speed, ambient temperature, barometric pressure, inlet conditions, etc., which may cause surge or choke phenomenon. To tackle the model uncertainty, several pieces of research have been adopted to design MPC strategies in the presence of uncertainties. Nonetheless, the worst-case events do not occur frequently, and the statistical information is neglected, which renders robust MPC strategies extremely conservative. In addition, the stochastic MPC controller design, which computes the probability of constraint satisfaction or performance improvement using expected indicators, does not rule out the incident of less probability to reduce the possibly catastrophic consequences. In view of the threat of the safety-critical control of turbofan engines, a risk measurement method and a risk-constrained model predictive control strategy are proposed. Meanwhile, to guarantee the safety and reliability of turbofan engines, we explicitly analyze the proposed algorithm’s closed-loop properties.}, archive = {J_TASE}, author = {Ran Chen and Yuzhe Li and Tianyou Chai}, doi = {10.1109/TASE.2024.3355152}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {872--882}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Risk-constrained MPC with application in safety-critical control of turbofan engines}, volume = {22}, year = {2025}, }
• (2025). A position controller for hydraulic excavators with deadtime and regenerative pipelines. TASE, 22, 855--871. (www)

  @article{ , abstract = {This paper proposes a position controller for commercial hydraulic excavators. It is constructed by combining a proportional-derivative (PD) controller and a sliding-mode controller as a differential algebraic inclusion and also is integrated with a recently-proposed hydraulic actuator model. The use of the PD control is intended to make the controller insensitive to the deadtime in the hydraulic system, which is typically 0.1 s to 0.6 s in commercial excavators. The use of the sliding-mode controller combined with the actuator model is for handling the saturation of the actuator force, which may happen when the target position is not close enough to the current position and when the relief valves open. Moreover, this paper extends the controller to deal with the effect of the regenerative pipelines, which are embedded in commercial excavators to realize efficient operations but act as a source of disturbance on the controller. This paper also shows an analysis that can be used for tuning the controller parameters. The proposed controller was validated with simulations and experiments using a 13-ton class excavator, in which some set-point control tasks and trajectory-tracking tasks were performed. Note to Practitioners—This paper proposes a position controller for hydraulic excavators. The controller was validated with the boom and arm actuators of a 13-ton class commercial excavator, with trajectory-tracking and set-point control tasks. Most of the controller parameters can be set referring to available specifications of the hydraulic circuit, such as the set pressures of the relief valves and the cross-sectional areas of the chambers. There are three parameters (the proportional gain, the derivative gain, and the time constant of the convergence) that should be carefully tuned, but their physical interpretations are relatively straightforward, and they can also be tuned along our guideline using pole locations of a particular transfer function. It has been shown that the proposed controller properly works despite the existence of the deadtime in the hydraulic systems, which are typically 0.1 s to 0.6 s. An extended version of the proposed controller handles the effects of the regenerative pipeline, which exists in some hydraulic systems, e.g., the arm actuator, for efficient operation and is not accounted for in the original version of the controller. The accuracy of the proposed controller will be further improved by combining it with better means to estimate external forces by using additional sensors, such as pressure sensors installed to actuator chambers.}, archive = {J_TASE}, author = {Yuki Yamamoto and Jinjun Qiu and Takayuki Doi and Takao Nanjo and Koji Yamashita and Ryo Kikuuwe}, doi = {10.1109/TASE.2024.3354952}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {855--871}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A position controller for hydraulic excavators with deadtime and regenerative pipelines}, volume = {22}, year = {2025}, }
• (2025). Electromagnetic-mechanical modeling and evaluation of a 2-DoF parallel-kinematic compliant nano-positioning stage based on normal-stressed electromagnetic actuators. TASE, 22, 841--854. (www)

  @article{ , abstract = {The normal-stressed electromagnetic actuators (NSEAs) have emerged as a promising actuation technology for developing high-performance compliant nano-positioning stages. But less attention was devoted to the modeling of the multi degrees-of-freedom (DoF) NSEA-based stages. This paper aims to improve the modeling accuracy of the stages’ static and dynamic performances by introducing a novel electromagnetic-mechanical modeling method. Unlike the previous studies, the proposed modeling method considers the effects of the NSEAs’ negative stiffnesses along both its actuation direction and the vertical-to-actuation direction in the electromagnetic modeling step. Together with the model of the mechanisms, the coupled electromagnetic-mechanical model is analytically derived. As an application case, the working stroke and resonant frequency of a NSEA-based parallel-kinematic 2-DoF compliant nano-positioning stage are evaluated with the proposed model. It is demonstrated by numerical, simulation, and experimental studies, that the proposed modeling method is accurate for predicting the performances of the NSEA-based nano-positioning stages. This is significant for the future development and applications of NSEA-based mechatronic systems. Note to Practitioners—The multi-DoF NSEA-based nano-positioning stages are the promising choice to develop long stroke, high natural frequency nano-positioning stages. However, some inherent electromagnetic-mechanical coupling effects are ignored in previous study when modeling the multi-DoF NSEA-based stages, which leads to a significant prediction error. By considering the electromagnetic characteristics both along and vertical to actuation direction, this study proposed an electromagnetic-mechanical model for the 2-DoF NSEA-based stages. The proposed model shows a higher accuracy to evaluate the static and dynamic performances of the NSEA-based stages, which are beneficial to practical applications such as trajectories tracking.}, archive = {J_TASE}, author = {Bocheng Yu and Xiangyuan Wang and Lingwen Tan and Qi Yu and Yixuan Meng and Linlin Li and LiMin Zhu}, doi = {10.1109/TASE.2024.3354943}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {841--854}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Electromagnetic-mechanical modeling and evaluation of a 2-DoF parallel-kinematic compliant nano-positioning stage based on normal-stressed electromagnetic actuators}, volume = {22}, year = {2025}, }
• (2025). Dynamic nonlinear indoor environment thermal state estimation with unknown inputs using PSO guided regularizer based adaptive EKF. TASE, 22, 830--840. (www)

  @article{ , abstract = {The present paper shows how the dynamic indoor temperature profile of an HVAC (Heating, Ventilation, and Air Conditioning) system in a building can be developed using Kalman filters, in presence of unknown inputs. An RC network based dynamic, nonlinear thermal model is first developed for the indoor environment with a novel consideration of relative humidity factor. Then an extended Kalman Filter based algorithm in presence of unknown inputs (called EKF-UI) and an adaptive variation of this EKF-UI algorithm (called AdEKF-UI) are developed for the real indoor environment under consideration. Next, a particle swarm optimization (PSO) guided adaptive extended Kalman filter with unknown inputs (PSOgAdEKF-UI) algorithm is proposed to overcome limitations of the EKF-UI and AdEKF-UI algorithms, especially under bad initialization situations. This PSOgAdEKF-UI algorithm proposes an effective utilization of regularizer based initializations for the initial state estimation error covariance matrix and the measurement noise covariance matrix. Extensive experiments showed that, overall, PSOgAdEKF-UI algorithm could outperform EKF-UI and AdEKF-UI algorithms by 46.59% and 20.66%, respectively, in terms of mean square error, while estimating an unknown state. Note to Practitioners—This paper was motivated to estimate the nonlinear dynamics of indoor HVAC thermal profile in presence of unknown inputs. The study explores a proposed Kalman filter-based heuristic regularizer-assisted adaptive filtering methodology for nonlinear state estimation that can circumvent the constraints imposed by current approaches. The proposed method demonstrates its applicability in actual nonlinear physical systems since many matrices needed for such state estimation algorithms do not have accurate initialization information. The nature of inferential stochastic inputs in practical HVAC system can be evaluated utilizing our novel state estimation method of the altering relative humidity coupled nonlinear dynamic thermal model. The thermal profile of a practical HVAC system, in presence of varying unknown inputs, can be more accurately modeled when temporal variations in relative humidity are included in the nonlinear dynamic model, as an additional influencing factor.}, archive = {J_TASE}, author = {Bed Prakash Das and Kaushik Das Sharma and Amitava Chatterjee and Jitendranath Bera}, doi = {10.1109/TASE.2024.3354930}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {830--840}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic nonlinear indoor environment thermal state estimation with unknown inputs using PSO guided regularizer based adaptive EKF}, volume = {22}, year = {2025}, }
• (2025). Business process modeling for industrial internet application via BPMN extension. TASE, 22, 813--829. (www)

  @article{ , abstract = {Business process modeling is widely used in modern organizations for business description. Business Process Modeling Notation (BPMN), as a de-facto modeling standard, represents business process models in graphical notations. Nevertheless, BPMN lacks intuitive modeling tasks for Industrial Internet application scenarios (e.g., IoT tasks and multi-instance tasks with constraints). Although there are some works on extending BPMN elements to improve the model representation, most of them stay in the conceptual model only without tool support, or they are confined to specific domains. In this paper, we extend both BPMN elements and attributes for application in the Industrial Internet context, and two modeling tools are implemented in a client version and Web version to support business process modeling via low-code, enabling the BPMN extension model from conceptual to executable level. Two real-world case studies in Industrial Internet are conducted to show the usefulness of process models with BPMN extension. Furthermore, a comprehensive user experiment is conducted to evaluate the extended process models and tools, and the experimental results show that process models with extension have better quality compared with traditional process models and provided tools are effective for business process modeling in Industrial Internet applications. Note to Practitioners—This article was motivated by the problem of business process modeling for Industrial Internet. Practically, existing methods extend BPMN elements and add text annotations to enhance the representation of process models. Although these methods facilitate process participants to understand the process models in detail, most of them focus on conceptual modeling and increase the complexity of process models. Moreover, the process models and multi-instance business constraints of business processes involving IoT elements are difficult to be represented by using native BPMN. This article proposes a comprehensive business process modeling language, named BPMN++, which extend both new modeling elements for Industrial Internet business requirements and attributes for BPMN multi-instance tasks with the corresponding graphical notations. Then, BPMN++ model could be executed on the process engine, enabling the model from conceptual to executable level. Finally, comprehensive experiments are conducted for BPMN++.}, archive = {J_TASE}, author = {Guosheng Kang and Hangyu Cheng and Jianxun Liu and Yiping Wen and Jun Peng}, doi = {10.1109/TASE.2024.3354929}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {813--829}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Business process modeling for industrial internet application via BPMN extension}, volume = {22}, year = {2025}, }
• (2025). Finite-horizon optimal control for nonlinear multi-input systems with online adaptive integral reinforcement learning. TASE, 22, 802--812. (www)

  @article{ , abstract = {In this paper, a novel adaptive integral reinforcement learning (AIRL) is utilized to online handle the finite-horizon optimum control policies of the partially unknown multi-input nonlinear system. Firstly, the concept of Nash equilibrium is introduced to make the multiple cost functions reach the saddle point. Then, dual neural networks (NNs) are applied to approach the performance index functions based on the integral reinforcement signal. Simultaneously, two novel learning algorithms are proposed to update the NN weights, in which the convergence of weights is proved. Then, the optimal strategies can be obtained by using the obtained weights. The designed controllers based on the data-driven AIRL scheme can avoid the internal state of the system and the derivatives of NN activations in the weight learning process. Finally, the stability of the controlled system is analyzed. An F-16 aircraft model and another nonlinear system are utilized to prove the validity and rationality of the algorithm. Note to Practitioners—There exist many multi-input systems in practical engineering, which includes multi-engine driven F-16 aircraft, large radar servo system and large artillery systems, etc. The finite-horizon optimal control of these systems is crucial for the better performance of system state. However, an accurate engineering model is difficult to obtain, and the finite-horizon cannot generally be achieved. To address these issues, this paper proposes the finite-horizon optimal control for these systems based on a novel adaptive integral reinforcement learning (AIRL). The AIRL can realize an optimal performance for these multi-input systems in finite-time without internal system dynamics, which is a good development for the multi-input system in practical engineering.}, archive = {J_TASE}, author = {Yongfeng Lv and Wan Zhang and Jun Zhao and Xiaowei Zhao}, doi = {10.1109/TASE.2024.3354830}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {802--812}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Finite-horizon optimal control for nonlinear multi-input systems with online adaptive integral reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Integrated planning of multiple spare parts inventory, warranty, and service engineers for a service-oriented manufacturer. TASE, 22, 788--801. (www)

  @article{ , abstract = {This paper considers a service-oriented manufacturer providing products and associated after-sales maintenance service with a warranty period to customers. The manufacturer gains revenue from both product sales and after-sales maintenance service which are affected by the warranty period length. The arrival tendency of customers for after-sales maintenance service is higher during the warranty period than that after the warranty expires. The manufacturer repairs stochastic failures of products with both spare parts of the right kind and employed service engineers. If the spare part is not available upon repair demand, the manufacturer can resort to the emergency suppliers (e.g., OEMs). We establish an original decision model for the service-oriented manufacturer and analytically derive the arrival rates of service demand under and out of warranty, and the repair call rates for requiring regular and emergency replenishment based on queuing theory. Then, a computationally efficient algorithm is designed to obtain the joint multiple spare parts inventory, service engineer employment and warranty period decisions. Numerical studies show that both the manufacturer and customers can benefit from the longer warranty period when customers are more inclined to require maintenance service out of warranty from the manufacturer. A higher service level target or quality levels of spare parts reduce the after-sales service dependence on emergency suppliers. Note to Practitioners—After-sales maintenance services often generate significant profit for many service-oriented manufacturers that sell products bundled with after-sales services. To offer efficient after-sales maintenance services, efficient joint management of spare part inventory, service engineers and warranty periods, which often interact with each other, is crucial for service-oriented manufacturers. However, in practice, facing the stochastic arrival of repair calls, it is challenging for manufacturers to jointly optimize the decisions of spare parts inventory, service engineers and warranty periods. Most existing studies optimized them separately. This research investigates the joint decision problem considering the three aspects based on queuing model. Based on the model analysis, an efficient and easy-to-implement heuristic algorithm is proposed to obtain reasonable joint decisions. Practitioners can implement the proposed approach to improve the integrated planning of spare parts inventory, warranty and service engineers and enhance profitability. Besides, sensitivity analysis is conducted to show how the changes of parameters such as payment to the supplier per emergency shipment and rate of the component lifespan will affect the joint decisions. Some managerial insights obtained can be useful to practitioners. First, if customers become more willing to require maintenance service from the manufacturer out of warranty, the manufacturer can benefit more from extending the warranty period length which is also beneficial to customers, leading to a win-win situation. Second, if the emergency replenishment cost becomes higher, the manufacturer should extend the warranty period length. Third, a higher service level target or quality levels of spare parts decline the dependence of the after-sales service on emergency suppliers.}, archive = {J_TASE}, author = {Qingwei Wang and Chang Liu and Meimei Zheng and Dong Wang and Ershun Pan}, doi = {10.1109/TASE.2024.3354422}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {788--801}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integrated planning of multiple spare parts inventory, warranty, and service engineers for a service-oriented manufacturer}, volume = {22}, year = {2025}, }
• (2025). Intelligent consensus asymptotic tracking control for nonlinear multiagent systems under denial-of-service attacks. TASE, 22, 776--787. (www)

  @article{ , abstract = {This work considers the adaptive output feedback consensus asymptotic tracking control problem for full-state-constrained nonlinear multiagent systems (MASs) under the denial-of-service (DoS) attacks and the actuator faults. Since MASs are regardered as a network, where each agent is considered as a network node, the DoS attacks that cut network communication links between agents are studied in this paper. Moreover, to compensate the unknown actuator fault signal, the Nussbaum function method is introduced into the controller design process for each agent. Due to the researched nonlinear MASs contain unknown nonlinearities and the system states of each agent are unmeasurable, an estimator based on fuzzy logic systems (FLSs) is constructed for each agent. Further, the secure control strategy based on the barrier Lyapunov functions (BLFs) is proposed for the researched nonlinear MASs, which guarantees the consensus tracking errors asymptotically converge to zero, all closed-loop signals remain bounded and the full-state constraint requirements are not broken. In the illustrative example, the proposed secure control strategy is applied to the MASs composed of multiple forced damped pendulums (FDPs). Note to Practitioners— In the framework of network communication, how to eliminate the impact of cyber attacks on nonlinear MASs that rely on communication links to transmit information is critical. Furthermore, due to practical systems are inevitably subject to multiple physical and environmental constraints, the problem of dealing with system constraints can not be ignored. Therefore, a secure control algorithm based on the BLFs is proposed to eliminate the effects of the DoS attacks on the researched nonlinear MASs and ensure that the full-state constraints are not broken. To realize the output feedback control and approximate the unknown nonlinearities of the studied system, the estimators combining with the intelligent approximation techniques based on FLSs are developed. The Nussbaum function is introduced into controller to solve the actuator faults problems for each agent.}, archive = {J_TASE}, author = {Yuqiang Jiang and Ben Niu and Tao Zhao and Xudong Zhao and Xinjun Wang and Huanqing Wang}, doi = {10.1109/TASE.2024.3354047}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {776--787}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Intelligent consensus asymptotic tracking control for nonlinear multiagent systems under denial-of-service attacks}, volume = {22}, year = {2025}, }
• (2025). A unified framework with incremental learning capacity for industrial fault detection and classification. TASE, 22, 765--775. (www)

  @article{ , abstract = {Detection and classification are two significant tasks for industrial fault diagnosis. However, conventional methods typically treat these tasks as separate and independent problems, and necessitate a retraining process when new fault samples or classes are collected. Therefore, an incremental support vector data description scheme using Gaussian kernel function is pro-posed for industrial process fault diagnosis in a unified frame-work. In this framework, the decision boundary is updated incrementally only based on the specific original support vectors and newly collected samples. An adaptive threshold and a restructured radius are proposed to promote accuracy in the fault detection. In the classification procedure, the hyperspheres for all known classes are constructed by decision tree. The new sample that does not belong to any known class is identified as an unknown class. Without a time-consuming retraining process, the proposed diagnosis method with the incremental learning capability can synchronously achieve the fault detection and classification task. Experimental results demonstrate the effectiveness and superiority in terms of diagnosis performance. Note to Practitioners—In practical industrial processes, new fault samples are collected and new fault classes emerge continually. Under such scenarios, fault diagnosis with incremental learning capability is becoming increasingly important. This work proposes a unified incremental framework for fault diagnosis based on support vector data description, which is able to achieve fault detection and classification synchronously. For fault detection, an adaptive threshold and a restructured radius are developed. For fault classification, hypersphere-shaped boundaries of all fault classes are given via the decision tree-based strategy. The proposed method is updated to include new fault samples and fault classes without dimensionality reduction or any distributional assumptions.}, archive = {J_TASE}, author = {Li Cai and Hongpeng Yin and Jingdong Lin}, doi = {10.1109/TASE.2024.3353535}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {765--775}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A unified framework with incremental learning capacity for industrial fault detection and classification}, volume = {22}, year = {2025}, }
• (2025). Self-adjustable performance-based adaptive tracking control of uncertain nonlinear systems. TASE, 22, 750--764. (www)

  @article{ , abstract = {This paper is concerned with the problem of prescribed performance tracking control for strict-feedback systems with parametric uncertainties and unmatched disturbances. An adaptive command-filtered control approach with performance guarantees is put forward to address the problem. By means of a new-type self-adjustable performance function and a barrier function, it is capable of achieving the practical prescribed time tracking and enhancing the reliability of control implementation simultaneously, without yet the requirement for the specific initial condition. In the control design, a class of finite-time command filters is adopted to solve the “explosion of complexity” problem, and the error compensation mechanism with practical finite-time stability is introduced to remove the impact of filtered errors. Moreover, considering that the overshoot of the tracking error fails to be quantified by the majority of performance functions with infinite initial values, a pair of asymmetric performance functions is constructed such that the trajectory tracking with the predefined overshoot, settling time, and accuracy is achieved, while preserving the capability of relaxing the initial condition. It turns out that the proposed approach warrants the performance-related constraint satisfaction and the boundedness of all the closed-loop signals. Three simulation studies are carried out to demonstrate the theoretical findings. Note to Practitioners—This paper is motivated by the potential fragility problem exhibited in the traditional constraint-handling methods in the presence of paroxysmal factors such as suddenly strong disturbances or highly fluctuating target trajectories. For practical applications (e.g., satellite docking, automobile production, target interception, and network congestion control) suffering from the above scenario, it might be favorable to widen prescribed performance boundaries in a proactive way from the perspective of the safe and reliable operation of the controlled system. On this basis, a novel command filter-based adaptive tracking control solution with self-adjustable performance guarantees is proposed. It not only achieves trajectory tracking with the preassigned overshoot, settling time, and accuracy in the context of raising the reliability of control implementation but also possesses the attributes of inexpensive computation burden and easy acceptability in practical applications.}, archive = {J_TASE}, author = {Haixiu Xie and Jin-Xi Zhang and Yuanwei Jing and Georgi Marko Dimirovski and Jiqing Chen}, doi = {10.1109/TASE.2024.3353380}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {750--764}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Self-adjustable performance-based adaptive tracking control of uncertain nonlinear systems}, volume = {22}, year = {2025}, }
• (2025). A data-driven Human–Machine collaborative product design system toward intelligent manufacturing. TASE, 22, 736--749. (www)

  @article{ , abstract = {In the era of big data, enterprises have accumulated large amounts of valuable data throughout the entire product life cycle (PLC). Such PLC data contains a wealth of design knowledge. Intelligent manufacturing seeks to establish a collaborative platform that integrates advanced data analytics and artificial intelligence into the manufacturing process, providing new opportunities for efficient and intelligent product design. Mining design knowledge from PLC data and applying it to the design stage is a critical issue that urgently needs to be addressed for data-driven product design (DDPD). To enhance the efficiency and adaptability of DDPD, this work proposes a comprehensive framework for extracting design knowledge from PLC data and utilizing the knowledge to inform the design process. A structured storage method is developed to manage PLC data with multi-source and heterogeneous characteristics. Then, human-machine collaborative pattern extraction, deep learning-based relation extraction, and other data mining techniques are used to extract knowledge from PLC data. Moreover, a product design knowledge network is constructed based on knowledge graph to achieve knowledge organization and management. Finally, a novel intelligent push method for product design knowledge, based on context navigation, is proposed as part of the framework. A case study showcases how data-driven human-machine collaborative patterns can be used to improve the flexibility and performance of product design. Note to Practitioners—Data-driven method can realize the closed-loop design of products while linking users, products and production processes to improve design efficiency. However, one of the major challenges in DDPD is the need to flexibly extract knowledge from PLC data and push them to designers. In this work, we propose a novel system that leverages human-machine collaboration and deep learning methods to realize DDPD toward intelligent manufacturing. It allows us to extract knowledge from product data, and then proactively push appropriate knowledge to designers for decision-making. The proposed system consists of three main components: product life cycle multi-source heterogeneous data processing, product design knowledge mining, and design knowledge intelligent pushing. Specifically, the human-machine collaboration mechanism improves the system’s capability to address uncertain and complex problems. A case study using shield machine PLC data has demonstrated the feasibility and effectiveness of the proposed framework.}, archive = {J_TASE}, author = {Wei Wei and Chuan Jiang and Yuzhe Huang}, doi = {10.1109/TASE.2023.3295571}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {736--749}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A data-driven Human–Machine collaborative product design system toward intelligent manufacturing}, volume = {22}, year = {2025}, }
• (2025). A hybrid QFD-based human-centric decision making approach of disassembly schemes under interval 2-tuple q-rung orthopair fuzzy sets. TASE, 22, 724--735. (www)

  @article{ , abstract = {Design for disassembly (DFD) is a basic design technology serving the scrap and recycling stages, which can relieve the environmental pressure and improve the economic benefits. Considering the wide variety of mechanical products and the shortening of product lifecycle, the product design integrating customer requirements can effectively improve the market competitiveness of products. Thus, this paper proposes a QFD-based human-centric decision-making approach for disassembly scheme selection. The theory of the interval 2-tuple q-rung orthopair fuzzy sets (I2q-ROFSs) is proposed to better describe the ambiguity of the environment and avoid information loss/distortion in the information aggregation stage. A scheme evaluation system based on the disassembly technical features is established. A hybrid multi-attribute decision-making (MADM) method combing BWM-FQFD (best worst method and fuzzy quality function deployment) and RT-EDAS (the regret theory based on distance from average solution) is presented to obtain the optimal alternative. A case study, i.e., four refrigerator DFD schemes, is applied to verify the effectiveness of the proposed method. The result demonstrates that this work provides an effective tool to select the optimal scheme of disassembly considering customer requirements and some references for designers. Note to Practitioners— The selection of DFD schemes is the last task of the disassembly design. The decision information of the disassembly design scheme has insufficient evaluation complexity ability and the information distortion of the evaluation information in the aggregation stage. To end this, this paper proposes I2q-ROFSs which is a fuzzy set with the higher level of fuzzy description and avoiding information distortion. At present, the disassembly design research lacks the correlation between customer demand and disassembly technology features. Therefore, a fuzzy QFD method can well assist manufacturing enterprises to improve the market competitiveness of their products. Then, RT-EDAS can describe consumer psychology, which can choose a more market-appropriate DFD scheme.}, archive = {J_TASE}, author = {Honghao Zhang and Zhongwei Huang and Guangdong Tian and Wenjie Wang and Zhiwu Li}, doi = {10.1109/TASE.2023.3301237}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {724--735}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hybrid QFD-based human-centric decision making approach of disassembly schemes under interval 2-tuple q-rung orthopair fuzzy sets}, volume = {22}, year = {2025}, }
• (2025). Multiple product hybrid disassembly line balancing problem with human-robot collaboration. TASE, 22, 712--723. (www)

  @article{ , abstract = {The advances of manufacturing technology accelerates the replacement of consumer products. The recycling of these out-of-date products not only has economic benefits but also contributes to environmental protection. Therefore, the disassembly and reuse of products have attracted great attention all over the world. The traditional human worker disassembly is characterized by high cost and low efficiency. Robots can work more efficiently, but they are not flexible enough to perform different tasks. On the other hand, the combination of a U-shaped disassembly line and a single-row linear disassembly line would offer unique advantages for various applications. This work studies a hybrid disassembly line balancing problem (HDLBP) based on human-robot collaboration. The special challenge with HDLBP is that we need to consider the work load balancing among different lines, in addition to workstations, to achieve optimal results. A combination of linear programming and integer one is proposed to solve the optimization model of HDLBP that is composed of linear and U-shaped disassembly lines, with the objective of maximal disassembly profit. The feasibility of the model is verified by commercial solver CPLEX in solving different size problem instances. Note to Practitioners—This work deals with issue of using human workers only or using robots alone in disassembly lines and the limitation of each type of disassembly line layout. Most of the existing disassembly operation assignment methods are based on the correlation between humans and robots and the factors that affect disassembly. This paper suggests that the selection of humans and robots based on an optimization model that can be solved CPLEX. To leverage the unique advantages offered by each type of disassembly layout, this paper suggests the use of hybrid disassembly lines. Based on the idea of mixed integer programming, a hybrid disassembly line model of human-robot collaboration is designed and solved by CPLEX. The experimental results show that the hybrid disassembly line of human-robot collaboration has obvious advantages over the disassembly line composed of worker-only or robot-only when disassembling products. In the future research, we will use reinforcement learning algorithm to solve the hybrid disassembly line balancing problem, and consider more details of the human-robot cooperative hybrid disassembly lines.}, archive = {J_TASE}, author = {Shujin Qin and Changsheng Xiang and Jiacun Wang and Shixin Liu and Xiwang Guo and Liang Qi}, doi = {10.1109/TASE.2024.3418898}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {712--723}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multiple product hybrid disassembly line balancing problem with human-robot collaboration}, volume = {22}, year = {2025}, }
• (2025). Edge-assisted epipolar transformer for industrial scene reconstruction. TASE, 22, 701--711. (www)

  @article{ , abstract = {Given a set of calibrated images, Multiple View Stereo (MVS) applies end-to-end depth inference network to recover scene structure. However, previous methods designed pixel-visibility modules to aggregate cross-view cost, ignoring the consistency assumption of 2D contextual features in the 3D depth direction. The current multi-stage depth inference model also relies on intensive depth samples, which requires high memory consumption. To alleviate these problems, this work exploits edge-assisted epipolar Transformer for multi-view depth inference. The improvements of this work are summarized as follows: 1) The epipolar Transformer block is developed for reliable cross-view cost aggregation, and the edge detection branch is designed to constrain the consistency of epipolar geometry and edge features. 2) The dynamic depth range sampling mechanism based on probability volume is adopted to improve the accuracy of uncertain areas. Comprehensive comparisons with the state-of-the-art works indicate that our work can reconstruct dense scene representations with limited memory bottleblock. Note to Practitioners—Learning-based MVS can obtain dense point clouds with accurate depth map estimation, which are widely applied in the fields of unmanned driving, battlefield environment perception and robot navigation. MVS-based scene reconstruction technology is the premise of the subsequent planning, decision-making and control of the human-machine system. To obtain dense scene representation with limited memory and runtime, this work proposes a multi-view stereo network with edge-assisted epipolar Transformer. Experiments on public benchmarks verify the feasibility and effectiveness of our model, which has good potential in battlefield environment reconstruction and human-computer interaction fields, and can provide intuitive and dense scene representation for decision-making assistance.}, archive = {J_TASE}, author = {Wei Tong and Xiaorong Guan and Miaomiao Zhang and Ping Li and Jin Ma and Edmond Q. Wu and Li-Min Zhu}, doi = {10.1109/TASE.2023.3330704}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {701--711}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Edge-assisted epipolar transformer for industrial scene reconstruction}, volume = {22}, year = {2025}, }
• (2025). Risk-aware decision making for Human–Robot collaboration with effective communication. TASE, 22, 691--700. (www)

  @article{ , abstract = {Human-robot collaboration is crucial for integrating robots into intelligent manufacturing (IM). However, a significant challenge is rational decision-making for the human-cyber-physical system (HCPS) in IM to enhance cognitive limits of human operators and overcome the potential irrationality. Since humans dominate the collaboration in IM, it is essential to address two critical issues: determining the next task for the robot and deciding whether the human operator should be informed. We propose a risk-aware decision-making framework for task allocation and human-robot interaction (HRI) to achieve a balance between the autonomy level of the human operator and task efficiency. To quantify the efficiency risk, we utilize conditional value-at-risk (CVaR) considering the uncertainty of human operators. We then obtain the optimal task allocation and selection for the robot by minimizing the efficiency risk. We also establish a necessary collection of tasks that must be performed by the human operator. Furthermore, we develop two criteria to quantify the necessity of explicit HRI. Experiments with a real mechanical arm platform demonstrate that our methods can enhance human-robot collaboration (HRC), reduce the need for extensive communication, and grant human operators greater execution freedom. Note to Practitioners—This research is motivated by the fact that the imperfect information sharing between human operators and robots in IM. It results in the inability of human operators to accurately comprehend the robot’s capability constraints and the robot’s limitations in perceiving the current or next action of the human operator. As a result, the effectiveness of HRC then inevitably decreases. Our proposed method introduces the efficiency risk to quantify the need of robot active intervention in HCPS. This approach aids in determining when and what to communicate with the human operator, thereby empirically ensuring efficiency and providing the human operator with a greater degree of freedom. Experimental studies suggest that the efficiency risk can serve as a new metric for balancing the amount of human-robot communication and the improvement of efficiency. Our research can also find utility in human-centric scenarios where communication between the human operator and the robot is constrained or costly.}, archive = {J_TASE}, author = {Yang Li and Yi Wang and Hongbo Li and Huaping Liu}, doi = {10.1109/TASE.2023.3306868}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {691--700}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Risk-aware decision making for Human–Robot collaboration with effective communication}, volume = {22}, year = {2025}, }
• (2025). Process monitoring and fault prediction of papermaking by learning from imperfect data. TASE, 22, 680--690. (www)

  @article{ , abstract = {Fault prediction is increasingly concerned in the industry due to complexity grows in the production process. Paper break, the most common process fault of papermaking, risks paper mills enormously on cost and efficiency. Data derived from papermaking workshop always involves imperfect issues of mismatching, losing, human intervention etc. which mask the inherent hints about paper break, prevent early warn. This study proposed pretreatment processes on data upon papermaking knowledge and analysis of paper breaks, exploited random forest to extract interrelated features, and establish a prediction model of paper break based on Gaussian mixture models (GMM) and Mahalanobis distance (MD). GMM clusters the datasets of extracted variable normally performed to form the health benchmark, and utilize MD to analyze the deviation of real time state of papermaking process from health, and determine whether to warn the operators of paper break through kernel density estimation. The verification results showed that the proposed model has a fault prediction accuracy of 76.8% and a recall rate of 72.5%, which allows paper break associated anomalous data to be observed in advance, providing valuable time for subsequent fault diagnosis. Note to Practitioners —This article is motivated by the problems of paper break in the papermaking process, which can be applied to fault prediction in papermaking and other associated complex process industries. Due to technical limitations, it is difficult to monitor all parameters of the entire production process to prevent faults from the manufacturing processes. This paper exploits existing imperfect production data, and analyzes breaking mechanisms in the papermaking process to interpret the meaning of certain data. It is obtained variables closely relating to production through data analysis and a framework consisting of health benchmark with deviation determination. Studies on corresponding actual production cases validated that the approach is feasible. The paper breaks can be predicted, but there are still some faults of which mechanisms are too unclear to support prediction. In the future research, it is encouraged to improve the accuracy and timeliness of prediction.}, archive = {J_TASE}, author = {Zhenglei He and Guojian Chen and Mengna Hong and Qingang Xiong and Xianyi Zeng and Yi Man}, doi = {10.1109/TASE.2023.3290552}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {680--690}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Process monitoring and fault prediction of papermaking by learning from imperfect data}, volume = {22}, year = {2025}, }
• (2025). Scheduling a single-arm two-cluster tool with a process module failure subject to wafer residency time constraints. TASE, 22, 667--679. (www)

  @article{ , abstract = {Multi-cluster tools are widely adopted in semiconductor manufacturing. When a process module (PM) at a step fails, a multi-cluster tool cannot complete the process recipe of work-in-process wafers and must be forced to enter a closedown process to be empty. To increase the throughput of a wafer fab, it is economically significant to shorten the failure-closedown process. However, due to the wafer residency time constraints, it is highly challenging to respond to a PM failure and find a corresponding optimal schedule. By assuming no parallel PM at a step, this work is the first to study this important issue for scheduling multi-cluster tools. We analyze the task sequences and synchronization conditions for robot activities to avoid deadlock in a shared buffer module. Upon these analyses, for process-dominant multi-cluster tools whose optimal steady-state schedule is known, algorithms are proposed to synthesize the proper sequences for robots in case of PM failures, then, a nonlinear program model is proposed to find an optimal schedule for the corresponding closedown process or decide no feasible solutions. The proposed model can uniformly deal with different scenarios of PM failures. Examples are given to illustrate the application of the proposed method. Note to Practitioners—In a wafer fab, there are hundreds, even thousands, of cluster tools. It is common that a failure of a processing module happens in a cluster tool. How to intelligently respond to such a random failure in a multi-cluster tool is an important issue in real-world production. This paper studies the scheduling problem of a multi-cluster tool in case of a failure at a PM. For a multi-cluster tool in case of a failure module, the proposed method can significantly reduce the loss of work-in-process wafers and shorten the closedown process.}, archive = {J_TASE}, author = {Qinghua Zhu and Jun Yuan and Genghong Wang and Yan Hou}, doi = {10.1109/TASE.2023.3275703}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {667--679}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling a single-arm two-cluster tool with a process module failure subject to wafer residency time constraints}, volume = {22}, year = {2025}, }
• (2025). Skin-inspired triple tactile sensors integrated on robotic fingers for bimanual manipulation in human-cyber-physical systems. TASE, 22, 656--666. (www)

  @article{ , abstract = {Collaborative robots are predicted to interact physically with humans in human-cyber-physical systems (HCPSs). Robotic hands are able to record force and temperature simultaneously through soft and conformable sensors applied over finger surfaces and conforming to the complex curved geometries of automatic machines with tactile perception. Here, skin-inspired triple tactile (SITT) sensors are integrated into robotic fingers to enable precise bimanual grasping. The SITT sensor has skin-inspired multilayer microstructures, which integrate three sensors, namely, an interdigital electrode sensor, a flexible force sensor, and a temperature sensor. The SITT sensor can simultaneously or independently measure a material’s dielectric property, tactile force and temperature. An HCPS based on SITT sensors, a data acquisition board, bimanual robotic hands, and human-robot interaction software is developed for efficient bimanual manipulation. Through 3C assembly experiments, the designed HCPS is demonstrated to execute complex tasks. This research presents a novel methodology for constructing robust tactile sensors for robotic fingers in a bimanual manipulation system, and it presents significant potential across various aspects of intelligent production, including sensitive object handling, adaptive manipulation, and interactive robotics applications. Note to Practitioners—This work aims to overcome the challenge of skin-inspired triple tactile (SITT) sensors developed for robotic fingers in a human-robot collaborative assembly scenario, which can also be used in many other similar human-robot/machine collaborations (e.g., wearable prosthetics with tactile feedback) with practical value. Its capability to accurately measure the touched objects’ information (e.g., material dielectric property, force, temperature) is crucial for the bimanual robot to successfully interact with human operators. HCPS is valuable for its potential to achieve complicated interactions among humans, cyber systems, and physical resources. Collaborative robots with tactile sensors are expected to interact physically with humans in the HCPS. In this paper, we report SITT sensors integrated on robotic fingers to enable precise bimanual grasping. They can simultaneously or independently measure material dielectric properties and tactile forces and record temperature. By combining tactile perception information and a related control method, our smart bimanual robotic hands with the developed HCPS demonstrates the capability to execute 3C assembly in intelligent manufacturing. In the future, additional application scenarios will be designed for the HCPS platform, and some traditional tasks (such as a single arm for collaborative assembly) will be replaced by our hardware and software systems.}, archive = {J_TASE}, author = {Shumi Zhao and Zhijun Li and Haisheng Xia and Rongxin Cui}, doi = {10.1109/TASE.2023.3320710}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {656--666}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Skin-inspired triple tactile sensors integrated on robotic fingers for bimanual manipulation in human-cyber-physical systems}, volume = {22}, year = {2025}, }
• (2025). Local observation based reactive temporal logic planning of human-robot systems. TASE, 22, 643--655. (www)

  @article{ , abstract = {Human-robot collaboration plays an important role in intelligent manufacturing. However, the main challenge is how the robot can make online reactive changes to the plan based on the observed human behavior to ensure the completion of user-defined tasks. Such a challenge is further exacerbated if eye-in-hand manipulation is considered since the local field of the camera view cannot capture global observations. Different from existing planning approaches that separate the perception and planning modules, and make strong assumptions about perception abilities, we develop a framework of real-time local reactive planning that enables the robot to quickly adapt its actions if necessary through its limited perception of surroundings using an eye-in-hand camera. Specifically, we develop a locally observable transition system (LOTS) and interpretably express the task using linear temporal logic (LTL). To improve the grasping performance using local visual perception, we propose a high-resolution grasp network (HRG-Net) that achieves state-of-the-art results on multiple datasets (99.50% in Cornell and 97.50% in Jacquard and 96% in Graspnet-1Billion) for the task. A physical experiment using a 7DoF Franka Emika Panda robot demonstrates the effectiveness of the reactive planning framework. Note to Practitioners—Intelligent manufacturing often requires the human operator to work collaboratively with the robot in a shared workspace. Due to possible (assistive or non-assistive) interference of human operators, it is highly desired that the robot can perceive human behaviors and react properly to ensure task accomplishment. Hence, this work is particularly motivated to develop a reactive planning framework that relies on real-time local visual perception (i.e., eye-in-hand camera) to quickly react to its dynamic surroundings and replan its motion when necessary. In future work, rather than using the observed human behavior, we will investigate how to predict human intentions to further improve human-robot collaboration.}, archive = {J_TASE}, author = {Zhangli Zhou and Shaochen Wang and Ziyang Chen and Mingyu Cai and Hao Wang and Zhijun Li and Zhen Kan}, doi = {10.1109/TASE.2023.3304842}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {643--655}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Local observation based reactive temporal logic planning of human-robot systems}, volume = {22}, year = {2025}, }
• (2025). New SLIM method for human reliability analysis based on consensus analysis and combination weighting. TASE, 22, 631--642. (www)

  @article{ , abstract = {Human reliability analysis (HRA) is a systematic activity to identify, analyze, and reduce human failure events in complex engineering systems. As a widely-used approach in HRA, the success likelihood index method (SLIM) mainly depends on experts’ judgments for the rating and weighting of performance shaping factors (PSFs). Additionally, the estimation precision of human error probabilities (HEPs) is greatly affected by experts’ opinions and PSF weights. In this paper, an improved SLIM by integrating the minimum cost consensus method and the game theory-based combination weighting method is developed to acquire the HEPs of operational tasks under a 2-tuple linguistic environment. To facilitate consensus reaching, the missing information in task state assessment matrices of experts is filled based on their knowledge-match levels and similarity degrees. Then, a minimum cost consensus method is used to resolve the conflicts among domain experts considering individual harmony. Besides, a game theory-based combination weighting method is introduced to determine the relative weights of PSFs by integrating their subjective and objective importance. Finally, the feasibility and advantages of the proposed SLIM are validated via an empirical case concerning bunkering operation on board. Note to Practitioners—This research is motivated by a human error assessment issue during the shipboard bunkering operation. It develops a new SLIM based on the minimum cost consensus method and the game theory-based combination weighting method. The purpose of this work is to reduce high-risk human errors in the bunkering operation and enhance maritime transportation safety. Using the proposed SLIM, practitioners can handle unknown information and discrepancies in experts’ opinions and effectively quantify human errors. The application results prove the superiority of the new model. Furthermore, the proposed SLIM can be modified and applied to the socio-technical system in other industries.}, archive = {J_TASE}, author = {Hua Shi and Jing-Hui Wang and Hu-Chen Liu and Song-Man Wu}, doi = {10.1109/TASE.2024.3392742}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {631--642}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {New SLIM method for human reliability analysis based on consensus analysis and combination weighting}, volume = {22}, year = {2025}, }
• (2025). Toward human motion digital twin: A motion capture system for human-centric applications. TASE, 22, 619--630. (www)

  @article{ , abstract = {Following the rule of human-centricity, Human Motion Digital Twin (HMDT) attempts to apply human motion data to ensure the development and well-being of human beings. Particularly, perception and estimation of human motion play fundamental roles in realizing HMDT. This work proposes an inertial motion capture system for human motion digital twin (InMoDT). The designed motion capture device is made up of a hub node and inertial measurement units attached to the human body. The proposed algorithm framework supported by sensor fusion and pose calibration algorithms, enables to acquire orientations of sensors and body segments. With the deployment of algorithms, InMoDT achieves an average root mean square error of 4.7° in estimating orientations when compared with an optical motion capture system. Experimental results show a great correlation (92.5%) and agreement (97.8%) between InMoDT and the optical system. The abilities of InMoDT are spotted in terms of human-centric applications based on the integration of human, cyber system, and physical system, such as motion monitoring and estimation, and human-robot teleoperation. Note to Practitioners—This paper is motivated by the problem of inadequate attention on humans in Cyber-Physical System (CPS) while the roles of operators have a significant effect on industry. With the popular applications of digital twins in CPS, HMDT is expected to monitor, analyze, and assess motion data for facilitating the Human-Cyber-Physical System (HCPS). In this research work, the authors present a system for whole-body motion capture. The proposed system based on a wearable inertial sensor-based device provides a solution to construct HMDT. Moreover, a novel algorithm framework is employed, which consists of the sensor fusion algorithm and the kinematic constraints-based pose calibration algorithm. Experimental results demonstrate the system’s effectiveness in motion sensing accuracy, correlation, and agreement in comparison with the gold standard. The validated applications of the system lie in motion monitoring and estimation, and human-robot teleoperation, showing the potential for enhancing HMDT.}, archive = {J_TASE}, author = {Huiying Zhou and Longqiang Wang and Gaoyang Pang and Huimin Shen and Baicun Wang and Haiteng Wu and Geng Yang}, doi = {10.1109/TASE.2024.3363169}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {619--630}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Toward human motion digital twin: A motion capture system for human-centric applications}, volume = {22}, year = {2025}, }
• (2025). Addressing a collaborative maintenance planning using multiple operators by a multi-objective metaheuristic algorithm. TASE, 22, 606--618. (www)

  @article{ , abstract = {Selective maintenance has a significant impact on the sustainable management of maintenance operations. The collaboration of multiple maintenance teams/operators is helpful to achieve sustainability for selective maintenance sequence planning. For products with a large number of components, a single maintenance team/operator is inefficient due to a long completion time which is not acceptable for emergency planning. Providing specific and efficient maintenance sequence planning is critical to effectively handle different types of emergencies (e.g., wartime) while avoiding vague task assignments to multiple maintenance teams/operators. For scheduling many maintenance jobs while improving the efficiency and quality of maintenance operations, this study proposes a collaborative maintenance planning based on the concept of imperfect maintenance. In this regard, this study develops a multi-objective optimization model to optimize parallel maintenance sequences considering maintenance profit, maintenance cost, maintenance team, and resource limitations. We show the feasibility of the proposed multi-objective optimization model through a real case of maintenance practice for the components of an assistor device. For analyzing the complexity of the proposed maintenance sequence planning problem, this study introduces a new multi-objective metaheuristic algorithm which is an enhanced multi-objective gravitational search algorithm (EMOGSA) to find high-quality Pareto solutions for the proposed problem. Different multi-objective evaluation metrics are used to study the performance of the proposed algorithm. From the results, the proposed model and developed solution algorithm can help maintenance decision-makers to determine complex maintenance planning. Note to Practitioners—This paper deals product with a maintenance and proposes gravitational search algorithm based on only maintenance task, which maintenance task. The goal of this paper is to analyze the maintenance problem from the perspective of collaboration of multiple maintenance teams/operators.}, archive = {J_TASE}, author = {Guangdong Tian and Lele Zhang and Amir M. Fathollahi-Fard and Qi Kang and Zhiwu Li and Kuan Yew Wong}, doi = {10.1109/TASE.2023.3269059}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {606--618}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Addressing a collaborative maintenance planning using multiple operators by a multi-objective metaheuristic algorithm}, volume = {22}, year = {2025}, }
• (2025). Safety compliant, ergonomic and time-optimal trajectory planning for collaborative robotics. TASE, 22, 594--605. (www)

  @article{ , abstract = {The demand for safe and ergonomic workplaces is rapidly growing in modern industrial scenarios, especially for companies that intensely rely on Human-Robot Collaboration (HRC). This work focuses on optimizing the trajectory of the end-effector of a cobot arm in a collaborative industrial environment, ensuring the maximization of the operator’s safety and ergonomics without sacrificing production efficiency requirements. Hence, a multi-objective optimization strategy for trajectory planning in a safe and ergonomic HRC is defined. This approach aims at finding the best trade-off between the total traversal time of the cobot’s end-effector trajectory and ergonomics for the human worker, while respecting in the kinematic constraint of the optimization problem the ISO safety requirements through the well-known Speed and Separation Monitoring (SSM) methodology. Guaranteeing an ergonomic HRC means reducing musculoskeletal disorders linked to risky and highly repetitive activities. The three main phases of the proposed technique are described as follows. First, a manikin designed using a dedicated software is employed to evaluate the Rapid Upper Limb Assessment (RULA) ergonomic index in the working area. Next, a second-order cone programming problem is defined to represent a time-optimal safety compliant trajectory planning problem. Finally, the trajectory that ensures the best compromise between these two opposing goals–minimizing the task’s traversal time and maintaining a high level of ergonomics for the human worker– is computed by defining and solving a multi-objective control problem. The method is tested on an experimental case study in reference to an assembly task and the obtained results are discussed, showing the effectiveness of the proposed approach. Note to Practitioners—Health and safety in workplaces are business imperatives, since they ensure not only a safe collaboration between industrial machinery and human operators, but also an increased productivity and flexibility of the entire industrial process. Hence, investing in health is a real driver for business growth. The key enabling technologies of Industry 4.0, such as collaborative robotics, exoskeletons, virtual and augmented reality, require standardization and indispensable technical safety requirements that cannot ignore physical, sensory, and psychological peculiarities of the human worker and aspects like usability and acceptability of these technologies in performing their activities. Against this ongoing industrial challenge, the aim of this paper is to provide researchers and practitioners with an innovative HRC trajectory planning methodology focused on enhancing production efficiency while respecting the SSM ISO safety requirement and guaranteeing the ergonomic optimal position of the operator during an assembly task. Therefore, the proposed methodology can be a convenient solution to be deployed in industrial companies, since it can support human operators by drastically reducing work-related musculoskeletal disorders and augmenting their performance in the working environment.}, archive = {J_TASE}, author = {Silvia Proia and Graziana Cavone and Paolo Scarabaggio and Raffaele Carli and Mariagrazia Dotoli}, doi = {10.1109/TASE.2023.3331505}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {594--605}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Safety compliant, ergonomic and time-optimal trajectory planning for collaborative robotics}, volume = {22}, year = {2025}, }
• (2025). A knee-guided evolutionary algorithm based navigation approach for mobile robots in intelligent manufacturing scenarios. TASE, 22, 582--593. (www)

  @article{ , abstract = {This paper proposes a novel navigation and optimization approach for mobile robots operating in intelligent manufacturing (IM) scenarios with unknown obstacles. First, we introduce a set of evaluation functions that simultaneously consider multiple metrics, including speed, security, sampling step, and obstacle avoidance ability, for scenarios with unknown dynamic obstacles. We then adopt a knee-guided multi-objective evolutionary algorithm capable of balancing parameter size and performance to optimize these conflicting objectives. Finally, we present a local navigation framework that integrates navigation and multi-objective optimization to enable obstacle avoidance in unstructured environments. The proposed algorithm is validated through simulations and practical scenarios, demonstrating its effectiveness in both static and dynamic scenarios.Note to Practitioners—The Human-cyber-physical System (HCPS) is the basic principle of the new generation of intelligent manufacturing (IM). It presents new requirements for the production method, and robot operate completely autonomously is expected to be the future solution for some complex tasks. However, traditional navigation methods often fail in these scenarios due to environmental changes and restricted workspaces. To address this challenge, this paper proposes a multi-objective optimization-based local navigation approach (MONA) that transforms the local navigation problem into a multi-objective optimization problem. The approach uses the knee-guided multi-objective evolutionary algorithm to optimize the control input for the mobile robot, ensuring safe and fast navigation in complex environments. The proposed approach has been validated through simulation and real scenario experiments, demonstrating its effectiveness.}, archive = {J_TASE}, author = {Yunhao Xia and Zhijun Li and Gary G. Yen and Haisheng Xia}, doi = {10.1109/TASE.2024.3394715}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {582--593}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A knee-guided evolutionary algorithm based navigation approach for mobile robots in intelligent manufacturing scenarios}, volume = {22}, year = {2025}, }
• (2025). Resilient event-triggered terminal sliding mode control design for a robot manipulator. TASE, 22, 570--581. (www)

  @article{ , abstract = {A novel non-singular terminal sliding mode controller (NTSMC) has been developed for the purpose of tracking and stabilizing tasks in uncertain electro-hydraulic robot manipulators. It is supposed that the controller communicates with the robot through a network that is vulnerable to cyber-attacks. To reduce the communication burden on the network layer and achieve resiliency against cyber-attacks, an event-based strategy is employed. For this purpose, the event-triggering rule is derived so that the Zeno-free behavior is guaranteed. Then, based on the cyber-attack characteristics, i.e., frequency and duration of the attacks, the resilient behavior of the proposed scheme in the presence of denial of service attacks, unmodelled dynamics, and external disturbance are analyzed. Moreover, to capture the nonlinear nature of the robot an experimentally validated analytical model of an uncertain 7-DoF manipulator with a hydraulic model of the joints and actuators, namely, Brokk-Hydrolek, is employed. Finally, the merits of the proposed methodology in terms of resiliency, robustness, and preservation of the communication resources are validated, and the results are compared to the state-of-the-art approaches based on the $\rho $ index criterion. Note to Practitioners—The aim of this study is to address the problem of network-based control of robotic manipulators. These systems, relying on the network layer for data collection and control commands, are highly vulnerable to catastrophic cyber-attacks. Furthermore, they should comply with network restrictions, such as limited bandwidth, to achieve the desired performance. Therefore, due to the collaborative behavior of robot manipulators in industries, it is vital for engineers and practitioners to be assured of achieving desired performance in the presence of these threats and limitations. As a first step to deal with these issues, a 7-DoF robotic manipulator model is mathematically formulated and experimentally validated. Then, a controller design procedure that guarantees the desired performance in spite of model uncertainties, denial-of-service cyber-attacks, and network restrictions is derived. Additionally, a clear relation between cyber-attack characteristics and designed parameters is defined while resilient behavior is maintained. Note that the proposed approach can be applied to a wide range of network-based nonlinear dynamic systems.}, archive = {J_TASE}, author = {Mobin Saeedi and Jafar Zarei and Mehrdad Saif and Declan Shanahan and Allahyar Montazeri}, doi = {10.1109/TASE.2023.3297119}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {570--581}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Resilient event-triggered terminal sliding mode control design for a robot manipulator}, volume = {22}, year = {2025}, }
• (2025). DuAK: Reinforcement learning-based knowledge graph reasoning for steel surface defect detection. TASE, 22, 557--569. (www)

  @article{ , abstract = {Surface defect is a crucial factor affecting the product quality of steel products. Current studies mainly focus on defect recognition and classification using machine vision-based algorithms, which lack the trace of potential causes and the reuse of experiential knowledge. To address this issue, we construct a knowledge graph for steel surface defects by fusing the multi-source and heterogeneous industrial data, including process parameters, chemical compositions, defect images, operation logs and empirical knowledge. A policy-based reinforcement learning approach is developed to solve the path reasoning problem over the industrial knowledge graph in defect detection and diagnosis. The approach employs two agents to explore the path efficiently from opposite directions, utilizes an integrated reward function that comprehensively considers the path direction, path length and entity distance to perform action selection, and adopts the path sharing mechanism and the prior knowledge to update selection policy. Experimental comparisons with the state-of-the-art knowledge reasoning algorithms on two benchmark datasets, NELL-995 and FB15K-237, validate the performance and merits of the proposed method. The effectiveness of the proposed method is also evaluated on a practical steel surface defect dataset, and the results show that our approach performs well in knowledge reasoning on the surface defect graph. Note to Practitioners—The surface quality of products has become a widely concerned focus in manufacturing industries. With the development of industrial IoT and Cyber-physical system technologies, more and more industrial data has been collected, and machine learning-based algorithms have been developed and applied to the recognition of detect defects. However, the algorithms do not take full advantage of the multi-source and heterogeneous defect-related data. On the other hand, it is also difficult to accumulate, inherit and reuse the experts’ knowledge of solving historical cases in the long-term production process. In order to deal with the above obstacles, we apply the knowledge graph for steel surface defect detection. In the proposed approach, a policy-based reinforcement learning algorithm is developed to solve the path reasoning problem over the industrial knowledge graph. To further improve the performance of our algorithm, we employ two agents to explore the path efficiently from opposite directions, utilize an integrated reward function which comprehensively considers the path direction, path length and entity distance to perform action selection, adopt the path sharing mechanism and updated selection policy to reuse the prior knowledge. As a result, our algorithm obtains high precision in knowledge reasoning tasks on two benchmark datasets and a practical steel surface defect dataset compared with some existing algorithms. Hence, it can be readily applied to real surface defect detection problems and facilitates intelligent manufacturing in steel production.}, archive = {J_TASE}, author = {Yufei Zhang and Hongwei Wang and Weiming Shen and Gongzhuang Peng}, doi = {10.1109/TASE.2023.3307588}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {557--569}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {DuAK: Reinforcement learning-based knowledge graph reasoning for steel surface defect detection}, volume = {22}, year = {2025}, }
• (2025). Modeling and analysis of a petri net-based system supporting implementation of additive manufacturing technologies. TASE, 22, 546--556. (www)

  @article{ , abstract = {The paper proposes a novel modelling and analysis technique of the system supporting implementation of additive manufacturing (AM) technologies specified by the interpreted Petri net. Nowadays, there is a need in a manufacturing enterprise to implement the new technologies supporting the realization of production, especially in terms of the possibility of unplanned crisis events. Managers expect such models to support their decisions in order to maintain or strengthen their competitive position thanks to the investments. Therefore, this paper proposes a system supporting the decision to implement AM technology, based on the interpreted Petri net, by utilizing its main advantages: the possibility of graphical modelling, and strong mathematical support of formal verification techniques. In particular, the proposed verification algorithm allows for detection of possible unbounded places. Contrary to the most popular analysis methods (which are bounded exponential in general case), it is proved that the presented technique is bounded by a cubic polynomial with the number of Petri net nodes. Finally, the presented modelling and analysis method is illustrated by a real-life case-study example of the decision-making model to support the implementation of AM technologies. Note to Practitioners—This paper is motivated by the limited performances in the existing modelling and analysis techniques of systems that support implementation of additive manufacturing (AM) technologies. In particular, the article proposes a system supporting the decision to implement AM technology, based on the interpreted Petri net. Existing approaches have important restrictions related to the exponential computational complexity of analysis methods. Therefore, verification of the model can be a real challenge to the designer. In contradistinction, the algorithm proposed in the paper permits the examination of the system in a polynomial time. The presented modelling and analysis ideas are illustrated by a real-life example of the decision-making system that utilizes implementation of AM technology. Managers of manufacturing enterprises, thanks to the use of the proposed approach, can assess the investment in AM technology in the following three unexpected situations: 1) delays in the delivery of materials and/or semi-finished products or the inability to receive materials and/or semi-finished products; 2) high employee turnover; 3) increasing costs of energy consumption. Thus, the application of this model allows for the harmonious conduct of manufacturing activities in conditions of unexpected situations.}, archive = {J_TASE}, author = {Remigiusz Wiśniewski and Justyna Patalas-Maliszewska and Marcin Wojnakowski and Marcin Topczak}, doi = {10.1109/TASE.2023.3301304}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {546--556}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Modeling and analysis of a petri net-based system supporting implementation of additive manufacturing technologies}, volume = {22}, year = {2025}, }
• (2025). Whole-body safety-critical control design of an upper limb prosthesis for vision-based manipulation and grasping. TASE, 22, 534--545. (www)

  @article{ , abstract = {In this paper, an upper limb prosthesis has been furnished with a novel vision-based manipulation and grasping strategy. The proposed whole-body safety-critical control design includes vision servoing, multiple tasks planning with strict priorities, which can be formulated as an hierarchical multi-task optimization (HMO) problem with safety conditions—expressed as control barrier functions (CBF). Firstly, a modified YOLOv7 algorithm with key points detection is developed to determine the grasping pattern of the object and extract its edge contour information using a depth camera. An HMO-based strategy with a notion of CBF, providing inequality constraints in the control input, is proposed to handle multiple prioritized tasks with various constraints to offer guarantees of safety with the whole-body motion in consideration. Then the HMO problem is solved by a neuro-dynamics optimization solution online. Finally, experiments are implemented by using a self-developed upper limb prosthesis. Experimental results validate the performance of the proposed whole-body control strategy. Note to Practitioners—The intuitive, convenient and autonomous control of prosthetics has always been the object of researchers’ efforts. This paper proposes a whole-body safety-critical control design, including artificial perception system, autonomous control system under visual guidance, and user volition control system. The HMO strategy with safety conditions-expressed as control barrier functions in the context of real-time optimization-based method can fully consider the task requirements of different priorities and realize the orderly execution of prosthetic tasks. The vision feedback can detect surrounding environments in real time to obtain fine position information and optimal grasping pattern, which are conducive to reduce the amputees’ cognitive burden and provide new ideas for the whole body control of prostheses.}, archive = {J_TASE}, author = {Zhijun Li and Jin Huang and Peihao Zhang and Peng Shi}, doi = {10.1109/TASE.2024.3412823}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {534--545}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Whole-body safety-critical control design of an upper limb prosthesis for vision-based manipulation and grasping}, volume = {22}, year = {2025}, }
• (2025). Hierarchical data fusion-based health state assessment of nuclear power plant operation under human-cyber-physical system architecture. TASE, 22, 523--533. (www)

  @article{ , abstract = {Human-cyber-physical system has emerged as a pivotal tool for intelligent production throughout the whole lifecycle. In this paper, a health state assessment approach based on hierarchical data fusion is proposed under human-cyber-physical system architecture. The human-cyber-physical system emphasizes the flexibility of dynamic operating states and the evolvability of the enabling models. First of all, a multiple hierarchy consisting of equipment, systems and function is established as a physical basis for human-cyber-physical system. Regarding the equipment-level assessment characterized by extensive input parameters, a dynamic weighting method is exploited to highlight the abnormal parameters. Thus, the degraded device can be significantly reflected in the system-level assessment by fusing equipment parameters. To evaluate the operation function of a plant, a Nearest Class Mean classifiers-based Random Forests algorithm is explored considering the small-scale and imbalanced instances in different health levels. However, the conventional fixed and changeable models threaten the accuracy and rationality of the human-cyber-physical system. We introduce incremental learning to reform Random Forests that can update promptly with new data and knowledge. Finally, the proposed scheme and technologies are applied in a nuclear power plant to demonstrate their effectiveness. Note to Practitioners—The modern production desires to be further intelligent for improvement of economic benefits but with the consistent requirements of high safety and reliability. The trades-off between efficiency and safety bring the greater challenges to the operation and maintenance for production. This paper proposes an evolvable human-cyber-physical system scheme that not only integrates more intelligent elements but also attaches some solution to the abovementioned problem. The scheme is implemented on the multi-hierarchy health states assessment which is the core in production maintenance against lagging. This equipment-system-function fusion and assessment provide a unified platform for multi-departments to collaborate and self-manage. The novel weighting method abandons the conventional subjective evaluation criteria with high time consumption. Computers are convenient to possess sensing data efficiently and highlight the anomaly obviously. In particular, it is mentioned that the intelligent models in human-cyber-physical systems are always fixed and unable to alter with the physical system changes. Incremental learning, a promising branch of machine learning, is introduced to reform the assessment model to be stainable developed with new data adding. The scheme takes a nuclear plant as example because it is more corresponding to these issues. The future studies can extend the scheme to other scenarios.}, archive = {J_TASE}, author = {Xiangyu Jiang and Yixiong Feng and Zhiwu Li and Zhaoxi Hong and Hengyuan Si and Jianrong Tan}, doi = {10.1109/TASE.2023.3321049}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {523--533}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Hierarchical data fusion-based health state assessment of nuclear power plant operation under human-cyber-physical system architecture}, volume = {22}, year = {2025}, }
• (2025). Autoencoder-embedded iterated local search for energy-minimized task schedules of Human–Cyber–Physical systems. TASE, 22, 512--522. (www)

  @article{ , abstract = {This work considers a task scheduling problem with deadline constraints in human-cyber-physical systems. To find its energy-efficient schedules in a short time, an autoencoder-embedded iterated local search algorithm is proposed to solve it. Iterated local search is selected as a main scheduler. In order to handle real-time requirements and high computational load involved in the problem solution, a Long Short-Term Memory-based AutoEncoder model (LSTM-AE) is constructed to capture the relevant and complementary features of the considered problem. The model is used to find low-order and high-fit sub-solutions via unsupervised end-to-end learning, and generates promising solutions in an informative low-dimensional solution space. To further reduce computational burden, a two-stage optimization framework is constructed, which includes an off-line training phase and an online optimization one. The former trains LSTM-AE by using expert knowledge and historical data. The latter designs optimal resource allocation strategies to build a high-quality initial solution. Then, LSTM-AE-assisted local search operators are proposed and used to reform the initial solution and generate better ones. Various numerical experiments are performed to compare the proposed method with several classic heuristics and some recently-developed methods. The results show its superiority over them.Note to Practitioners—In human-cyber-physical systems, a task scheduling problem is usually solved by using heuristics due to the limited computational resources. Nevertheless, fast dispatching rules tend to perform poorly. Meta-heuristics can find a relatively high-quality schedule but are time-consuming, especially for a population-based algorithm that requires to evaluate a fitness function for many candidate solutions at each iteration. To balance computational burden and solution quality, our idea is to combine machine-learning methods with meta-heuristics. Specially, we integrate a long short-term memory-based autoencoder model into iterated local search to improve the latter’s optimization ability. The combination of meta-heuristics and machine learning techniques makes it possible to obtain a high-quality schedule for the concerned problems in a short time. Theoretic analysis and experimental results show that the proposed method well outperforms its competitive peers.}, archive = {J_TASE}, author = {Chengran Lin and ZhengCai Cao and MengChu Zhou}, doi = {10.1109/TASE.2023.3267714}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {512--522}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Autoencoder-embedded iterated local search for energy-minimized task schedules of Human–Cyber–Physical systems}, volume = {22}, year = {2025}, }
• (2025). Human-cyber-physical system for industry 5.0: A review from a human-centric perspective. TASE, 22, 494--511. (www)

  @article{ , abstract = {Industry 5.0 heralds a new wave of the industrial revolution, placing a spotlight on human-centric intelligent manufacturing. At the core of Industry 5.0 lies the human-cyber-physical system (HCPS), a composite intelligent system where interactions among humans, cyberspace, and physical assets are orchestrated across diverse manufacturing levels and phases. Understanding the pivotal roles played by humans in these advanced systems is of paramount importance. Nonetheless, the exploration of HCPS within the context of Industry 5.0 remains in its infancy. This paper presents a holistic literature review of industrial HCPS from a human-centric perspective. A united architecture is employed to encompass the aspects of cognitive-to-technology integration and human-to-human interaction in HCPS, highlighting human-in-the-loop, human-on-the-loop, and human-in-the-society paradigms. The mechanisms of these paradigms and their effects on design, production, and service are investigated to expand the research landscape of intelligent manufacturing in Industry 5.0. Key enabling technologies that facilitate harmonious tri-space integration are introduced, and the future challenges of industrial HCPS are discussed. This work is expected to attract more open discussions and in-depth research on HCPS in the new industrial revolution era. Note to Practitioners—This paper is motivated by the emergence of Industry 5.0 that integrates humans into cyber-physical systems to offset drawbacks on both sides. It presents an overview of HCPS-related works to identify the state-of-the-art and open problems in the Industry 5.0 era. The review of HCPS applications in the design, production, and service phases can benefit engineers in the intelligent manufacturing area. Key enabling technologies on human ability augmentation, human-robot interaction, digital twin, human-cyber-physical data fusion, crowdsourcing, and system modeling, are analyzed to attract researchers in broader research fields to join in the development of industrial HCPS.}, archive = {J_TASE}, author = {Shanhe Lou and Zhongxu Hu and Yiran Zhang and Yixiong Feng and MengChu Zhou and Chen Lv}, doi = {10.1109/TASE.2024.3360476}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {494--511}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Human-cyber-physical system for industry 5.0: A review from a human-centric perspective}, volume = {22}, year = {2025}, }
• (2025). Guest editorial: Human-cyber-physical systems for intelligent manufacturing: An emerging area. TASE, 22, 489--493. (www)

  @article{ , archive = {J_TASE}, author = {MengChu Zhou and Yixiong Feng and Jan Faigl and Chen Lv and Weihong Grace Guo}, doi = {10.1109/TASE.2024.3523762}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {489--493}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guest editorial: Human-cyber-physical systems for intelligent manufacturing: An emerging area}, volume = {22}, year = {2025}, }
• (2025). Jointed observer-based sliding mode predictive control for interconnected power systems with input delays. TASE, 22, 480--488. (www)

  @article{ , abstract = {This paper is concerned with the sliding mode predictive control problem for the load frequency control (LFC) of interconnected power systems with input delays and load disturbance. To deal with input delays, unknown disturbance, the interconnected influence existing simultaneously in power systems, the state-observer and disturbance-observer are respectively designed by employing the reduction transformation technique for state observer and introducing the sliding mode surface for the extended disturbance observer. Then the jointed observers based integral sliding mode controller with active disturbance compensation is designed. Furthermore, by constructing a novel Lyapunov function and matrix inequality technique, the sufficient conditions of exponential stability are provided for the closed-loop observer error systems in linear matrix inequality form, which can be solved by linear matrix inequality (LMI) toolbox in Matlab. Simulation is given to verify the effectiveness of the proposed approach. Note to Practitioners—This paper was motivated by the practical control problem of load frequency control systems with input delay. The input delay, unknown load disturbance and frequency fluctuation of tie-lines make it difficult to design a suitable control for load frequency control systems. The main practical contributions are summed as follows: A new jointed observer-based sliding mode predictive control approach is proposed to the interconnected power systems with input delays. By employing the reduction transformation technique, a jointed observer based integral sliding predictive mode controller with active disturbance compensation is developed to deal with input delay, unknown disturbance, the interconnected influence existing simultaneously in practical power systems.}, archive = {J_TASE}, author = {Yingchun Wang and Yang Liu and Xiaoqi Yu and Huaguang Zhang}, doi = {10.1109/TASE.2024.3353166}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {480--488}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Jointed observer-based sliding mode predictive control for interconnected power systems with input delays}, volume = {22}, year = {2025}, }
• (2025). A task-adaptive deep reinforcement learning framework for dual-arm robot manipulation. TASE, 22, 466--479. (www)

  @article{ , abstract = {Closed-chain manipulation occurs when several robot arms perform tasks in cooperation. It is complex to control a dual-arm system because it requires flexible and adaptable operation ability to realize closed-chain manipulation. In this study, a deep reinforcement learning (DRL) framework based on actor-critic algorithm is proposed to drive the closed-chain manipulation of a dual-arm robotic system. The proposed framework is designed to train dual robot arms to transport a large object cooperatively. In order to sustain strict constraints of closed-chain manipulation, the actor part of the proposed framework is designed in a leader-follower mode. The leader part consists of a policy trained from the DRL algorithm and works on the leader arm. The follower part consists of an inverse kinematics solver based on Damped Least Squares (DLS) and works on the follower arm. Two experiments are designed to prove the task adaptability, one of which is manipulating an object to a random pose within a defined range, the other is manipulating a delicate structural object within a narrow space. Note to Practitioners—In common industrial manipulation scenarios, there are requirements to employ robotic arms to transport a large object relative to the robotic arm, e.g., moving a payload onto a loader and assembling big craft parts. It is a cost-effective way to use a dual-arm system to extend the loading capacity of robotic arms while preserving the flexibility of manipulation. Moreover, the dual-arm system is expected to manipulate different objects without complicated reprogramming, especially in small batch production scenarios. This study proposes a task-adaptive deep reinforcement learning framework for dual-arm robot manipulation. The task adaptability includes two specific aspects, one being adaptability in targeting the pose, such as manipulating an object to a random pose within a specified range. The other is the adaptivity on the task prerequisites such as manipulating a delicate structural object within a narrow space. For future research, the dual-arm system may autonomously plan the grab positions, and additional investigations should address more common scenarios involving various object shapes.}, archive = {J_TASE}, author = {Yuanzhe Cui and Zhipeng Xu and Lou Zhong and Pengjie Xu and Yichao Shen and Qirong Tang}, doi = {10.1109/TASE.2024.3352584}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {466--479}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A task-adaptive deep reinforcement learning framework for dual-arm robot manipulation}, volume = {22}, year = {2025}, }
• (2025). Guided model-based policy search method for fast motor learning of robots with learned dynamics. TASE, 22, 453--465. (www)

  @article{ , abstract = {Reinforcement learning recently has achieved impressive success in allowing robots to learn complex motor skills in simulation environments. However, most of these successes are difficult to transfer to physical robots since current algorithms require lots of practical training and complex sim-to-real transfer skills. To improve the learning efficiency and adaptability of physical robots, this article proposes a guided model-based policy search (GMBPS) algorithm inspired by a hypothetical model-free (MF) and model-based (MB) actor-critic brain implementation. This approach bridges the gap between MF and MB control processes, overcoming the suboptimality of MB methods and speeding up the learning rate of MF methods. Additionally, a one-step predictive control framework is proposed for minimizing the impact of delayed sensorimotor information in real-world tasks. This helps to accurately control the action cycle time and ensures the feasibility of MB planning for physical robots. The simulation and experimental results demonstrate that the proposed approach enables a 6-DOF UR5e robot arm to learn various reaching tasks in a few minutes with better policies and higher learning efficiency. Note to Practitioners—Reinforcement learning is becoming a popular framework that allows robots to learn complex motor skills without building analytical models of controlled plants. However, low learning efficiency severely limits its application in practical robots, where robots have to quickly adapt to dynamically changing environments in micro-data situations. To solve the inefficiency problem of physical robot learning from scratch, this paper proposes a MF and MB fusion control algorithm inspired by a hypothetical MF and MB actor-critic brain implementation. The motion decision process is modeled as an optimization problem with inequality constraints. The global MF value function is incorporated into the MB objective function, extending the short-term optimization into a long-term version to overcome the suboptimality of conventional MB methods. The MB policy is searched based on the quadratic penalty method with the guide of the MF policy, which helps improve the quality of policy at every decision-making step. Moreover, since the model dynamics is fitted by a probabilistic neural network, the proposed method is not only applicable to joint-driven robots but also provides a feasible solution for the control of various robotic systems with complex dynamics, such as soft robots and musculoskeletal robots.}, archive = {J_TASE}, author = {Xiao Huang and Xingfang Wang and Yan Zhao and Jiachen Hu and Hui Li and Zhihong Jiang}, doi = {10.1109/TASE.2024.3352580}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {453--465}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Guided model-based policy search method for fast motor learning of robots with learned dynamics}, volume = {22}, year = {2025}, }
• (2025). A less conservative robust control approach and its application to truck-trailer system. TASE, 22, 442--452. (www)

  @article{ , abstract = {This paper presents a robust $H_{\infty }$ performance analysis method for an uncertain linear parameter-varying (LPV) system to achieve less conservative disturbance rejections. Specifically, by adopting Lyapunov function, a feedback controller is designed based on linear matrix inequalities (LMIs) framework to eliminate the influences of unknown uncertainties for continuous-time and discrete-time LPV systems. Meanwhile, a convex optimization technique is explored for solving optimization problems of bilinear matrix inequalities (BMIs), in which several sufficient conditions guaranteeing system stability are relaxed. Furthermore, compared with some existing approaches, the superior performances of the presented scheme are demonstrated by rigorous theoretical analysis and numerical simulations on a practical truck-trailer system. Note to Practitioners—This paper is motivated by the problem of minimal control domain range (also called conservativeness) in robust $H_{\infty }$ control approaches for LPV systems. Generally, LMI conditions are accompanied by a certain level of conservativeness for the robust stability of LPV systems. However, excessive conservatism makes the proposed control strategy impossible to implement in practical engineering. Considering this gap, we formulate a lemma to reduce the conservativeness in robust stability conditions while dealing with the effects of system uncertainty. Strict theoretical proofs and comparative simulation experiments are also conducted to verify the effectiveness of the proposed robust control scheme in continuous and discrete LPV systems.}, archive = {J_TASE}, author = {Bing Liu and Zhongmei Li and Rong Nie and Wenli Du}, doi = {10.1109/TASE.2024.3351996}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {442--452}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A less conservative robust control approach and its application to truck-trailer system}, volume = {22}, year = {2025}, }
• (2025). Integral barrier lyapunov function-based adaptive event-triggered control of flexible riser systems. TASE, 22, 433--441. (www)

  @article{ , abstract = {This paper presents an adaptive boundary control design for flexible riser systems with external disturbances and boundary position constraint. An integral barrier Lyapunov function (iBLF) is used to solve the boundary position constraint problem. Since the iBLF directly constrains the boundary position, this relaxes the conservative restriction on the state constraint of conventional BLF control. A new auxiliary signal is designed to offset the effect of the coupling term that cannot be eliminated. Compared to time triggering, the controller and actuator communicate less when using event triggering. Therefore, an event-triggered control scheme is designed to achieve effective suppression of riser vibration by introducing a relative threshold strategy. The Lyapunov stability theory is used to demonstrate that the flexible riser system is finally constrained. The efficiency of the control strategy is then further verified by numerical simulations. Note to Practitioners—This paper investigates the control problem of flexible riser systems with external disturbances. Since the riser needs to consider the uninterrupted marine distribution disturbance and external disturbance, the riser will inevitably distort and vibrate. So the stability of the ship is extremely challenging. The riser is regarded as an Euler-Bernoulli beam construction because of its small diameter and lengthy length. Using Hamilton’s principle, the riser is represented as a fourth-order partial differential equation and two ordinary differential equations. In contrast to the logarithmic BLF and tangent BLF, the integral BLF has direct constraints on the boundary positions, which relaxes the conservative restrictions on state constraints imposed by conventional BLF control. Event-triggered control has attracted the interest of many flexible system control researchers due to its benefits in preserving communication, cost and other resources, and is widely used in practical engineering fields. The complexity and uncertainty attributed to the PDE system itself makes the design of event-triggered strategies more difficult. Based on the original ODE event triggering, it discusses the PDE-based event-triggered strategy of flexible riser systems.}, archive = {J_TASE}, author = {Xin-Yu Zhang and Xiangpeng Xie and Yan-Jun Liu and Jiayue Sun}, doi = {10.1109/TASE.2024.3351739}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {433--441}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Integral barrier lyapunov function-based adaptive event-triggered control of flexible riser systems}, volume = {22}, year = {2025}, }
• (2025). Robust neural differential solution for time-dependent nonlinear optimization with noises rejection. TASE, 22, 418--432. (www)

  @article{ , abstract = {Model uncertainty, exogenous disturbance, and time-dependent computing are significant issues in constrained nonlinear optimization and related practical problems. This paper highlights two limitations of current methods for time-dependent nonlinear optimization, i.e., disturbance handling and restrictions on the objective function and constraints. First, a novel robust neural differential (RND) solution is proposed for solving equality- and inequality-constrained time-dependent nonlinear optimization (EIC-TDNO) problems involving the suppression of nonspecific disturbances. Second, the dynamic vector-valued error equation equivalent to the EIC-TDNO method is derived. Theoretical proofs ensure that the proposed RND solution converges to zero at an exponential rate, eliminates polynomial noise, and effectively reduces the time-varying disturbance. Finally, through simulation and applications to 2-D filter design and robot motion planning, the RND solution is verified to exhibit rapid convergence, high precision, strong robustness, and versatility in comparison with the existing models. Note to Practitioners—The primary impetus behind this paper is the disadvantage of current optimization methods when solving real-world application problems. First, most of the existing optimization methods consider time-invariant objective functions and constraints. Additionally, the latest time-varying optimization methods limit their application to unconstrained or equality-constrained situations and are capable of handling disturbances of only specific types and within specific numerical ranges. Note that the interplay between modules may be introduced if other components, such as filters, are used to handle disturbances. Therefore, in this paper, an optimization approach is intrinsically designed for time-dependent optimization with its own ability to stabilize the whole system while being perturbed by various types of time-varying disturbances, which could reduce large lagged errors and system complexity. Theoretical analysis provides performance guarantees as well as guidance on parameter tuning to ensure solution accuracy and robustness. This paper primarily investigates and verifies how the proposed robust neural differential solution solves constrained time-dependent nonlinear optimization problems with the presence of different time-varying disturbances; this approach can be applied to numerous practical engineering problems, such as robot control, energy dispatch, and image processing.}, archive = {J_TASE}, author = {Lin Wei and Long Jin}, doi = {10.1109/TASE.2024.3351197}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {418--432}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robust neural differential solution for time-dependent nonlinear optimization with noises rejection}, volume = {22}, year = {2025}, }
• (2025). A USDE-based control for bilateral teleoperation systems with time-varying delays. TASE, 22, 407--417. (www)

  @article{ , abstract = {In this paper, a simple but effective control scheme is proposed for bilateral teleoperation systems with variable communication time-delays. The robotic models are first reformulated and an unknown system dynamics estimator (USDE) is constructed to estimate the lumped uncertainties without using the accelerations, where only low-pass filtering operations and trivial algebraic calculations are used. The USDE is then incorporated into the controller design to compensate for the effects of uncertainties. The stability of the closed-loop system is rigorously proven via the Lyapunov theorem for all three motion states: free motion, tracking motion, and contact motion. The salient feature over traditional control strategies is that apart from position and velocity synchronization, compliance can be guaranteed in the tracking motion and force transparency can be obtained in the contact motion. Moreover, the proposed USDE and control have simple structures and less parameters, which facilitate their practical application. Both simulation and experimental results are presented to demonstrate the effectiveness of the proposed methods. Note to Practitioners—Bilateral teleoperation systems are widely used in nuclear science, space research, telesurgery, microassembly and other fields, where human or unmanned robotics cannot perform certain tasks. However, in the teleoperation, several factors such as external forces and communication delays between local and remote robotics will affect the control performance, e.g., synchronization, compliance and transparency. This paper proposes a USDE-based control method that can handle unknown system dynamics and variable time-delays without using the acceleration signals, while the system’s stability in all the three motion states: free motion, tracking motion and contact motion can be retained, and the corresponding position and velocity synchronization objectives can be achieved. This method has the certain advantages including simpler structures, less tuning parameters and better control performances, facilitating its practical applications. Finally, both simulation and experiments are illustrated to verify the theoretical claims.}, archive = {J_TASE}, author = {Baorui Jing and Jing Na and Hao Duan and Xian Wang and Yingbo Huang}, doi = {10.1109/TASE.2024.3351157}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {407--417}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A USDE-based control for bilateral teleoperation systems with time-varying delays}, volume = {22}, year = {2025}, }
• (2025). An adaptive generalized super-twisting algorithm via event-triggered control. TASE, 22, 393--406. (www)

  @article{ , abstract = {This paper presents a novel event-triggered adaptive generalized super-twisting algorithm (ET-AGSTA) for uncertain nonlinear systems. The remarkable features of the developed method lie in that the control gains can be dynamically adjusted to avoid overestimation, and the transmission of unnecessary signals can be reduced to save computational costs and network resources. Under the proposed ET-AGSTA, it can be proved by the Lyapunov theory that the sliding variable finite-time converges to a desired domain around the origin. The size of the domain can be achieved by modifying the control parameters. Moreover, the lower bound of the triggering time intervals is verified to be always a positive constant, which guarantees Zeno-free behavior in the whole system. Finally, simulation results are given to demonstrate the effectiveness of the proposed scheme. Note to Practitioners— In this paper, the motivation is to handle the problem of the ET-AGSTA design for uncertain nonlinear systems. Although the super-twisting algorithm (STA) is an effective tool for studying uncertain nonlinear systems, there exist two problems in practical applications. On the one hand, the control gains will be overestimated because the upper bounds of the derivatives of the uncertainties are often unknown. This will lead to severe chattering due to the discontinuous term existing in the STA. On the other hand, in actual scenarios, the state signals are periodically sampled and then transmitted to the controllers through the limited network. As the sampling period is usually fixed, it will yield the transmission of redundant signals, resulting in a waste of resources. Therefore, based on the Lyapunov theory, a novel ET-AGSTA is designed to overcome the drawbacks caused by the discontinuous term in the STA. Future work will apply the proposed method to the robotic systems.}, archive = {J_TASE}, author = {Wenhui Dou and Shihong Ding and Ju H. Park and Keqi Mei}, doi = {10.1109/TASE.2024.3351122}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {393--406}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An adaptive generalized super-twisting algorithm via event-triggered control}, volume = {22}, year = {2025}, }
• (2025). Robot-assisted deep venous thrombosis ultrasound examination using virtual fixture. TASE, 22, 381--392. (www)

  @article{ , abstract = {Deep Venous Thrombosis (DVT) is a common vascular disease with blood clots inside deep veins, which may block blood flow or even cause a life-threatening pulmonary embolism. A typical exam for DVT using ultrasound (US) imaging is by pressing the target vein until its lumen is fully compressed. However, the compression exam is highly operator-dependent. To alleviate intra- and inter-variations, we present a robotic US system with a novel hybrid force motion control scheme ensuring position and force tracking accuracy, and soft landing of the probe onto the target surface. In addition, a path-based virtual fixture is proposed to realize easy human-robot interaction for repeat compression operation at the lesion location. To ensure the biometric measurements obtained in different examinations are comparable, the 6D scanning path is determined in a coarse-to-fine manner using both an external RGBD camera and US images. The RGBD camera is first used to extract a rough scanning path on the object. Then, the segmented vascular lumen from US images are used to optimize the scanning path to ensure the visibility of the target object. To generate a continuous scan path for developing virtual fixtures, an arc-length based path fitting model considering both position and orientation is proposed. Finally, the whole system is evaluated on a human-like arm phantom with an uneven surface. The code (https://github.com/dianyeHuang/RobDVTUS) and intuitive demonstration video (https://www.youtube.com/watch?v=3xFyqU1rV8c) can be publicly accessed. Note to Practitioners—Robotic ultrasound (US) systems have attracted attention for various applications in the past decades. However, the existing studies are not mature and intelligent enough for some challenging applications, such as DVT exam, which requires rich contact interaction between patients and clinicians. To tackle with this challenge, this study presents a novel human-centric robotic DVT exam program using the technique of virtual fixture. The coarse-to-fine path planning module ensures the repeatability of US acquisitions carried out at different times. During DVT exam, the proposed continuous 6D path virtual fixture can guide clinicians to freely move the probe along the scan path while limiting the probe motion in other directions. In order to perform the compress-release exam, a decoupled position/force controller is developed to precisely generate the contact force conveyed by clinicians and to restrict the probe motion along the probe centerline. We believe such a robot-assisted system is a promising solution to take both advantages of robots about the accuracy and repeatability and human operators about the advanced physiological knowledge.}, archive = {J_TASE}, author = {Dianye Huang and Chenguang Yang and Mingchuan Zhou and Angelos Karlas and Nassir Navab and Zhongliang Jiang}, doi = {10.1109/TASE.2024.3351076}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {381--392}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Robot-assisted deep venous thrombosis ultrasound examination using virtual fixture}, volume = {22}, year = {2025}, }
• (2025). A control architecture for safe trajectory generation in Human–Robot collaborative settings. TASE, 22, 365--380. (www)

  @article{ , abstract = {This paper introduces a control architecture that enables a robotic system to ensure the safety of human operators entering its workspace. The proposed method utilizes an appropriate metric to measure safety levels and adjusts the robot’s motion to maintain this metric above a minimum threshold. To guarantee safety, the robot scales down and deviates from its intended path. For redundant robots, internal motion is exploited to enhance safety levels further. The approach is incorporated into a Hierarchical Quadratic Programming control framework, allowing the robot to address other control objectives simultaneously, such as handling joint limits. Experimental results with a dual-arm mobile robot developed as part of the EU-funded CANOPIES project demonstrate the effectiveness of the proposed method. Note to Practitioners—This paper was motivated by the problem of ensuring human safety in unstructured environments shared with human operators. We propose a control architecture that allows complex dual-arm robotic systems to operate effectively in such scenarios. The devised architecture gives the robot the capability to slow down a trajectory to follow as well as to deviate from a nominal path to keep a human operator safe. We tested the devised approach in a precision farming setting; however, it can be adopted in any human-robot interaction scenario.}, archive = {J_TASE}, author = {Jozsef Palmieri and Paolo Di Lillo and Martina Lippi and Stefano Chiaverini and Alessandro Marino}, doi = {10.1109/TASE.2024.3350976}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {365--380}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A control architecture for safe trajectory generation in Human–Robot collaborative settings}, volume = {22}, year = {2025}, }
• (2025). Algorithms for cameras view-frame placement problems in the presence of an adversary and distributional ambiguity. TASE, 22, 353--364. (www)

  @article{ , abstract = {In this paper, we introduce cameras view-frame placement problem (denoted by CFP) in the presence of an adversary whose objective is to minimize the maximum coverage by $p$ cameras in response to input provided by $n$ autonomous agents in a remote location. We allow uncertainty in the success of attacks, incomplete information of the probability distribution associated with the uncertain data, and varying levels of risk-appetite of the adversary. We present an exact cutting planes based algorithm to solve this problem and provide conditions under which it is finitely convergent. Since this approach solves deterministic CFP in each iteration, we also present improved exact method for CFP with $p=1$ , approximation algorithm and heuristics for Multi-CFP with $p\geq 2$ , and Multi-CFP with fixed tilt of the cameras. To evaluate the effectiveness and performance of the proposed approaches, we conduct computational experiments using randomly generated instances and simulation experiments where these approaches are utilized to find a hidden object in a remote location. Note to Practitioners—Telerobotic cameras have been widely used for a variety of applications in environment where it is tedious for humans to collect information such as surveillance, natural environment observation, search and rescue, satellite imaging, and many more. Therefore, computationally efficient approaches proposed in this paper for placement of view-frame of camera(s), by adjusting their pan, tilt, and zoom, will improve the effective utilization of a telerobotic cameras system. Additionally, before operating such systems in a military environment, a decision maker needs to analyze vulnerable cameras in the system whose disruption can significantly impact the information acquisition process. The algorithms presented for adversarial camera view-frame placement problem can identify the set of cameras (or vehicles carrying them) that are susceptible to attacks by a reasonable (risk-averse) attacker. Likewise, the proposed algebraic modeling framework and solution approaches are also applicable for planning interdiction actions to minimize the information acquisition by an evader/enemy. These results can be leveraged by autonomy solutions developed by the Army for both logistics (Autonomous Ground Resupply program) and combat missions (Combat Vehicle Robotics program).}, archive = {J_TASE}, author = {Seonghun Park and Manish Bansal}, doi = {10.1109/TASE.2024.3350973}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {353--364}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Algorithms for cameras view-frame placement problems in the presence of an adversary and distributional ambiguity}, volume = {22}, year = {2025}, }
• (2025). A novel hybrid ordinal learning model with health care application. TASE, 22, 339--352. (www)

  @article{ , abstract = {Ordinal learning (OL) is a type of machine learning models with broad utility in health care applications such as diagnosis of different grades of a disease (e.g., mild, modest, severe) and prediction of the speed of disease progression (e.g., very fast, fast, moderate, slow). This paper aims to tackle a situation when precisely labeled samples are limited in the training set due to cost or availability constraints, whereas there could be an abundance of samples with imprecise labels. We focus on imprecise labels that are intervals, i.e., one can know that the a sample belongs to an interval of labels but cannot know which unique label it has. This situation is quite common in health care datasets due to limitations of the diagnostic instrument, sparse clinical visits, or/and patient dropout. Limited research has been done to develop OL models with imprecise/interval labels. We propose a new Hybrid Ordinal Learner (HOL) to integrate samples with both precise and interval labels to train a robust OL model. We also develop a tractable and efficient optimization algorithm to solve the HOL formulation. We compare HOL with several recently developed OL methods on four benchmarking datasets, which demonstrate the superior performance of HOL. Finally, we apply HOL to a real-world dataset for predicting the speed of progressing to Alzheimer’s Disease (AD) for individuals with Mild Cognitive Impairment (MCI) based on a combination of multi-modality neuroimaging and demographic/clinical datasets. HOL achieves high accuracy in the prediction and outperforms existing methods. The capability of accurately predicting the speed of progression to AD for each individual with MCI has the potential for helping facilitate more individually-optimized interventional strategies. Note to Practitioners—Machine learning (ML) algorithms have been widely adopted to support disease diagnosis and prognosis. In some situations, the outcome variable of interest is on an ordinal scale, i.e., it includes several classes with a natural order. For example, the variable of interest can be the grade of a disease as mild, moderate, or severe; or it can be the progression speed of a disease as very fast, fast, moderate, or slow. Ordinal learning (OL) is the type of ML algorithms for ordinal variable prediction. Most existing OL algorithms can only include samples with precise labels in training. However, it is common to have samples with imprecise/interval labels, i.e., we know that a sample belongs to a range of classes/labels but do not know which specific class/label it belongs to. This situation can happen due to a variety of different reasons such as use of less accurate diagnostic instrument under cost or availability constraints, sparse clinical assessment, and patient dropout. We propose a Hybrid Ordinal Learner (HOL) to integrate samples with both precise and interval labels to train a robust OL model. HOL is evaluated using four public benchmarking datasets and shows superior performance compared to existing methods. Also, we apply HOL to a real-world dataset for predicting the speed of progressing to Alzheimer’s Disease (AD) for individuals with Mild Cognitive Impairment (MCI). MCI is the prodromal stage of AD. Individuals with MCI show noticeable signs of memory loss and cognitive declines, but these symptoms are not severe enough to interfere their independent living. HOL achieves high accuracy in predicting the speed of progressing to AD for each MCI subject (e.g., the speed of ‘very fast’, ‘fast’, ‘moderate’, or ‘slow), which could potentially help facilitate the development of more individually-optimized interventional strategies.}, archive = {J_TASE}, author = {Lujia Wang and Hairong Wang and Yi Su and Fleming Lure and Jing Li}, doi = {10.1109/TASE.2024.3350894}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {339--352}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A novel hybrid ordinal learning model with health care application}, volume = {22}, year = {2025}, }
• (2025). Distributed adaptive asymptotic consensus tracking control for stochastic nonlinear MASs with unknown control gains and output constraints. TASE, 22, 328--338. (www)

  @article{ , abstract = {This paper studies the asymptotic consensus tracking control problem for a class of stochastic nonlinear multiagent systems (MASs) with output constraints and unknown control gains. Firstly, the Nussbaum technique is introduced to solve the difficulty of the unknown control gains in the stochastic nonlinear MASs. Meanwhile, a $tan$ -type nonlinear mapping (NM) function is used to ensure that the output of each agent satisfies the predefined output constraints. Furthermore, the “explosion of complexity” problem caused by the traditional backstepping design methods is handled by using the command filter technique. The developed distributed adaptive asymptotic consensus tracking control strategy ensures that all the signals in the closed-loop system are bounded in probability and the consensus tracking errors of all agents converge to zero in probability. Finally, a simulation example proves the effectiveness of the proposed control strategy. Note to Practitioners—In this paper, the asymptotic consensus tracking control problem is studied for a class of the stochastic nonlinear MASs. In nature, there are many meaningful movements of multi-agents with stochastic disturbances. It is particularly challenging to achieve the asymptotic consensus tracking control problem for stochastic nonlinear MASs, which involves the unknown control gains and output constraints. Therefore, the Nussbaum technique is used to solve the difficulty of the unknown control gains, meanwhile the command filter technique is introduce to solve the “explosion of complexity” problem in the backstepping design process. Moreover, the designed control strategy and stability analysis for the studied system is based on the nonlinear mapping technique and Lyapunov method, which makes the developed methodology more engineering-oriented.}, archive = {J_TASE}, author = {Yan Zhu and Ben Niu and Zihao Shang and Zhenhua Wang and Huanqing Wang}, doi = {10.1109/TASE.2024.3350547}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {328--338}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed adaptive asymptotic consensus tracking control for stochastic nonlinear MASs with unknown control gains and output constraints}, volume = {22}, year = {2025}, }
• (2025). Design and quantitative assessment of teleoperation-based Human–Robot collaboration method for robot-assisted sonography. TASE, 22, 317--327. (www)

  @article{ , abstract = {Tele-echography has emerged as a promising and effective solution, leveraging the expertise of sonographers and the autonomy of robots to perform ultrasound scanning for patients residing in remote areas, without the need for in-person visits by the sonographer. Designing effective and natural human-robot interfaces for tele-echography remains challenging, with patient safety being a critical concern. In this article, we develop a teleoperation system for robot-assisted sonography with two different interfaces, a haptic device-based interface and a low-cost 3D Mouse-based interface, which can achieve continuous and intuitive telemanipulation by a leader device with a small workspace. To achieve compliant interaction with patients, we design impedance controllers in Cartesian space to track the desired position and orientation for these two teleoperation interfaces. We also propose comprehensive evaluation metrics of robot-assisted sonography, including subjective and objective evaluation, to evaluate tele-echography interfaces and control performance. We evaluate the ergonomic performance based on the estimated muscle fatigue and the acquired ultrasound image quality. We conduct user studies based on the NASA Task Load Index to evaluate the performance of these two human-robot interfaces. The tracking performance and the quantitative comparison of these two teleoperation interfaces are conducted by the Franka Emika Panda robot. The results and findings provide guidance on human-robot collaboration design and implementation for robot-assisted sonography. Note to Practitioners—Robot-assisted sonography has demonstrated efficacy in medical diagnosis during clinical trials. However, deploying fully autonomous robots for ultrasound scanning remains challenging due to various constraints in practice, such as patient safety, dynamic tasks, and environmental uncertainties. Semi-autonomous or teleoperation-based robot sonography represents a promising approach for practical deployment. Previous work has produced various expensive teleoperation interfaces but lacks user studies to guide teleoperation interface selection. In this article, we present two typical teleoperation interfaces and implement a continuous and intuitive teleoperation control system. We also propose a comprehensive evaluation metric for assessing their performance. Our findings show that the haptic device outperforms the 3D Mouse, based on operators’ feedback and acquired image quality. However, the haptic device requires more learning time and effort in the training stage. Furthermore, the developed teleoperation system offers a solution for shared control and human-robot skill transfer. Our results provide valuable guidance for designing and implementing human-robot interfaces for robot-assisted sonography in practice.}, archive = {J_TASE}, author = {Weiyong Si and Ning Wang and Chenguang Yang}, doi = {10.1109/TASE.2024.3350524}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {317--327}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Design and quantitative assessment of teleoperation-based Human–Robot collaboration method for robot-assisted sonography}, volume = {22}, year = {2025}, }
• (2025). Accurate and automatic dental crown components segmentation with multi-scale attention based U-net and hybrid level set models. TASE, 22, 305--316. (www)

  @article{ , abstract = {This paper presents a two-step method to automatically and accurately segment the dental crown components from CT images. Firstly, a multi-scale attention based U-Net model is proposed for pulp segmentation, which is embedded with global and local attention modules. The constructed attention modules can automatically aggregate pixel-wise contextual information and focus on catching the real dental pulp region. Secondly, two efficient level set models are proposed: one is the shape constraint-based level set model for enamel and dentin segmentation, the other is the region mutual exclusion-based level set model for neighboring teeth segmentation. The proposed shape constraint term can better handle topology changes of teeth and the region mutual exclusion term can more effectively avoid intersecting segmentation. Besides, a starting slice initialization method is introduced to achieve automatic segmentation, and an accurate contour propagation strategy is developed for slice-by-slice segmentation. We set up a series of comparative experiments for evaluation. Experimental results verify that the proposed method obtains promising performance for each crown component segmentation, and outperforms state-of-the-art tooth segmentation methods in terms of accuracy. This suggests that the proposed method can be used to accurately segment the crown components for precise tooth preparation treatment. Note to Practitioners—The motivation of this work is to reduce the burden on dentists during tooth preparation treatment, which requires accurate segmentation of crown components (i.e., enamel, dentin, and pulp) from dental CT images. Existing methods only focused on the segmentation of teeth or alveolar bone. Therefore, we present a novel automatic segmentation model for the dental crown components with high accuracy. A key strength of this study is the combination of a data-driven method (deep learning) and model-driven methods (level-set), which can provide good accuracy under limited training samples. This ability is highly desirable for practitioners by saving labor-intensive, costly labeling efforts. Furthermore, our proposed method will provide tools to help reduce subjectivity and human errors, as well as streamline and expedite the clinical workflow. This will significantly facilitate tooth preparation automation.}, archive = {J_TASE}, author = {Dongyue Li and Mingzhu Zhu and Shaoan Wang and Yaoqing Hu and Fusong Yuan and Junzhi Yu}, doi = {10.1109/TASE.2024.3350088}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {305--316}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Accurate and automatic dental crown components segmentation with multi-scale attention based U-net and hybrid level set models}, volume = {22}, year = {2025}, }
• (2025). A hankel matrix-based multivariate control chart with shrinkage estimator for condition monitoring of rolling bearings. TASE, 22, 293--304. (www)

  @article{ , abstract = {Rolling element bearings are crucial key components in rotating machinery, and it is essential to monitor their condition to prevent unexpected breakdown or safety incidents. However, features that indicate the operating condition are often affected by strong background noise, leading to inaccurate detection. Additionally, classical monitoring methods often rely on a multivariate normal distributed (MND) assumption, which may not be suitable in real-world applications due to noise interference, outliers, and sampling errors. To address the aforementioned issues, this paper proposes a novel Hankel matrix-based multivariate homogeneously weighted moving average control chart with shrinkage estimator (HMHS chart). The first step involves segmenting the bearing vibration signal to create Hankel matrix sequences, and the singular features are obtained via singular value decomposition (SVD). To determine the optimal singular sequence, a criterion called mode entropy (ME) is proposed. Secondly, a shrinkage estimator is applied to establish a robust statistic with low covariance bias. Thirdly, a health index is constructed using the HMHS chart to track the bearing’s degradation process. Simulation studies and two real case studies are performed with comparison to classical charts to demonstrate the effectiveness and accuracy of the proposed method. The results demonstrate the effectiveness and accuracy of the proposed method in the condition monitoring of rolling bearings. Note to Practitioners—A novel condition monitoring method of rolling bearings is proposed in this work. Existing methods often rely on traditional statistics (such as kurtosis, root mean square), which can be easily disturbed by noise and outliers in training samples. Moreover, the MND assumption needed is not always tolerated in real applications, leading to poor performance in condition monitoring. To solve this problem, a Hankel matrix-based multivariate control chart is proposed in this paper, which includes several steps for implementation. These steps include: 1) collecting vibration signals of rolling bearings; 2) conducting Hankel matrix-based adaptive feature extraction; 3) constructing the monitoring statistic with robust estimators, and estimating the control limit; 4) conducting online health monitoring of rolling bearings. Extensive simulation studies and two real case studies are conducted to evaluate the accuracy and effectiveness of the proposed method in bearing condition monitoring.}, archive = {J_TASE}, author = {Wei Fan and Fan Jiang and Yongxiang Li and Zhike Peng}, doi = {10.1109/TASE.2024.3350032}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {293--304}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A hankel matrix-based multivariate control chart with shrinkage estimator for condition monitoring of rolling bearings}, volume = {22}, year = {2025}, }
• (2025). Disturbance-compensation-based predictive sliding mode control for aero-engine networked systems with multiple uncertainties. TASE, 22, 276--292. (www)

  @article{ , abstract = {This paper investigates the compound control problem of aero-engine networked systems with multiple uncertainties and saturation constraints. To address these challenges, a novel sliding mode controller (SMC) with extended state observer (ESO) is first designed to implement anti-disturbance control, which can alleviate the chattering phenomenon without sacrificing robustness by two parallel ways: adaptive switching term and disturbance compensation. Then, the predictive control strategy is introduced to further optimize the reaching phase, and two disturbance-compensation-based predictive SMC schemes are proposed to coordinate diverse control requirements in the presence and absence of system constraints. When there are no saturation limits, the optimization problem is formulated as a convex one, and its analytic solution is derived explicitly. Furthermore, both state and control limits are considered in the system synthesis, and a standard prediction problem with multiple constraints is established. For high-frequency sampling needs, the Laguerre function is designed to reconstruct the input variables of the prediction sequence, which can effectively reduce the calculation complexity without compromising the dynamic performance. The experiment simulations show that the proposed compound schemes have strong robustness to multiple uncertainties and saturation constraints, and achieve promising control performance in both transient and steady-state phases. Note to Practitioners—As the heart of aircraft, aero-engine system is developing towards networked and intelligent. Subsequently, some new challenges are exposed to be addressed, such as communication delays, saturation constraints, electromagnetic disturbances, etc., which make the existing linear schemes fail to guarantee diverse requirements. Although SMC owns good robustness against multiple uncertainties, most relevant results still face some challenges, such as high switching gain, known disturbance boundary assumption, and sufficient control period. To address these difficulties, we propose two disturbance-compensation-based predictive SMC schemes to coordinate diverse control requirements with and without system constraints. The proposed schemes do not rely on excessive computing resources and are applicable to high-frequency sampling systems. The results of this paper can also provide guidance for the design of robust compound strategies for other networked systems with multiple uncertainties.}, archive = {J_TASE}, author = {Pengtao Song and Qingyu Yang and Donghe Li and Guangrui Wen and Zhifen Zhang and Jingbo Peng}, doi = {10.1109/TASE.2024.3350020}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {276--292}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Disturbance-compensation-based predictive sliding mode control for aero-engine networked systems with multiple uncertainties}, volume = {22}, year = {2025}, }
• (2025). Event-based adaptive sliding-mode containment control for multiple networked mechanical systems with parameter uncertainties. TASE, 22, 264--275. (www)

  @article{ , abstract = {The issue related to distributed containment control for multiple networked mechanical systems with inherent nonlinearities, dynamic leaders, unknown external disturbances, parameter uncertainties, and constrained network communication is investigated by designing distributed adaptive event-triggered sliding-mode controllers in this study. To lessen the number of state updates and network resource loss of networked mechanical systems, a time-varying-threshold-based adaptive event-triggered mechanism is constructed. An adaptive sliding-mode estimator is established to estimate the inaccurate states. Then, integrated with the aforementioned event-triggered strategy and adaptive sliding-mode estimator, discontinuous and continuous distributed adaptive event-triggered sliding-mode control laws without requiring each follower to get the upper bounds of the leaders’ states derivatives are, respectively, devised to compensate for the influences of nonlinearities, disturbances, and parameter uncertainties. To further attenuate the negative effects of unknown disturbances, inherent nonlinearities, and chattering, a distributed adaptive event-triggered sliding-mode control protocol with boundary layer function is designed. Eventually, the Lyapunov stability theory is utilized to testify that the adaptive error and containment error are uniformly ultimately bounded. A practical example is furnished to verify the validity of the present sliding-mode containment control strategies. Note to Practitioners—This work aims to develop the distributed containment control approach for multiple networked nonlinear mechanical systems, which is of great significance in the fields of deep space exploration, environment monitoring and joint rescue. We put forward an event-based adaptive estimation and containment control framework for multiple networked nonlinear mechanical systems, which solves practical problems such as transmission frequency, limited computation capability, and network resources. An adaptive sliding-mode estimator and adaptive event-triggered mechanism are, respectively, devised to estimate the inaccurate states and reduce data transmission frequency. Despite the effects of communication interruption and unknown disturbances, an event-triggered adaptive control protocol based on sliding-mode estimator is designed, which is applied to a planar manipulator with two degree-of-freedom.}, archive = {J_TASE}, author = {Deyin Yao and Yuyang Wu and Hongru Ren and Hongyi Li and Yang Shi}, doi = {10.1109/TASE.2024.3349634}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {264--275}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Event-based adaptive sliding-mode containment control for multiple networked mechanical systems with parameter uncertainties}, volume = {22}, year = {2025}, }
• (2025). Data-driven H−/H∞ fault detection and control in finite-frequency domain with application to steel rolling process. TASE, 22, 252--263. (www)

  @article{ , abstract = {This study investigates the problem of data-driven fault detection and control (FDC) with the help of $H_{-}$ and $H_{\infty} $ performance indices for systems with unknown model parameters. To guarantee disturbance attenuation performance and fault sensitivity performance under data-driven framework, two data sets will be excited from disturbances and faults respectively. However, different data sets make it hard to solve a common control gain which satisfies the above two performance constraints simultaneously. Regarding this, a new data-driven parametrization method is developed by introducing a separation technology, which achieves the separation of excited data from the analytical solution of controller gain. Then, the joint design of controller and detector is realized by solving the $H_{-}/H_{\infty} $ optimization problem. Particularly, in terms of single-objective $H_{-}$ and $H_{\infty} $ optimization problems in finite frequency domain, sufficient and necessary controller design conditions are formulated. Finally, the proposed method is applied to the FDC problem of hot strip rolling and cold strip rolling whose model parameters are hard to obtain, and the effectiveness is verified by simulation experiments. Note to Practitioners—The determination of model parameters in both hot strip and cold strip rolling systems is restricted by the forward slip equation. However, it may be difficult to determine the forward slip equation accurately due to complex coupling effects, which makes the model-based control and detection problems challenging. In view of this, an improved data-driven method combining with $H_{-}$ and $H_{\infty} $ optimization technologies is proposed in this paper, which can be used to solve the FDC problem of industrial systems with unknown model parameters. The proposed data-driven method overcomes the difficulty of using two different data sets excited from disturbances and faults, respectively, to design a unique controller that satisfies both $H_{-}$ and $H_{\infty} $ performances. Particularly, faults and disturbances in practical industrial systems tend to occur in finite frequency ranges, which inspires us to further develop the finite frequency data-driven FDC technology. The effectiveness of the results in the simulation experiments is validated for both cold strip and hot strip steel systems using data generated based on the simulation model as well as field rolling data.}, archive = {J_TASE}, author = {He Liu and Xiao-Jian Li}, doi = {10.1109/TASE.2024.3349387}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {252--263}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Data-driven H−/H∞ fault detection and control in finite-frequency domain with application to steel rolling process}, volume = {22}, year = {2025}, }
• (2025). Virtual cell-based scheduling approach to single-robotic-arm cluster tools subject to wafer residency time constraints. TASE, 22, 240--251. (www)

  @article{ , abstract = {Scheduling single-robotic-arm cluster tools subject to wafer residency time constraints has received much attention. Compared to some scheduling strategies that use all processing modules (PMs) to process wafers, it is much more challenging to schedule a more general case whose optimal scheduling strategy is not limited to the case of using all PMs. The strategy of only adjusting the robot’s waiting time may fail to produce a desired schedule. When a tool using all PMs is not schedulable, it may become schedulable if only some PMs of a type are used. Therefore, it is very important to select an appropriate number of PMs to process wafers. This work studies the cyclic scheduling problem of wafer-residency-time-constrained single-robotic-arm cluster tools by simultaneously adjusting the number of PMs and robot waiting time. We build a virtual cell that includes an appropriate number of PMs to process wafers with the maximal productivity. We establish sufficient and necessary conditions under which the system is schedulable. The schedulability conditions are less conservative than the state-of-the-art one. A polynomial algorithm is developed to find the optimal cyclic schedule, a virtual cell’s configuration, and robot waiting time. We illustrate the practicability of the proposed algorithm via several examples, and its superiority over the existing one. Note to Practitioners—This paper addresses the optimal cyclic scheduling problem of wafer-residency-time-constrained single-robotic-arm cluster tools that are used in every wafer fabrication factory. This work for the first time simultaneously adjusts the number of PMs and the robot waiting time given such a cluster tool such that it can be scheduled with the highest productivity. It presents the optimal scheduling method with polynomial complexity. We confirm that the proposed approach can improve schedulability of single-robotic-arm cluster tools over existing methods by using multiple examples. It can thus be readily applied to industrial wafer fabrication.}, archive = {J_TASE}, author = {Jufeng Wang and Chunfeng Liu and MengChu Zhou and Abdullah Abusorrah}, doi = {10.1109/TASE.2024.3349380}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {240--251}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Virtual cell-based scheduling approach to single-robotic-arm cluster tools subject to wafer residency time constraints}, volume = {22}, year = {2025}, }
• (2025). Process model-assisted optimization control method based on model predictive control and deep reinforcement learning for coordinated control system in coal-fired power unit. TASE, 22, 228--239. (www)

  @article{ , abstract = {As the large-scale integration of renewable energy into the power grid, coal-fired power units have to undertake the task of load-frequency regulation, the optimization control of the coal-fired power units is becoming increasingly important. Reinforcement learning is being used for optimization control of the process increasingly but may encounter high learning costs. In this paper, a process model-assisted optimization control framework called PMA-OC is proposed to achieve the optimization control of coal-fired power unit. In this framework, a low-precision process model developed by process operation knowledge is employed to achieve the initial control based on model predictive control algorithm (MPC) and improve the dynamic characteristics of the coordinated control system. Then the deep reinforcement learning algorithm (DRL) is utilized to improve the initial control continuously, through constantly interacting with the improved process. With the full mining of process knowledge from the power unit, the improved process offers the better training samples for DRL, reduces the learning complexity of DRL, and strengthens the learning efficiency of DRL. Finally, the proposed control framework is employed to achieve the optimization control of the coordinated control system, and simulation results indicate that the proposed control framework outperforms the existing standalone MPC and DRL methods in terms of anti-interference ability, dynamic response time, and performs well control capability. Note to Practitioners—The flexible and stable operation of coal-fired power units is of great significance to the peak load regulation and supply guarantee of the power grid. For practitioners of the coal-fired power unit, this paper proposes a process model-assisted optimization control framework to promote the process knowledge utilization. In this framework, process knowledge is used to implement preliminary control on coal-fired power units to reduce the complex dynamic characteristics of coal-fired power units. Then, the reinforcement learning algorithm is used to continuously improve the control performance of coal-fired power units by interacting with the improved process. Finally, the effectiveness of the proposed method is verified by a supercritical unit. Although the proposed method is applied to the optimal control of coal-fired power units, its ideas can be used for the reference in other complex industrial processes.}, archive = {J_TASE}, author = {Bangwu Dai and Fuli Wang and Yuqing Chang and Fei Chu and Shengjun Song}, doi = {10.1109/TASE.2023.3349196}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {228--239}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Process model-assisted optimization control method based on model predictive control and deep reinforcement learning for coordinated control system in coal-fired power unit}, volume = {22}, year = {2025}, }
• (2025). Scheduling a distributed permutation flowshop with uniform machines and release dates. TASE, 22, 215--227. (www)

  @article{ , abstract = {Global manufacturing optimizes production efficiency and cost, enabling enterprises to compete effectively. Distributed production is emerging for the trend of globalization and the requirements of diversified manufacturing, for which distributed scheduling plays an important role in enterprises enhancing efficiency and conserving energy. This study investigates a heterogeneous distributed permutation flowshop scheduling problem to minimize the makespan, in which release date and factory speed are incorporated to mirror a real-world production scenario. A mixed integer programming model is established to address this NP-complete problem using a business optimizer. The findings assist in identifying optimal properties for algorithm development and assessing the performance of a proposed branch and bound algorithm, which includes a problem-specific pruning rule and defined lower and upper bounds to reduce the solution space. For industrial scheduling, a dispatching rule named Dynamic Largest Processing Volume on the Fastest Machine First is proposed, offering asymptotic optimality and ensuring uninterrupted production. Additionally, a discrete artificial bee colony algorithm with a novel availability rule and an abandonment criterion is introduced to improve efficiency. Simulation results demonstrate the efficacy of the proposed algorithms. Note to Practitioners—Distributed production leverages information technology to integrate geographically decentralized manufacturing units for cost reduction and production enhancement. This mode finds extensive application in automobile, electronics, and medical equipment manufacturing. In mass production scenarios, the online Dynamic Largest Processing Volume on the Fastest Machine First rule provides a viable alternative to optimal algorithms, delivering a convergent schedule rapidly and preventing unnecessary delays. In complex industrial settings, the discrete artificial bee colony algorithm offers effective solutions without relying on a mathematical model, proving more efficient in such environments.}, archive = {J_TASE}, author = {Danyu Bai and Tianyi Liu and Yuchen Zhang and Feng Chu and Hu Qin and Liang Gao and Yue Su and Mingjie Huang}, doi = {10.1109/TASE.2023.3349167}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {215--227}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Scheduling a distributed permutation flowshop with uniform machines and release dates}, volume = {22}, year = {2025}, }
• (2025). Dynamic-memory event-triggered sliding-mode secure control for nonlinear semi-markov jump systems with stochastic cyber attacks. TASE, 22, 202--214. (www)

  @article{ , abstract = {This paper presents an investigation of the sliding-mode secure control for nonlinear semi-Markov jump systems in the presence of non-periodic denial-of-service attacks and false data injection attacks. First of all, we employ Takagi-Sugeno fuzzy model to describe the nonlinear semi-Markov jump control systems. In order to optimize transmission efficiency and control performance, a novel dynamic-memory event-triggered mechanism is developed by incorporating auxiliary dynamic variable and historical transmitted data. Then, a memory-based fuzzy sliding surface is put forward to attenuate the influences of stochastic cyber attacks with the aid of event-triggered state information. Moreover, by utilizing Lyapunov stability theory, sufficient conditions are derived to guarantee the exponentially mean-square stability of the system with an $H_{\infty }$ performance index, even in the cases of generally uncertain and unknown transition rates. Furthermore, a memory-based sliding mode secure controller is designed to ensure the reachability of the predefined switching surface and desirable sliding motion within finite time. Finally, the efficacy of the proposed control scheme is demonstrated through a tunnel diode circuit model. Note to Practitioners—This study focuses on the issue of secure control for nonlinear semi-Markov jump systems, which holds practical significance across various domains, including applications in robotic manipulators, circuit models, and DC motors. We broaden the scope by considering more general jump parameter matrices to align more closely with the real-world system environment. Moreover, networked environments pose two primary challenges: network bandwidth constraints and the external network attacks. To tackle these issues, this paper introduces an innovative dynamic memory event triggering mechanism to enhance network transmission efficiency and optimize communication resource utilization. Meanwhile, to address cyber attacks resulting from the inherent openness of networks, the current study adopts a defensive sliding mode control strategy to provide robust protection against network attacks and disturbances. The effectiveness of the suggested approach is confirmed through a circuit system model. It is worth mentioning that the proposed method has the potential for broader application within real-world engineering scenarios, particularly those involving network-based nonlinear systems with various practical constraints.}, archive = {J_TASE}, author = {Yushun Tan and Jiajing Liu and Xiangpeng Xie and Engang Tian and Jinliang Liu}, doi = {10.1109/TASE.2023.3349150}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {202--214}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Dynamic-memory event-triggered sliding-mode secure control for nonlinear semi-markov jump systems with stochastic cyber attacks}, volume = {22}, year = {2025}, }
• (2025). Prescribed-time optimal control of nonlinear dynamical systems with application to a coupled tank system. TASE, 22, 191--201. (www)

  @article{ , abstract = {This article presents a solution to the problem of achieving optimal prescribed-time stability and stabilization for nonlinear dynamical systems. In contrast to existing prescribed-time control methods, this article initiates by establishing sufficient conditions for prescribed-time stability through the use of continuous Lyapunov candidate functions. Building upon these conditions, we introduce an optimal prescribed-time stabilization method that incorporates specific differential inequalities. This method complies with the Hamilton-Jacobi-Bellman steady-state equation, ensuring both optimality and prescribed-time stability. Furthermore, we derive a set of optimal prescribed-time stabilizing control laws for a class of affine nonlinear dynamical systems. Finally, we demonstrate the effectiveness of the proposed approach through simulations and experiments involving the reference level tracking of a coupled tank system, thus ensuring that the tracking performance aligns with practical user specifications. Note to Practitioners—This article was instigated by the challenge of devising optimal feedback control strategies for a specific class of nonlinear dynamical systems at predetermined time instances. In recent years, there has been a growing interest in prescribed time stability and stabilization approaches, driven by their potential applications across diverse fields, including control engineering, robotics, and aerospace engineering. These methods facilitate the regulation of nonlinear dynamical systems to reach a desired steady state within a predefined finite time, offering a valuable solution for situations demanding rapid stabilization. In this article, we introduce a novel optimal prescribed-time stabilization method that relies on specific differential inequalities. This method not only adheres to the Hamilton-Jacobi-Bellman steady-state equation but also guarantees both optimality and prescribed-time stability. Furthermore, we derive a family of optimal prescribed-time stabilizing control laws tailored to a particular class of affine nonlinear dynamical systems. To validate the effectiveness of our proposed stabilization approach, we conduct experiments focusing on tracking the desired water level within a coupled tank system. Ultimately, the presented prescribed-time optimal feedback control strategy marks a significant stride forward in the advancement of optimal and efficient control methods for nonlinear dynamical systems, offering solutions that hold immense promise in practical applications.}, archive = {J_TASE}, author = {Vijay Kumar Singh and Shyam Kamal and Bijnan Bandyopadhyay and Sandip Ghosh and Thach Ngoc Dinh}, doi = {10.1109/TASE.2023.3349068}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {191--201}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed-time optimal control of nonlinear dynamical systems with application to a coupled tank system}, volume = {22}, year = {2025}, }
• (2025). Prescribed performance adaptive containment control for full-state constrained nonlinear multiagent systems: A disturbance observer-based design strategy. TASE, 22, 179--190. (www)

  @article{ , abstract = {This paper focuses on the prescribed performance adaptive containment control problem for a class of nonlinear nonstrict-feedback multiagent systems (MASs) with unknown disturbances and full-state constraints. First, the radial basis function neural networks (RBF NNs) technology is employed to approximate the unknown nonlinear functions in the system, and the problem of “explosion of complexity” caused by repeated derivation of virtual controls is solved by using the dynamic surface control (DSC) technology. Then, the nonlinear disturbance observers are designed to estimate the external disturbance, and the barrier Lyapunov functions (BLFs) and the prescribed performance function (PPF) are combined to achieve the control objective of prescribed performance without violating the full-state constraints. The theoretical result shows that all signals in the closed-loop system are semiglobally uniformly ultimately bounded (SGUUB), and the local neighborhood containment errors can converge to the specified boundary. Finally, two simulation examples show the effectiveness of the proposed method. Note to Practitioners—The containment control problem is a hot topic in the field of control, which plays an important role in practical engineering. Especially for this problem of nonlinear MASs, the mathematical models are difficult to be obtained accurately. This paper investigates the prescribed performance adaptive containment control problem for the nonlinear nonstrict-feedback MASs, whose model can be extended to more complex engineering applications, such as unmanned aerial vehicle formations and intelligent traffic management. It is worth noting that external disturbances and state constraint problems often exist in practical applications. Therefore, the disturance observers are designed to compensate for the system disturbances, which can eliminate the impacts of disturbances on the systems. By introducing BLFs, it is ensured that all states of the system are constrained within the specified regions. To sum up, the paper proposes a prescribed performance adaptive containment control strategy, which contributes to the development of containment control for MASs in practical applications.}, archive = {J_TASE}, author = {Jihang Sui and Chao Liu and Ben Niu and Xudong Zhao and Ding Wang and Bocheng Yan}, doi = {10.1109/TASE.2023.3348978}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {179--190}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Prescribed performance adaptive containment control for full-state constrained nonlinear multiagent systems: A disturbance observer-based design strategy}, volume = {22}, year = {2025}, }
• (2025). Observer-based event-triggered bipartite consensus for nonlinear multi-agent systems: Asymmetric full-state constraints. TASE, 22, 168--178. (www)

  @article{ , abstract = {This paper addresses the challenge of adaptive fuzzy event-triggered bipartite consensus control for a class of nonlinear multi-agent systems (MASs) with full-state asymmetric constraints, incorporating both cooperative and adversarial communication between agents. To ensure consensus tracking performance in the presence of asymmetric full-state constraints, the nonlinear MASs are augmented with an enhanced nonlinear function. Fuzzy logic systems (FLSs) are employed to handle unfamiliar nonlinearities effectively. To reduce the communication burden and enhance transient performance, an event-triggered control strategy is introduced, accompanied by the construction of an observer utilizing triggered output signals to assess unmeasured states. Furthermore, the control scheme is formulated by integrating dynamic surface control (DSC) and adaptive backstepping techniques. The validity of the proposed approach is verified through Lyapunov’s theory, demonstrating that all signals within the closed-loop system are bounded. Simulation results further support the effectiveness of the proposed methodology. Note to Practitioners—With the rapid development and deep integration of intelligent control theory, the control design of MASs is still challenging in many engineering problems. However, in practical applications, there is not only cooperation but also competition between agents, and the system states are unmeasured and subject to various forms of constraints. To solve this problem, we introduce an improved nonlinear transformation function. When network resources are limited, continuous communication is sometimes not feasible in engineering applications. To minimize the communication burden, we propose an event-triggered control strategy, which combines dynamic surface control technology and adaptive backstepping technology to design the trigger controller. The proposed scheme can effectively save resources and deal with various constraints of system states and may be applied to other similar engineering fields.}, archive = {J_TASE}, author = {Shuai Sui and Dongyu Shen and Wenshan Bi and Shaocheng Tong and C. L. Philip Chen}, doi = {10.1109/TASE.2023.3348236}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {168--178}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Observer-based event-triggered bipartite consensus for nonlinear multi-agent systems: Asymmetric full-state constraints}, volume = {22}, year = {2025}, }
• (2025). Multiplayer stackelberg game-based intelligent frequency control of power system with line loss uncertainty. TASE, 22, 154--167. (www)

  @article{ , abstract = {The power fluctuation and system inertia degradation in a short time put forward higher requirements on the rapidity and robustness of frequency control. This paper proposes an online multiplayer Stackelberg game control framework to achieve frequency control of multi-area power system with the uncertainties of load, renewable energy and line loss. Firstly, an optimal frequency control model incorporating efficient management of load aggregators (LA) is designed based on Stackelberg game, where the LA acts as the leader and all micro-turbines act as the followers. The two-level optimization problem of leader and followers is converted into solving the coupled Hamilton-Jacobi equations with the constraints of the follower’s costate equation. Then, an improved integral reinforcement learning is designed to enhance the wide adaptability of the control strategy online by introducing a general function with uncertainty upper bounds and line losses for all players, while satisfying the uniform ultimate bounded stability of the closed-loop system. A single critic neural network structure is utilized to obtain the optimal strategy. And the convergence of neural network weights is proven. Last, comparative simulation results verify the effectiveness of the proposed method. Note to Practitioners—High accuracy and robustness in frequency control of power systems is important for system safety. Demand-side participation in the interactive regulation of the system makes frequency control strategies more flexible and diversified. However, unreasonable control allocation schemes can incur large payment costs, while changes in control can cause dynamic changes in line losses to affect the supply-demand balance, leading to frequency deviations. This paper proposes an intelligent frequency control method based on the Stackelberg game. The designed IRL algorithm is able to solve the trade-off between system control performance and payment costs online without the priori knowledge of the system and with wide adaptability to uncertainty. Simulation results show that the proposed method outperforms non-game strategies and is robust.}, archive = {J_TASE}, author = {Xin Tong and Dazhong Ma and Rui Wang and Qiuye Sun and Huaguang Zhang}, doi = {10.1109/TASE.2023.3347754}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {154--167}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Multiplayer stackelberg game-based intelligent frequency control of power system with line loss uncertainty}, volume = {22}, year = {2025}, }
• (2025). Stealthy false data injection attacks detection and classification in cyber-physical systems using deep reinforcement learning. TASE, 22, 141--153. (www)

  @article{ , abstract = {Cyber-physical systems (CPSs) are increasingly threatened by stealthy false data injection (SFDI) attacks, which compromise system integrity by manipulating control signals and introducing false sensor data. These attacks are particularly challenging due to their diversity and often indistinguishable nature. In response to this issue, our work uncovers the fundamental causes behind SFDI attacks in linear time-invariant (LTI) systems and elucidates the principles enabling their stealth. We present a novel virtual extended system framework designed to eliminate strictly stealthy attacks within the entire CPS. Utilizing deep reinforcement learning (DRL) methodologies, we pioneer the use of detection results for real-time SFDI attack classification. Through numerical simulations, we validate our proposed method’s effectiveness, demonstrating a classification accuracy of no less than 95%. Notably, even in scenarios where attackers manage to breach the framework partially, our method continues to provide a reliable success rate in SFDI attack detection and classification, showcasing its robustness and efficacy. Note to Practitioners—Cyber-physical systems (CPSs), a critical component of modern industries, are becoming increasingly susceptible to stealthy false data injection (SFDI) attacks. These attacks compromise system integrity by subtly manipulating control signals and feeding false sensor data, making them challenging to detect. Our research presents an innovative framework that uses deep reinforcement learning techniques to detect and classify these elusive attacks, achieving a classification accuracy of over 95%. The information on SFDI attack categories, ascertained by this method, lays the groundwork for the development of subsequent defence strategies. For professionals working in sectors reliant on CPSs, such as manufacturing, healthcare, and transportation, this framework offers a promising tool to enhance system security. Even in scenarios where the system has been partially compromised, our method continues to provide reliable detection and classification, underscoring its robustness and practical utility. The system remains effective despite full breach attempts on specific attack types, ensuring resilience against a broad range of SFDI attacks. In conclusion, our research offers a substantial advancement in protecting CPSs against cyber threats.}, archive = {J_TASE}, author = {Liang Xin and Guang He and Zhiqiang Long}, doi = {10.1109/TASE.2023.3347538}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {141--153}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Stealthy false data injection attacks detection and classification in cyber-physical systems using deep reinforcement learning}, volume = {22}, year = {2025}, }
• (2025). Socially intelligent reinforcement learning for optimal automated vehicle control in traffic scenarios. TASE, 22, 129--140. (www)

  @article{ , abstract = {In this paper, a novel approach is presented for modeling the interaction dynamics between an ego car and a bicycle in a traffic scenario using a hybrid reinforcement learning framework combined with a social value orientation (SVO) model. The proposed framework leverages the SARSA algorithm to learn the optimal policy for the ego vehicle while incorporating risk cost as the negative log-likelihood of collision. Additionally, a customized SVO model is introduced to capture the social preferences of the ego car and the bicycle, defining the SVO of each agent as a continuous variable between egoistic and cooperative orientations. Furthermore, a weight parameter is incorporated in the framework to regulate the influence of the SVO model on the learning process. We demonstrate the effectiveness of our approach through extensive simulations, showing that the ego car can balance between maximizing its reward and avoiding collisions while considering the social preferences of the agents. The obtained results are compared to other models in the literature, and it is shown that the proposed method contributes to the development of safe and efficient autonomous driving systems that interact with human-driven vehicles in a socially intelligent manner. Note to Practitioners—This proposed framework is motivated by the pressing challenge of navigation for autonomous cars in complex urban driving scenarios and mixed traffic situations. With the increasing prevalence of autonomous vehicles on roads, developing intelligent navigation systems that can effectively interact with other road users has become essential. Our novel framework addresses this need by leveraging the SARSA algorithm to learn the optimal policy for the ego vehicle while incorporating risk cost as the negative log-likelihood of collision. Additionally, a customized SVO model is introduced to capture the social preferences of the ego car and the bicycle, defining the SVO of each agent as a continuous variable between egoistic and cooperative orientations. This enables autonomous vehicles to make informed decisions and navigate safely and efficiently. Our framework can enormously help the field of autonomous vehicle navigation and contribute significantly to developing safe, human-centric, and reliable transportation systems. The versatility of our approach is evident in its potential to support a network of autonomous vehicles interacting with multiple road users, thereby enhancing scalability. By leveraging the power of machine learning, our solution provides a robust and adaptable approach that can handle the diverse and ever-changing conditions of urban driving scenarios.}, archive = {J_TASE}, author = {Hamid Taghavifar and Chongfeng Wei and Leyla Taghavifar}, doi = {10.1109/TASE.2023.3347264}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {129--140}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Socially intelligent reinforcement learning for optimal automated vehicle control in traffic scenarios}, volume = {22}, year = {2025}, }
• (2025). ADP-based optimal control for discrete-time systems with safe constraints and disturbances. TASE, 22, 115--128. (www)

  @article{ , abstract = {In this paper, a novel adaptive dynamic programming (ADP)-based optimal control method is developed for discrete-time systems subject to constraints and disturbances. Particularly, a safe policy iteration scheme is designed to handle state and input constraints, including both hard and soft constraints, by converting the original policy improvement strategy into a constrained optimization problem with a prescribed state cost function. After that, an actor-critic-disturbance framework is introduced to address the constrained optimal control problem. The robust safety against disturbances is treated as a two-player zero-sum game, where the actor and disturbance neural networks are used to approximate the optimal control input and the disturbance policy, respectively. The convergence property of the proposed algorithm is analyzed, and the multi-step version of the proposed ADP scheme is derived based on this property. Simulation results are demonstrated and discussed to validate the effectiveness and performance of the proposed method. Note to Practitioners—Addressing constraints in optimal control problems is essential for guaranteeing the safe operation of controlled systems. However, conventional ADP algorithms struggle to simultaneously manage state and control input constraints during the search for the optimal solution. In real-world applications, another critical and common issue is the presence of external disturbances, where disturbances that cause the control object to deviate from the safe region must be constrained while seeking an optimal control policy. Bearing these factors in mind, this study presents a novel ADP scheme for solving optimal control problems of discrete-time systems, taking into account state and control constraints as well as the impact of disturbances. Moreover, the convergence analysis of the proposed SADP scheme is provided, offering a powerful theoretical foundation for guaranteeing the safety and feasibility of the controlled system during operation.}, archive = {J_TASE}, author = {Jun Ye and Hongyang Dong and Yougang Bian and Hongmao Qin and Xiaowei Zhao}, doi = {10.1109/TASE.2023.3346876}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {115--128}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {ADP-based optimal control for discrete-time systems with safe constraints and disturbances}, volume = {22}, year = {2025}, }
• (2025). An attribution feature-based memetic algorithm for hybrid flowshop scheduling problem with operation skipping. TASE, 22, 99--114. (www)

  @article{ , abstract = {An actual hybrid flow shop scheduling (HFSS) problem with operation skipping is investigated from the steelmaking continuous casting (SCC) process, which plays a vital role in the productive processing of a slab in iron and steel enterprises. Firstly, a mixed integer mathematical model is explored for this problem based on the previous survey. Secondly, a block heuristic method based on double-layer right shift is presented on the basis of the problem-specific characteristics, which can generate the better initial solution for the problem. Thirdly, an improved memetic algorithm (IMA) with double-vector-representation based on attribution feature is proposed for dealing with the problem, which includes the block heuristic for initialization, the novel mutation structure on the basis of the problem-specific characteristics, and the enhanced local research to improve the exploitation ability. Finally, to test the performance of the IMA, a large number of instances from a steel plant are adopted. By statistical analysis, the results of experiment evaluation indicate that the proposed IMA has highly effective performance and obvious advantages comparing with other well-known algorithms. Note to Practitioners—Due to the change of order and the technology requirements of SCC process, partial charges must be further refined to eliminate the impurities or heat molten steel. This study models a novel hybrid flow shop scheduling problem with operation skipping, in which all charges undergo the primary refining stage, and partial ones undergo the double or triple refining stage. The average sojourn time, the total earliness and the total tardiness are taken as the objective functions, and the constraint for operation skipping is added. We develop an IMA based on the attribution feature of charge, in which double-vector-representation is designed, and a novel mutation structure on the basis of the problem-specific characteristics is presented, and an enhanced local research is proposed to improve the exploitation ability. Furthermore, a block heuristic method with double-layer right shift is presented to generate good initial individuals. The performance of IMA is analyzed by comparing with six modern meta-heuristics algorithms, the results show that the IMA is more superior than others. Because of the complex of scheduling problem in SCC, some dynamic influence, such as machine breakdowns, the influence of transportation tools, should be considered. The paper’s work can be extended to the above actual dynamic problem. Moreover, the presented IMA can also be developed to other hybrid flow shop scheduling problem with operation skipping.}, archive = {J_TASE}, author = {Yang Yu and Quan-Ke Pan and Xinfu Pang and Xiaochu Tang}, doi = {10.1109/TASE.2023.3346446}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {99--114}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An attribution feature-based memetic algorithm for hybrid flowshop scheduling problem with operation skipping}, volume = {22}, year = {2025}, }
• (2025). Real-time 2D/3D registration via CNN regression and centroid alignment. TASE, 22, 85--98. (www)

  @article{ , abstract = {Registration of pre-operative 3D volumes and intra-operative 2D images is critical for neurological interventions. In various 2D/3D registration tasks, deep learning-based approaches have become popular and achieved tremendous success. However, due to vast space of transformation parameters, estimation errors are significant in these approaches. To tackle above issues, a novel learning-based framework for 2D/3D registration is proposed, consisting of CNN regression and centroid alignment. The former introduces a residual regression network (Res-RegNet) to preliminarily estimate transformation parameters. To further reduce estimation errors, the latter utilizes target vessel centroids to refine projected images. The proposed framework is individually trained and evaluated on three patients, reaching mean Dice of 76.69%, 78.51%, and 85.39%, respectively, all outperforming baseline methods. Extensive ablation studies demonstrate centroid alignment can significantly improve registration performance. As a normalization layer in Res-RegNet, SPADE can modulate activations using binarized inputs through a spatially-adaptive, learned transformation. Semantic information of inputs is preserved to learn better representations for parameter estimation. Moreover, the inference rate of our framework is about 21 FPS combined with the state-of-the-art segmentation model, significantly surpassing real-time requirements (6 $\sim $ 12 FPS) in clinical practice. These promising results indicate the potential of the framework to facilitate various 2D/3D registration tasks. Note to Practitioners—This paper was motivated by the problem of image-guided neurological interventions. Existing 2D/3D registration methods suffer from 1) long iteration times, which are difficult to meet real-time clinical necessities, or 2) significant parameter estimation errors, leading to poor registration accuracies. Therefore, this paper suggests a new registration framework, combining with CNN regression to give predictions of transformation parameters via a single forward propagation, and centroid alignment to reduce estimation errors by translation transformation. The framework is trained and tested on three patients separately and achieves state-of-the-art performance, demonstrating its superiority. Furthermore, the proposed framework is a learning-based method that is adaptable to various image modalities. Therefore, it has latent capacities to be integrated into surgical navigation systems.}, archive = {J_TASE}, author = {De-Xing Huang and Xiao-Hu Zhou and Xiao-Liang Xie and Shi-Qi Liu and Zhen-Qiu Feng and Zeng-Guang Hou and Ning Ma and Long Yan}, doi = {10.1109/TASE.2023.3345927}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {85--98}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Real-time 2D/3D registration via CNN regression and centroid alignment}, volume = {22}, year = {2025}, }
• (2025). A feature extraction approach over workpiece point clouds for robotic welding. TASE, 22, 75--84. (www)

  @article{ , abstract = {The programming of welding trajectories hinders the development of welding robots. With the development of machine vision, the welding path planning based on 3D vision can eliminate the teaching and programming work of the robot. However, the current weld extraction research relies on the geometric information of the workpiece, resulting in the low generality and robustness of the existing methods. The edge and corner features of the point cloud are the basis for extracting welds. Therefore, in view of the above problems, this paper proposes a method for edge and corner extraction of workpieces applied to 3D vision welding. First, the LOBB feature descriptor is proposed to build a feature space for the workpiece point cloud. Then, to enhance the robustness of feature clustering, nonlinear activation is performed on the feature description. Finally, a hierarchical clustering method based on K-means is designed to achieve the extraction of crease points, boundary points, and corner points. Experiments show that the method in this paper can not only extract the edges of various workpieces completely and with low noise, but also can identify corners efficiently, which is overall better than the existing methods. Note to Practitioners—Welding trajectory programming has been a bottleneck in the development of welding robots. However, advancements in machine vision have opened up new possibilities for welding path planning using 3D vision, which can eliminate the need for manual teaching and programming of robots. Unfortunately, existing methods for weld extraction from point clouds rely heavily on geometric information, resulting in limited applicability and robustness. To address these challenges, this paper proposes a novel method for extracting edge and corner features from workpieces using 3D vision in welding applications. The key contribution is the introduction of the LOBB feature descriptor, which creates a feature space for the workpiece’s point cloud. By applying nonlinear activation to the feature description, the robustness of feature clustering is enhanced. Additionally, a hierarchical clustering method based on K-means is designed to extract crease points, boundary points, and corner points. Experimental results demonstrate that the proposed method can effectively extract edges from various workpieces with minimal noise and efficiently identify corners. Overall, the method outperforms existing approaches in terms of completeness and accuracy of edge extraction. This advancement holds significant potential for practitioners in the field of welding robotics, as it reduces the programming complexity and improves the reliability of welding robots in real-world applications.}, archive = {J_TASE}, author = {Yuankai Zhang and Yusen Geng and Xincheng Tian and Fuquan Zheng and Yong Jiang and Min Lai}, doi = {10.1109/TASE.2023.3345868}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {75--84}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {A feature extraction approach over workpiece point clouds for robotic welding}, volume = {22}, year = {2025}, }
• (2025). Online geometric memory generation and maintenance for visuomotor navigation in structural dynamic environments. TASE, 22, 62--74. (www)

  @article{ , abstract = {Autonomous navigation substantially depends on memory mechanisms for enhancing path optimality. However, the dynamic nature of environments can cause inconsistencies between the stored scene memory and real-time perception, leading to potentially catastrophic navigation errors. Existing memory structures often fall short in addressing these challenges, as they typically only account for object-level dynamics and falter when faced with long-term alterations in environmental structure. To counter these constraints, this paper presents a learning-based framework designed to construct and maintain a geometric memory, thereby facilitating improved visuomotor navigation under structural dynamics. This proposed framework employs visual inputs to construct a geometric representation of the environment, and identifies structural changes by assessing the consistency between the established memory and instantaneous perception. To update the geometric memory efficiently, we introduce a memory updater grounded in a structure storage pool. Furthermore, a two-phase hierarchical planner is proposed to decompose navigation tasks and formulate smooth navigation strategies. Experimental results from photorealistic simulations underscore the efficacy of the proposed system in managing long-term dynamics and navigation control. The effectiveness of the proposed system is further corroborated through deployment and testing in real-world environments. Note to Practitioners—Classic geometry-based navigation methods hinge on the construction of a global map for spatial reasoning and optimized robot control. Within a learning-based pipeline, the dependence on memory information becomes critical for enabling robots to develop spatial and temporal awareness. While the maintenance of memory structures can significantly expand the field of the robot’s perception in time and space, discrepancies between historical memory and real-time perception in dynamic environments can lead to misguided decisions. Unlike most existing research that addresses short-term dynamic issues at the object level, this paper centers on long-term, large-scale dynamics precipitated by changes in environmental structure. We put forward a learning-based framework that incorporates dynamic perception, map maintenance, and hierarchical navigation. The experimental results highlight the efficiency and real-time processing capability of this method in handling structural dynamics, hence enhancing navigation efficiency.}, archive = {J_TASE}, author = {Qiming Liu and Neng Xu and Zhe Liu and Hesheng Wang}, doi = {10.1109/TASE.2023.3344771}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {62--74}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Online geometric memory generation and maintenance for visuomotor navigation in structural dynamic environments}, volume = {22}, year = {2025}, }
• (2025). Distributed reduced-order observer-based consensus of multiple euler-lagrange systems over switching networks. TASE, 22, 53--61. (www)

  @article{ , abstract = {In this paper, we investigate leader-follower consensus (LFC) of multiple Euler-Lagrange systems (MELSs) under the case that followers can only measure partial output information of leader system, and the communication network among followers is switching and jointly connected. Firstly, in order to accurately estimate complete state information of the leader system in real time, we propose a distributed reduced-order observer, which is valid over a jointly connected switching network. Then, by applying this distributed reduced-order observer, we further design an observer-based controller for multiple Euler-Lagrange systems (MELSs) to reach leader-follower consensus (LFC). Distinct with the existing controllers for LFC of MELSs, this controller is robust for bounded external disturbances, and independent of the structure and features of Euler-Lagrange (EL) system. At last, a simulation example is given to confirm the effectiveness of the proposed distributed reduced-order observer and robust controller. Note to Practitioners—Many practical engineering applications (such as spacecrafts, mobile robot manipulators) can be modeled as multiple uncertain Euler-Lagrange systems, where each agent suffers external input disturbances, and only a portion of followers can measure partial components of leader’s output. However, most relevant works assumed that the dynamic information of each agent is completely known, agents do not suffer external input disturbances, and all follower agents can measure full-dimensional output of leader. How to achieve LFC for such MELSs has become a main focus of control researches. Thus, this paper proposes an adaptive distributed reduced-order observer-based consensus algorithm for such MELSs, whose communication network is switching and jointly connected. Finally, the effectiveness of the proposed algorithm is illustrated by a simulation example.}, archive = {J_TASE}, author = {Mingkang Long and Housheng Su}, doi = {10.1109/TASE.2023.3339888}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {53--61}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Distributed reduced-order observer-based consensus of multiple euler-lagrange systems over switching networks}, volume = {22}, year = {2025}, }
• (2025). The demand-side management and control of smart grids based on weighted network congestion games. TASE, 22, 43--52. (www)

  @article{ , abstract = {Based on the semi-tensor product of matrices, this article considers the demand-side management and state-flipped control of smart grids. Firstly, the smart grids with single-layer and multi-layer power companies can be modelled by a potential game under some specific conditions. Secondly, in the evolutionary process with single-layer power companies, a state-flipped control is proposed such that the evolutionary game converges to the optimal Nash equilibrium. An algorithm is constructed to reduce the control cost and the flipped node set will be found. Then, the corresponding flipped node subsets can be obtained to make as few evolution steps as possible. Finally, the correctness of the theoretical results is verified by numerical simulations. Note to Practitioners—This paper has great practical applications in the demand-side management of smart grids. Considering different loads of each player, we use a weighted network congestion game to model the smart grids. The weights in the game are specific to each player and represent their congestion impact, which is more practical in engineering problems. Then we search for the conditions that this kind of game is a potential game and hence has Nash equilibria. This starting point is different from the existing works. Furthermore, we propose a novel state-flipped control to intervene the game process and make it converge to the optimal objective. Compared with the existing theoretical results, the calculation is more convenient and the Nash equilibrium can be easily found. This paper shows an effective method to solve the demand-side management of smart grids in practice and also provides ideas for other engineering problems in such as resource allocation.}, archive = {J_TASE}, author = {Jinhuan Wang and Min Liu and Han Wu}, doi = {10.1109/TASE.2023.3336640}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {43--52}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {The demand-side management and control of smart grids based on weighted network congestion games}, volume = {22}, year = {2025}, }
• (2025). An optimal synchronization control method of PLL utilizing adaptive dynamic programming to synchronize inverter-based resources with unbalanced, low-inertia, and very weak grids. TASE, 22, 24--42. (www)

  @article{ , abstract = {When it comes to integrating inverter-based resources (IBRs) into modern grids with varying characteristics like unbalanced systems, low-inertia networks, or very weak grids, synthesizing the synchronization control method (SCM) of the IBR’s phase-locked loop can be a challenging task. This paper provides a unique solution to enhance the three-phase IBR’s SCM using the adaptive dynamic programming (ADP) method based on reinforcement learning. By making the SCM more intelligent and self-learning, IBRs can be easily integrated into diverse grids. To this end, this article investigates the synchronization process’s detailed dynamics, including all incorporating disturbances and parameters required for the first step in designing the ADP method. Afterward, this research synthesizes an optimal controller using an ADP method. It is a data-driven and practically sound approach to the problem under investigation. The new methodology is based on the adaptive optimal control employing measurement feedback to control the output regulation problem of uncertain synchronization process dynamics via the internal model principle. The proposed SCM design deploys an ADP learning methodology to tackle uncertain parameters and unknown disturbance signals to synchronize IBRs during transients, thereby enhancing IBRs’ synchronization in challenging conditions of modern power systems with unbalanced, low-inertia, and very weak grids. For comparison purposes, this paper applies a robust controller based on the well-established $\mu $ synthesis approach (benefiting from the well-known $D{-}K$ iteration process). Comparative simulations are performed; experiments are conducted to reveal the effectiveness and practicality of the ADP-based optimal SCM proposed in this paper. Note to Practitioners—As different nations strive to combat global warming and accelerate decarbonization, power and energy systems are undergoing a significant shift. Inverter-based resources are being used as an essential component to achieve these goals. However, studies have revealed that designing synchronization control methods of the inverter-based resources’ phase-locked loop in unbalanced, low-inertia, and very weak grids is challenging due to the need for accurate dynamic models and other factors. This study revisits the synchronization process’s detailed dynamics. It also proposes a novel adaptive dynamic programming strategy using intelligent self-learning approaches to the synchronization control method associated with inverter-based resources. This method utilizes an optimal control to synthesize the adaptive dynamic programming control strategy for the inverter-based resources’ synchronization process. Besides, it employs measurement feedback to control the output regulation problem of uncertain dynamics of inverter-based resources’ synchronization process via the internal model principle. As a result, this paper makes this process data-driven. It utilizes a learning methodology using adaptive dynamic programming to address uncertain parameters and unknown disturbance signals associated with the dynamics derived and formulated for the problem under investigation. Thus, the proposed method applies to controlling inverter-based resources’ synchronization process even in cases with slow parameter variations caused by different factors. It can compensate for all functional disturbance signals affecting the dynamics of the systems. In fact, unlike traditional methods that need an exact dynamic model of the inverter-based resources’ synchronization process to design and tune the controller to achieve a proper transient response, the proposed control system trains itself and does so. This study’s simulations and experiments reveal that the above points give the proposed approach a competitive edge over the existing methodologies.}, archive = {J_TASE}, author = {Masoud Davari and Weinan Gao and Amir Aghazadeh and Frede Blaabjerg and Frank L. Lewis}, doi = {10.1109/TASE.2023.3329479}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {24--42}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {An optimal synchronization control method of PLL utilizing adaptive dynamic programming to synchronize inverter-based resources with unbalanced, low-inertia, and very weak grids}, volume = {22}, year = {2025}, }
• (2025). Learning-based optimal cooperative formation tracking control for multiple UAVs: A feedforward-feedback design framework. TASE, 22, 11--23. (www)

  @article{ , abstract = {Notwithstanding the successful design of state-of-the-art cooperative control protocols to accomplish formation tracking for multiple unmanned aerial vehicles (UAVs), the assurance of performance optimality cannot be guaranteed in the face of complex disturbances affecting these multi-UAV systems. In order to surmount this challenge, this research endeavor aims to establish a feedforward-feedback learning-based optimal control methodology to facilitate cooperative UAV formation tracking in the presence of intricate disturbances. To be more precise, by leveraging backstepping-based feedback control, the problem of UAV formation tracking is transformed into an equivalent optimal regulation problem. Consequently, a learning-based feedforward control scheme is devised, wherein the cooperative policy iteration algorithm is formulated based on a two-player zero-sum game. The critic-only echo state network (ESN) is employed to approximate the optimal feedforward control policies, with the inclusion of an online adaptive tuning law and compensation terms to alleviate the persistence of excitation condition and eliminate the need for an initial admissible control. As a result, the closed-loop stability is guaranteed in terms of uniformly ultimately boundedness for tracking errors and ESN weights. Note to Practitioners—In real-world scenarios, the flight of multiple UAVs is invariably affected by intricate disturbances, resulting in compromised tracking precision. There is an urgent need to enhance resistance to disturbances and ensure optimal performance for cooperative formation tracking of multiple UAVs. Beyond the capabilities of model-based controllers, the integration of reinforcement learning has shown promise in achieving robust control actions. By introducing the cooperative policy iteration algorithm based on a two-player zero-sum game, the tracking performances of UAV formation can be further optimized. In order to facilitate the practical application of reinforcement learning in UAV systems, our proposed algorithm addresses the persistency of excitation condition by incorporating innovative compensation terms into the ESN tuning law. Furthermore, we resolve the requirement for initial admissible control by introducing a novel piecewise compensation term into the ESN tuning law, which is based on a newly proposed Lyapunov function.}, archive = {J_TASE}, author = {Boyang Zhang and Maolong Lv and Shaohua Cui and Xiangwei Bu and Ju H. Park}, doi = {10.1109/TASE.2023.3322028}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {11--23}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Learning-based optimal cooperative formation tracking control for multiple UAVs: A feedforward-feedback design framework}, volume = {22}, year = {2025}, }
• (2025). Attention-based spatiotemporal graph fusion convolution networks for water quality prediction. TASE, 22, 1--10. (www)

  @article{ , abstract = {In many fields, spatiotemporal prediction is gaining more and more attention, e.g., air pollution, weather forecasting, and traffic forecasting. Water quality prediction is a spatiotemporal prediction task. However, there are several challenges in water quality prediction: 1) Water quality time series has a complex nonlinear relationship, making it difficult to predict; 2) Water quality sensors are distributed on the river networks and have a strong spatial dependence on water quality prediction; and 3) Poor long-term forecast accuracy. To solve these problems, this work proposes a spatiotemporal prediction model called a Fusion Spatio-temporal Graph Convolution Neural network (FSGCN). First, This work uses a temporal attention mechanism to solve the nonlinear problem of water quality time series. Second, It adopts a graph convolution to extract spatial dependencies of river networks, and the fusion of spatiotemporal can more easily capture spatiotemporal features. Third, it adopts a temporal convolution residual mechanism, improving long-term series prediction accuracy. This work adopts two real-world datasets to evaluate the proposed FSGCN, and experiments demonstrate that FSGCN outperforms several state-of-the-art methods in terms of prediction accuracy. Note to Practitioners—This work considers the critical problem of spatiotemporal water quality prediction. Accurate water quality prediction can effectively prevent environmental pollution. Traditional water quality prediction only focuses on time series features without considering spatial features. In this work, a novel spatiotemporal prediction approach is proposed that combines spatial-temporal graph fusion construction networks for water quality time series prediction in a real-time manner. This work shows that this approach can achieve longer forecasting sequences and more accurate results than traditional forecasting methods. As a practical consequence of this research, spatiotemporal graph fusion convolution networks for water quality prediction can effectively integrate multi-dimensional data and improve the accuracy of long-term water quality prediction. This approach can also be applied to other fields, including intelligent transportation, smart manufacturing, finance, the Internet of Things, and urban computing.}, archive = {J_TASE}, author = {Junfei Qiao and Yongze Lin and Jing Bi and Haitao Yuan and Gongming Wang and MengChu Zhou}, doi = {10.1109/TASE.2023.3285253}, journal = {IEEE Transactions on Automation Science and Engineering}, pages = {1--10}, shortjournal = {IEEE Trans. Autom. Sci. Eng.}, title = {Attention-based spatiotemporal graph fusion convolution networks for water quality prediction}, volume = {22}, year = {2025}, }